Out of the Margins: The Rise of Chinese Vernacular Fiction

OUT OF THE MARGINS The Rise of ChineseVernacular Fiction

Liangyan Ge

Out of the Margins

Out of the Margins The Rise of Chinese Vernacular Fiction

Liangyan Ge

University of Hawai‘i Press Honolulu

© 2001 University of Hawai‘i Press All rights reserved Printed in the United States of America 01 02 03 04 05 06

6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data Ge, Liangyan. Out of the margins : the rise of Chinese vernacular fiction / Liangyan Ge. p. cm. Includes bibliographical references. ISBN 0-8248-2370-2 (alk. paper) 1. Shui hu zhuan. I. Title: Rise of Chinese vernacular fiction. II. Title. PL2694.S53 G4 2001 895.1'346—dc21 2001027357 University of Hawai‘i Press books are printed on acid-free paper and meet the guidelines for permanence and durability of the Council on Library Resources. Designed by Carol Colbath Printed by The Maple-Vail Book Manufacturing Group

Contents

Acknowledgments

vii

Note on Chinese Romanization

ix

Introduction

1

1. Vernacularization before Shuihu zhuan

10

2. Told or Written: That Is the Question

36

3. The Narrative Pattern: The Uniform versus the Multiform

64

4. From Voice to Text: The Orality-Writing Dynamic

101

5. The Engine of Narrative Making: Audience, Storytellers, and Shuhui xiansheng

144

6. Literary Vernacular and Novelistic Discourse

179

Notes

199

Glossary

245

Selected Bibliography

261

Index

283

Acknowledgments Although it has lived through a number of incarnations, Out of the Margins has its origin in my Ph.D. dissertation at Indiana University in the early 1990s. Eugene Eoyang, who provided the guidance for my graduate studies, has played a major role in my intellectual development by letting me share his vast knowledge and wisdom over the years. To him, I am deeply indebted. I am also grateful to Sandra K. Dolby, William Hansen, H. James Jensen, Oscar Kenshur, and Yingjin Zhang for their help and advice during the years of my graduate education. My thanks also go to Irving Lo, from whom I received not only academic advice but also personal care that was almost parental. I benefited greatly from the many suggestions by the late Clifford Flanigan, whose premature death robbed me of a teacher and a friend. Since 1995 I have been teaching at the University of Notre Dame. I am grateful to my colleagues, especially Michael Brownstein, Lionel Jensen, Peter Moody, Dian Murray, Lili Selden, Margaret B. Wan, and Xiaoshan Yang, for their support and friendship, which have given me an environment conducive to the furtherance of my research. I thank the Notre Dame Graduate School for a Faculty Research Grant in 1998, which facilitated my project. Thanks are also due the College of Arts and Letters and the East Asian Department for a one-semester research leave. I owe special thanks to the Notre Dame Library’s Office of Interlibrary Loans for indulging my numerous requests for books and articles. Among those who read my manuscript, David L. Rolston spent more time on it than it deserved, offering incredibly detailed comments and sparing me numerous embarrassing errors. Robert E. Hegel, despite his tight schedule at the time, examined my scribbles and offered many helpful suggestions for improvement. Victor Mair’s comments on the manuscript and his warm encouragement proved crucial for the completion of the work. To them, I am profoundly grateful. I also wish to extend my sincere thanks to other leading scholars in the field of Chinese vernacular literature, especially Patrick Hanan, Wilt L. Idema, and Andrew H. Plaks, whose brilliant scholarship has been a source of inspiration for my study. During my trip to China in 1998, I bene-

vii

viii

Acknowledgments

fited from stimulating conversations with Professors Zhang Peiheng and Han Jiegen of Fudan University. Thanks are also due the two anonymous readers at the University of Hawai‘i Press for their comments on my work and suggestions for further revision. Needless to say, all remaining faults and shortcomings are mine. Yaohua Shi (now at University of Massachusetts) and Zhijie Jia (now at Harvard) read an early version of my manuscript and shared with me their views on it. Jian Liu and Xiaoying Hu of Indiana University Library helped me gain access to various research materials. To these friends of mine, and many others, I am deeply thankful. Several editors associated with the University of Hawai‘i Press—Sharon Yamamoto, Pamela Kelley, Lee S. Motteler, Masako Ikeda, and Cheri Dunn— were involved in the production of this book at different stages. I thank all of them for their great help, which has been instrumental in bringing this work to print. Finally, let me thank my wife Yongqing and my daughter, Sherry, who gave me the warmth of both heart and hearth through the years. I am eternally grateful to my departed parents. By example, they showed me the ways to work persistently and relentlessly, long before I started my own career. To the memory of my parents, this book is dedicated, with love and piety.

Note on Chinese Romanization In this book the pinyin system is used for Chinese romanization. For citations from sources where the Wade-Giles system is used, all Chinese names and terms in the text are converted to pinyin for the sake of consistency. The Wade-Giles romanization for titles of books and articles, however, remains unchanged. I beg the original authors for their understanding.

ix

Introduction In both the West and the East, the relationship of writing to speech has a direct bearing on literature, especially narrative literature. Around the eighth century b.c., the Greeks adopted the writing shapes from the Phoenician syllabary and invented what is known to be the earliest alphabetic system. The advent of the Greek writing system marks the beginning of a new era of Western civilization, with written literature as one of the most immediate results. The Homeric epics, hitherto existing only orally, now became written, although there is no consensus how that was accomplished. The significance of this process has been most succinctly summarized by Eric Havelock in the title of one of his books, The Muse Learns to Write. Writing, as the newly arrived visitor, knocked at the door of the oral world and was hospitably received by the Muse, daughter of the goddess of Memory and mistress of oral literature. Only with the integration of the old and the new did it become possible for the Homeric verses to appear as literature, in the literal sense of the word. In a literate age, there is another kind of relationship between writing and orality that has received less scholarly attention. The positions of the old and the new are now reversed. Writing, by nature resistant to changes, can become the agent for what is old and conservative. It necessarily lags behind the more dynamic and fluid development of orality and therefore only represents yesterday’s language. Once the gap grows excessively wide, writing becomes senile and decrepit, to be rejuvenated only by a drastic dose of orality. On the other hand, the rapidly changing oral speech persistently seeks new surrogates in written script. Voice’s never-ending quest for a chirographical body results in constant reinvigoration of writing. Such changes in writing have been a driving force behind the evolution of literary forms. Very often, when the linguistic medium in literature was realigned with the latest living speech, new genres and styles were born, keeping literature fresh and vigorous. Histori-

1

2

Introduction

cally, therefore, what happened between writing and orality was by no means a one-way street. If the Muse had to learn to write in order to bring ancient Greece out of the preliterate “dark age,” the forces of literate culture afterwards also had to draw vitality from spoken words: The Muse-turned-writer had to return to the world of orality and learn—or relearn—to “babble.” In the history of Western literature, a significant moment of the writer learning to babble was in the thirteenth and fourteenth centuries, when men of letters in Europe who had written exclusively in Latin—which was predominantly a literary language seldom spoken out of the church—turned to orality and brought writing more in line with speech. The results were vernacular works such as Giovanni Boccaccio’s Decameron and Geoffrey Chaucer’s Canterbury Tales, which heralded the ascendance of European national literatures written in the vernaculars. In China, due to particular social and cultural conditions and especially to the unique nature of the Chinese script, vernacularization was a much longer and more arduous process. While the bulk of Chinese literature up to the end of the nineteenth century was written in classical Chinese, or wenyan, vernacular literature had existed stubbornly and persistently for many centuries. Yet the written vernacular, or baihua, did not become a full-fledged literary language until the emergence of voluminous novels in the Ming period (1368–1644), especially the works generally known as the “Four Masterworks” (si da qishu), namely Shuihu zhuan (Water Margin/ Outlaws of the Marsh), Sanguo yanyi (Romance of the Three Kingdoms), Xiyou ji ( Journey to the West), and Jin ping mei (Golden Lotus/The Plum in the Golden Vase). To be sure, the dominance of wenyan was to continue in most of China’s cultural and literary sectors until the early twentieth century, when it was finally replaced by the vernacular as the nation’s standard written language—but in narrative literature that revolutionary change had started a few centuries earlier. After the “Four Masterworks,” many narrative pieces, especially short tales such as those by Pu Songling (1640–1715) and Ji Yun (1724–1805), continued to be written in the classical language, but the use of the vernacular became increasingly a genre convention in prose fiction, especially in the novel.1 Among the “Four Masterworks,” it was Shuihu zhuan that played the most crucial role in establishing the written vernacular as the new literary language. Two separate pages from an early edition of Shuihu zhuan, with the title of Jingben Zhong yi zhuan, were found accidentally in the Shanghai Library in 1975. Most scholars agree that the fragments are from a Zhengde (1506–1521)-Jiajing (1522–1566) edition, if not earlier, possibly the earliest among all extant editions of Shuihu zhuan, full or incomplete. From a different Jiajing edition entitled Zhong yi Shuihu zhuan, five chapters are extant, previously in the possession of Zheng Zhenduo (1898–1958) and now housed in the Beijing Library.

Introduction

3

The earliest known reference to a printed edition of Shuihu zhuan occurs in the catalogue entitled Baichuan shuzhi, compiled by the Jiajing scholar Gao Ru, which lists a hundred-chapter Zhong yi Shuihu zhuan.2 Whether it was of the same edition as the Zhong yi Shuihu zhuan fragments in the Beijing Library remains unknown. Since Baichuan shuzhi carries a preface dated 1540 (the nineteenth year of the Jiajing reign), we can safely mark 1540 as the latest possible year for the earliest edition of Shuihu zhuan. The earliest reference to Jin ping mei was made by Yuan Hongdao (1568–1610) in his letter dated 1596 to the famous painter and calligrapher Dong Qichang (1555–1636). It is indicated in the letter that Yuan had obtained part of a manuscript copy of the novel from Dong.3 During the following two decades, manuscript copies of the work may have been circulating among literati readers, as Shen Defu (1578–1642) reports in Wanli yehuo bian that he had copied the entire manuscript from Yuan Zhongdao (1570–1623), Hongdao’s younger brother, in 1609.4 The work’s earliest known edition in print is Jin ping mei cihua, with one of the two prefaces dated 1617–1618 by a pseudonymous Dongwu nongzhu ke. The year 1617 is therefore the earliest possible date for the edition, which was antedated by the Jiajing editions of Shuihu zhuan by at least two-thirds of a century. When Jin ping mei was written, the writer may have looked to Shuihu zhuan for a model, as evidenced by the fact that the story in Jin ping mei itself sprouts from an episode in the earlier work. Both Xiyou ji and Sanguo yanyi had a long process of textual evolution similar and, for the most part, temporally parallel to that of Shuihu zhuan. However, extant fragments from budding textual precursors notwithstanding, the earliest exemplar of Xiyou ji in its fully developed form—the Shidetang edition— is dated 1592, at least half century later than the Jiajing editions of Shuihu zhuan. The earliest known edition of Sanguo yanyi, entitled Sanguo zhi tongsu yanyi, carries two prefaces, one by Jiang Daqi and the other by Zhang Shangde, which are dated 1494 and 1522 respectively. In all likelihood, the edition was contemporaneous with or slightly predated its Shuihu counterpart.5 Yet while Shuihu zhuan appears in a language that is distinctly vernacular, Sanguo yanyi is called a vernacular novel only when the word vernacular is more generously defined, as it employs a linguistic medium that mingles baihua with simple wenyan.6 Shuihu zhuan is, therefore, China’s earliest full-length fictional narrative in the true vernacular prose. If we take into account the fact that the earliest anthology containing vernacular short stories, Liushi jia xiaoshuo (Sixty Short Stories) was not published until around 1550,7 indeed we may call Shuihu zhuan the trailblazer for Chinese vernacular fiction at large. To say this, of course, is not to repudiate the efforts in vernacularization prior to Shuihu zhuan; on the

4

Introduction

contrary, as we will see, Shuihu zhuan is to be considered in the present study as part of a larger vernacularizing process. It is in Shuihu zhuan, however, that vernacular prose extends to an unprecedented length and the degree of vernacularity ascends to an unprecedented level. Writing and speech, divorced in wenyan, are now brought much closer to each other in the full-fledged vernacular prose of this novel. The last statement may call for a little qualification here. Chinese characters, whether employed in wenyan or baihua, are not a phonetic system of writing like alphabetic scripts. The peculiar relationship of Chinese characters to speech has long been a focus of the scholarly debate. The terminology that has been applied to Chinese characters has almost become a nomenclatorial kaleidoscope. They are called variously “sinographs,” “pictographs/pictograms,” “ideographs/ideograms,” “logographs,” “lexigraphs,” “morphographs,” “phonograms,” “phonosemantics,” “phonoideograms,” “logo-syllabics/word-syllabics,” and many others.8 This messy situation reflects the formidable difficulty of defining the Chinese script in terms of its relationship to the semantic and phonetic aspects of the language. Contending that no writing system represents ideas without regard to sound, John DeFrancis argues that terms such as “ideographic” should be consigned to “the Museum of Mythological Memorabilia along with unicorn horns and phoenix feathers.” DeFrancis designates the Chinese script, which he believes to be basically phonetic, as “morphosyllabic,” but even he has to admit that “the Chinese syllabary is only partially reliable in representing the pronunciation of the Chinese characters.”9 More recently, William Hannas tries to reconcile the “aphonic” and the phonetic views of the Chinese writing system by proposing that a character is part of a morpheme, which he defines as the conventional interface between meaning and a phonological form.10 Hannas, too, is quite aware of the inefficiency and inaccuracy of the phonetic denotation by the characters. “Lacking a systematic relationship between symbol and sound, or even an efficient protocol for describing the structure of its units,” Chinese writing “places an enormous burden on the user in comparison to alphabetic systems,” as Hannas puts it.11 The written vernacular, which inherits many—in fact most—of its characters from classical Chinese, may be no more phonetic than the latter. But what is relevant here is not each written symbol’s representation of an isolated sound, but the correspondence between a string of written symbols and a string of sounds in a speech utterance. A character in wenyan, of course, was and is pronounceable, and one of the most important pedagogical methods in traditional Chinese education was making students read aloud writings in wenyan until they could recite them from memory. Most of the characters, the so-called xingshengzi (semantic-phonetic composites), even have a component as a cue

Introduction

5

for pronunciation.12 But if one reads aloud a line of characters in typical wenyan, the string of sounds will usually represent a highly stylized utterance. On the other hand, the written vernacular can be much more accommodating to common speech, although it can be stylized in different ways as well. Characters in the written vernacular can therefore correspond to speech sounds, usually one character to one syllable, although each character is not phonetic in the same way as a letter or a syllabogram in an alphabetic script. It is in this sense that we can say that writing was finally realigned with the speaking voice in Shuihu zhuan, bringing to fruition the repeated and tenacious attempts of vernacularization for several centuries. This book is devoted to that historic moment when written vernacular prose—the product of a gradual acclimatization of writing to speech—became established in Shuihu zhuan as the new literary language for Chinese narrative literature. The argument here is that Shuihu zhuan, in its fullest and artistically most complex form, is the product of long-term interaction of oral and written traditions, reflective of the fascination with oral language used by generations of professional storytellers. Although the thorny issue of recensions will be addressed, this is not a study of the textual history of Shuihu zhuan punctuated by the different fanben (full-version) and jianben (simple-version) editions in post-Jiajing times. Rather, it is more about the poesis of the text, or the pre-Jiajing life of the narrative, before the fanben-jianben scramble ever had a chance to start. Chapter 1 is a review on the long and arduous process of vernacularization in literary prose during the few centuries prior to the fanben editions of Shuihu zhuan. Vernacular texts associated with various oral and performative genres in different historical periods will be discussed. Such a survey is necessary, as it informs us of the general linguistic conditions for the evolution of Shuihu zhuan, in both oral and textual forms. As the discussion will demonstrate, what happened to many other works in early vernacular literature may have been quite similar to the cumulative textualization of Shuihu zhuan. This makes the evolution of the Shuihu zhuan not an anomaly but representative of the general trend of vernacularization in literary prose. In chapter 2, efforts will be made to reclaim at least part of the formative process of the narrative in Shuihu zhuan. A number of oral and performative genres and their possible roles in the evolution of the narrative will be discussed. After that, the chapter will provide a discussion of the historical and contemporary critical views on the formation of the narrative, focusing on the divergent opinions on the relationship of the narrative discourse to the oral tradition. I will argue against the either-or binary logic that often lurks behind the divergent critical judgments on Shuihu zhuan’s connection to popular orality.

6

Introduction

And I will propose that Shuihu zhuan, as the earliest vernacular novel wrought by the long-term orality-writing interplay, should be considered both a work of oral provenance and a great literary innovation by men of letters. The subsequent chapters will substantiate that proposition. Chapter 3 discusses the oral provenance of the narrative discourse in Shuihu zhuan. One conceivable way to determine the oral derivation of the narrative would be to track down in reverse chronological order all its textual precursors, but that approach is obviously unfeasible, as the only extant text of Shuihu narrative before the fanben editions of Shuihu zhuan itself is the sketchy account of the Liangshan rebellion contained in Xuanhe yishi (Unrecorded Events of the Xuanhe Period), a historical narrative probably published in the early Yuan dynasty (1279–1368). Everything else was lost, including the earliest stories featuring Shuihu figures that circulated during the Southern Song period (1127–1279). Consequently, it is “difficult to show any direct textual connections to these materials, even to the Xuanhe yishi segments that are sometimes taken as a kind of blueprint for the novel,” as Andrew Plaks rightly points out.13 But the difficulty goes even deeper. Even if we could collect all necessary textual clues that would enable us to trace all the way back to the very earliest textual prototype, we would still have to face the same question about the nature of that version: Was it written or derivative of the oral? In the present study, therefore, what are considered as features inherited from the novel’s oral antecedents will be demonstrated through other means. Chapter 3 examines the thematic patterns of the narrative that are frequently recurrent, with numerous stretches paralleled either within the novel itself or elsewhere in early vernacular literature. The recurrence of the thematic patterns will be discussed in terms of the dialectic between the uniform and the multiform in the economy of the oral mode of story making. A major factor in the twentieth-century study of oral and oral-derived literature is the Parry-Lord theory on the making of oral epics. The universality of the theory has, however, met considerable resistance. One of the controversial issues is the very definition of orality itself. Parry-Lord studies seem to suggest a “pure” type of oral culture—what Walter Ong calls the “primary orality”—as opposed to an equally “pure” type of literate society. A singer of tales, as Albert Lord believed in his earlier works, can never be a literate or semiliterate man; if so, he would cease to be a singer. Therefore, a text is “either one or the other . . . either oral or written.”14 This categorical separation of orality from literacy, where the oral sensibility and the mode of composition-in-performance precludes literate mentality and the mode of writing, may be an accurate description of the Yugoslav oral world, where Lord did the field study, but the absolute polarization of orality to writing may not be univer-

Introduction

7

sally applicable. For instance, Ruth Finnegan, with the support of her fieldwork in the South Pacific, forcefully challenges what she calls the “binary typology” of orality and literacy by showing the possibility of a “mixture” of oral and literate modes of transmission: “For how useful is this binary typology when it turns out that most known cultures don’t fit? In practice a mixture of media (oral and written) is far more typical than a reliance on just one, with writing being used for some purposes, oral forms for others. . . . This kind of mixture is and has been a common and ordinary feature of cultures throughout the centuries rather than the ‘abnormal’ case implied by the ideal types model.”15 Another example of such a “mixed” mode of transmission is the Icelandic saga. The Njáls Saga, for instance, went through a formative period in which the oral and written sources not only existed side-by-side but also alternated and were often convertible to each other.16 This brings us to a conception of the “Chinese-type” of popular orality behind the genesis of the vernacular narrative of Shuihu zhuan. Chinese popular orality was by no means absolutely isolated from the literate culture. Rather it was in constant interaction with writing and with an extremely rich written literature in wenyan. Chapter 4 discusses the long and cumulative process of textualization of the narrative based on the writing-orality dynamic and reciprocity. In a philological analysis of the fanben text of Shuihu zhuan, some stylistic and linguistic features will be examined against a context constituted by other works in early vernacular literature, especially vernacular stories. The examination shows in the fanben text a sedimentary accretion of stylistic and linguistic features typical of different historical periods, which attests to a prolonged and gradual process of amplification of written vernacular prose. What I mean by “textualization” is different from “transcription.” To transcribe an oral narrative is to make a written record of the spoken words by using a preexisting writing script, and it is necessarily notational, as a modern American folklorist would do in collecting “personal narratives.” To “transcribe” the Shuihu narrative complex would suggest that developed and mature written vernacular prose was already available as an established literary language, which was not the case, as we will see in our survey of the pre-Shuihu vernacular works. Not writing in a mature vernacular prose but somehow writing toward it, the effort to turn the Shuihu complex into the novel Shuihu zhuan had to be a long process punctuated with successive written versions, both notational and compositional, each representing a certain point on the axis of transition from voice to print.17 As the cumulative result of such a long process of textualization, the novel may not be directly derivative of any particular single oral presentation. Instead it makes better sense to say that what is registered in print is based on the words from the oral tradition of the Shuihu cycles.

8

Introduction

Shifting from the text analyses in the foregoing chapters to a historical approach, chapter 5 speculates about the context of communication for the oral mode of existence of Shuihu zhuan’s precursors and expounds on some aspects of the work’s narrative formation and value orientation in terms of the raconteur-audience relationship. This chapter will also discuss the role in the textualization of the narrative played by the men of letters who had been relegated to the circles of oral entertainment. In the light of the historical conditions, the textualization of Shuihu zhuan will be considered a social protest as well as a literary innovation. It was with a defiant spirit of rebellion—which was in perfect accord with the subject matter of the narrative itself—that the frustrated men of letters revolted against the craft literacy in the wenyan tradition and anchored the storytellers’ voice in written words. The significance of registering voice in print and its literary ramifications deserves a full recognition. The concluding chapter discusses how the ascendance of literary vernacular completely reoriented China’s narrative literature. As the Shuihu stories became textualized with the storytellers’ dramatization of all the speaking voices of the characters, what we have in print is an exuberance of dialects of various social, professional, and geographical groups. Vis-à-vis wenyan, vernacular prose enjoyed a much-enlarged referential capacity; but more importantly, this stratified linguistic structure became an image of society itself, which is necessarily congested with different types of linguistic-ideological consciousness. Vernacular prose, as the vehicle for realistic representation of the characters’ linguistic experience, makes Chinese vernacular fiction quite compatible with the Western novel in that regard. This book thus considers the evolution of Shuihu zhuan as the most important terrain for the vernacularization of literary prose, which laid the linguistic groundwork for the rise of Chinese vernacular fiction. It does not claim to offer conclusive answers for all those perplexing questions about the inception of Chinese vernacular fiction in general and the formation of Shuihu zhuan in particular. Instead, this is a book suggesting some possibilities that have hitherto been largely overlooked. The ascendance of vernacular fiction during the Ming period was of course an extremely complicated issue. Apart from the maturation of vernacular prose, many other social, cultural, and economic factors were involved—including, for instance, the printing industry, the literacy level, and the emergence of a middle-brow readership. Those, however, are beyond the scope of this book. As we all know, the title of the novel, Shuihu zhuan, literally means “a story that happened on the water margin,” as the lair of the rebels, Liangshan, is on the side of a lake. Hence one of the novel’s English translations is entitled Water Margin.18 The word “shuihu” (water margin) originated in the Shijing (Classic

Introduction

9

of Poetry/Book of Songs). In the poem “Mian,” Danfu, the forefather of the Zhou people, is said to have led a westward exodus along the riverside (shuai xi shuihu) to the foot of Mount Qi, where he settled down and established the base for the future kingdom of Zhou.19 In our novel, the “water margin” indeed seems to be where the rebels belong, for as outlaws they are marginalized from society— politically as well as geographically. As Jin Shengtan expounds in one of his prefaces to the novel, the word “shuihu” refers to a place beyond the waters on the fringe of the emperor’s land, a place from which a member of the mainstream society is supposed to keep away.20 But it is from the “margin” that the bandits, like the self-exiled Danfu, grow into such a formidable force as to challenge the rule of the imperial court, the center of the geopolitical map. What we have said of the bandit heroes on the “margin” can be said of the Shuihu tradition itself as well. As a story complex it existed, for a long period, on the social margins before it crystallized into Shuihu zhuan, which was to be incorporated into the mainstream culture as a masterwork of fiction. What we call Water Margin is therefore a novel that has itself arisen from the margins. Indeed, as we will see in this book, the formative period of Shuihu zhuan was representative of the embryo stage of vernacular fiction at large. From the cultural margins, where the belles lettres interlocked with the popular and writing infiltrated into orality, tremendous energy was accumulated that eventually gave rise to Chinese vernacular fiction as a new and vigorous literary genre.

1 Vernacularization before Shuihu zhuan It is extremely difficult to determine exactly when the writing script of ancient China began to diverge from speech. Bernhard Karlgren estimated that Chinese writing and speech started to part ways at the beginning of the Christian era, which was roughly equivalent to the end of the Western Han period (206 b.c.–a.d. 22): “In the written language from the pre-Christian era right down to our own day, people have continued to use the original short and concise word material; in other words, the authors have continued to write classical Chinese, regardless of the fact that the spoken language had gone its own way.”1 Many Chinese scholars are fundamentally in agreement with Karlgren. The compilers of Gudai Hanyu, a standard college textbook on classical Chinese in contemporary China, make the observation that “Classical Chinese as a literary language is not only very different from modern Chinese but also out of line with the spoken language in all the dynastic periods following the Qin and Han.”2 Some other scholars believe that the disunion started much earlier. According to John DeFrancis, the divorce of the script from speech “most likely goes back to the earliest known stages of Chinese writing,” the Shang period (sixteenth to eleventh centuries b.c.), when oracle inscriptions had to be terse and concise for the practical purpose of saving space on the precious bones and shells on which the inscriptions were carved.3 Lü Shuxiang, however, adopts an approach that is more than just chronological. Instead of focusing on a possible age for the start of the disunion of writing from speech, he suggests that the bifurcation between what he calls yutiwen (writing in speech-based style) and chaoyutiwen (writing in ultraspeech style) could occur contemporaneously. He puts the Lunyu (The Analects) in the category of yutiwen, to which he believes most of the Mengzi ( Mencius) also belongs. The Xunzi and the Zhuangzi,

10

Vernacularization before Shuihu zhuan

11

according to Lü, are probably not as closely aligned with speech of the time and therefore closer to chaoyutiwen. Lü’s terms “yutiwen” and “chaoyutiwen,” as he makes clear in the article, are not synonymous with the terms “baihua” and “wenyan,” only partially overlapping.4 Indeed, one has to wonder at the fact that, despite the prolonged estrangement from living speech, wenyan performed its function with admirable efficacy for over two millennia. With some exceptions in the Yuan period (1279–1368) when China was under Mongol rule, virtually all the legal, administrative, technical, and educational writings throughout China’s entire dynastic history were written in wenyan. The transcendence of the written language over all mutations of the living tongue throughout time and space may have contributed tremendously to the agglomeration, cohesion, dissemination, and endurance of Chinese culture. At the same time, the predominance of wenyan was responsible for some salient features of Chinese literature. Until the advent of vernacular fiction, Chinese literature through all the centuries was written almost exclusively in wenyan, despite the constant infiltration of vernacular elements. As we will see in this chapter, vernacularization in Chinese literature was slow and arduous, as determined, first of all, by some conceptions of language and writing deeply rooted in ancient Chinese culture.

Zi, Wen, and Yu in Ancient Chinese Culture In Chinese classical literature, wenyan’s divorce from speech was not only accepted as a necessity but also acclaimed as a blessing. To be sure, in all cultures alike there is a stylistic difference between literary writing and oral delivery, but in Chinese classical literature the exaltation and sanctification of written characters, or zi, as opposed to yu (speech utterances), went well beyond the boundary of any stylistic considerations. In fact, the privilege of written characters was celebrated precisely for those properties that were largely extralinguistic. This is well reflected in the polysemy of the word “wen,” which plays a key role in traditional Chinese literary theories. Wen, as later superseded by the compound word wenxue, is often taken to be a Chinese equivalent to “literature.” Among other things, however, the word is also often used to refer to the writing symbols. These two seemingly separate lines of signification are actually intertwined and based on the same analogy: Both literary writings and written characters were believed to represent patterned forms of configuration or embellishment in the universe. Even the structure of the character “wen” itself was perceived to be composed of “intersecting strokes, representing a criss-cross pattern.”5 In his Wenxin diaolong, generally considered the most important work of

12

Chapter 1

literary theory in ancient China, Liu Xie (fl. 500) elaborates on the comparability of human wen with the patterns in the natural world so as to argue for a position for literature in the cosmic order: Wen, as a power, is great indeed! It was born together with heaven and earth. How can we say so? The “black one” [i.e., heaven] and the “yellow one” [i.e., earth] interspersed different colors; and the “square one” [i.e., earth] and the “round one” [i.e., heaven] keep their bodies apart. The sun and moon, like twin jade discs, exhibit the pattern affixed to heaven; mountains and rivers, like brilliant damask, display the configuration covering the earth. These are really the wen of the Dao. . . . When we extend our observations to all classes of things, we see both animals and plants have their wen. The dragon and the phoenix present auspicious signs with their painting-like colorful decorations; the tiger and the leopard invigorate their appearances with their brilliant and profuse patterns. The ornate colors [diaose]6 of the sunlight-reflecting clouds outmatch the painter’s wondrous art; the bright blossoms of the plants do not need the brocade-weaver’s marvelous skills. How can these be adornments wrought from without? No, they are so all by themselves. . . . Now if things that are without effect and intelligence appear in such colorful patterns, how can the being that is endowed with a mind be without his wen?7

Literature, therefore, is nothing but a human analogue to the configuration of cosmic phenomena, or a human “pattern” that is compatible to the “patterns” in the natural world. This model, as we will see, was also adopted in the traditional Chinese perception of their writing symbols. The pictographic origins of many Chinese characters seemed to offer themselves as a testimony to an unmediated imitative relationship. A legend goes that Cang Jie, the emperor Huangdi’s historian, invented characters upon watching the prints of birds’ talons on sand: A human pattern came into being with the inspiration from a pattern in the outer world.8 Another tradition on the origin of the characters, probably from a different source, is strikingly similar. In the Yijing (Book of Changes), the invention of the bagua or the Eight Trigrams—often regarded as the prototype of Chinese characters—is also said to have been based on an observation of patterns in the outer world: In ancient times, when Bao Xi was the king of the world, he raised his head to observe the signs in heaven and bent down to examine

Vernacularization before Shuihu zhuan

13

the arrangement on earth. He contemplated the patterns [wen] on birds and animals and the harmony [ yi] on earth. He was inspired both by his own person and by objects afar. Thereupon he started to invent the Eight Trigrams.9

If writing symbols originated from an attempt to embody natural patterns, as assumed, they could as a matter of course be perceived as the most congenial vehicle for literature, which was similarly believed to be a human counterpart to the patterned configuration of the outer world. Indeed, the polysemy of wen as referring both to literature and to writing script seems to reflect the striking affinity between the traditional Chinese perception of literature and that of the script used to write it. Modern linguistics tells us that writing symbols, once adopted by language as surrogates for speech sounds, refer to concepts or ideas only through the mediation of speech. Apart from speech, writing should have no life of its own, as Eric Havelock insists.10 The Chinese writing system, despite its largely pictographical origins and more complicated and elusive relationship to the phonetics of the language, is considered no exception to the rule. Against the “aphonic” view on Chinese characters, William Hannas argues, “Etymologically, we have seen . . . that most characters owe their shapes to the sounds of the language they represented. This is a panlinguistic phenomenon that applies to Chinese no less than to any other developed writing systems.”11 To the ancient Chinese, however, the presumed signifying capacity of Chinese characters bypassing speech sounds did seem to bestow on the script a life of its own. Not regarded as merely companions to the phonetics of the language, the written characters were recruited to serve in a mission that was quite elitist in nature—a mission called wenxue, meaning “literature” as well as the “learning of the embellished patterns.” It was believed that the gap must be carefully maintained between the words in common speech and those represented by characters in literary writings. The term “wenyan,” or “yan zhi wen” as Liu Xie calls it, means the “pattern/configuration/embellishment” of words.12 As the literary language, wenyan was believed to have derived its potency not from ties to speech but from its affinity to the patterns of the outer world. “The wen of words,” as Liu Xie exclaims enthusiastically, “is the heart of the Universe.”13 The conviction that only “patterned words” were to be used in literary writings died hard. As late as the nineteenth century, after the dominance of wenyan was repeatedly challenged by vernacular literature, Ruan Yuan (1764–1849) was still championing wenyan as the sole legitimate literary language by appealing to no less an authority than Confucius:

14

Chapter 1

Confucius himself gave the title “Embellished Words” [Wenyan] to his commentary on the hexagrams Qian and Kun [in the Book of Changes]: this is the ancestor of the literary compositions [wenzhang] of all ages. Those engaged in literary composition, who . . . merely use single [i.e., non-parallel] sentences and write in such wild abandon that, as soon as they begin, they will not stop until they have reached thousands of words, do not realize that what they write is what the ancients called “speech” [ yan], which means “straightforward speech,” or “talk” [ yu], which means “argument,” but not language [ yan] that has embellishments [wen], not what Confucius called wen [embellished words/literature].14

The distinction between “embellished words” and “straightforward speech” and the exclusion of the latter from literary writings had a tremendous impact on the development of Chinese literature. If the writing symbols with their crisscrossed strokes were considered human “patterns,” and if these patterns were found in correspondent relationship with the patterns in the cosmic world independent of—or at least in addition to—their relationship to the phonetics of the language, then such patterns could be “woven” into literary compositions that were also believed to be manifestations of outer-world patterns. This conception of the relationship between script, literature, and the world seemed to allow the written characters to serve a symbolizing function in literature while maintaining only a slim connection to speech. Related to the perception that script was detachable from speech, there was also a deep-rooted distrust of language per se as a vehicle for communication. The most celebrated Daoist aphorism, “The dao that can be spoken of is not the constant Dao, and the name that can be named is not the constant Name,” suggests an insurmountable inadequacy in linguistic signification.15 As Laozi suggests, language, although it can be of temporary use, can never be a reliable means for conveying the ultimate truth. It is in this sense that Laozi observes, paradoxically, that “The one who knows does not speak, and the one who speaks does not know.”16 The ultimate purpose in communication, as ancient Chinese thinkers believed, was not to understand words per se but to get to what the words were supposed to convey, or yi.Yi, however, was never merely the semantics of words. Rather, it referred to the prelinguistic mental process of meaning formation. While the concepts represented by language were necessarily limited in number, there was a boundless possibility for different nuances and subtleties in meaning. Yi, therefore, was to a large extent ineffable, that which words might approximate but would inevitably fall short in their attempt to convey, as Zhuangzi puts it in this famous message:

Vernacularization before Shuihu zhuan

15

What gets written is nothing but words. There is something that words are subordinated to. What the words are subordinated to is meaning [ yi ]. Meaning resides where it belongs. Where meaning belongs is beyond the conveyance of words.17

Confucius may also have been among the first to see the slippage of meaning over the chain from concepts in mind to speech and finally to written words. In the Yijing, he is quoted to have said: “What is written lacks the fullness of what is/was said; what is/was said lacks the fullness of the meaning in the mind.”18 As a result, the sages who wrote the Yijing had to deviate from natural language in order to transit the meanings more adequately: “To give a full expression to their meaning the sages set up the images, and to distinguish fully what is true from what is false they invented the hexagrams.”19 The view held by the early Daoist and Confucian masters on the inadequacy of language as a means of communication had a strong impact on literary theorists of later ages. Lu Ji (261–303), for instance, expresses concern in his Wen fu that the meaning in his writing would not tally with what was in reality and that what was written would not capture what had been intended as meaning.20 Referring to Zhuangzi’s story of the wheelwright who deplores the incapability of language to convey the tricks in his craft,21 Lu Ji regrets that he is likewise at the end of his linguistic resources: “This is what Wheelwright Bian could not put in words, and therefore even brilliant language will not be able to bring out its quintessence.”22 Liu Xie, too, was keenly aware of the constraints of language. As seen by Liu Xie, ideas, however wondrous they may be, will inevitably suffer a diminution once they are put into words. The reason, according to him, lies in the fact that words, unlike ideas that have full swing in the realm of imagination, are bound to and restrained by a system of concepts predetermined by reality. Like Lu Ji, Liu Xie is compelled to declare that the expressive power of language is sharply limited: “The subtle meanings in our thought and the cryptic mental workings inexpressible in words are not to be reached by language, and this is where one has to be sensible enough to halt his pen.”23 This obstinate skepticism toward the competence of language contributed to the dominance of the expressive theory in Chinese poetics. Since words in natural language were thought to be neither reliable nor accurate in transmitting meaning, literary writing, in order to be truly meaningful, had to claim a more direct and immediate connection to preverbal mental phenomena. Traditional Chinese poetics suggests an almost mystic interaction between the human mind and the ordered cosmic world. Only the human mind has in itself the power to manifest the aesthetic patterns latent in all phenomena, and literary composition (wenzhang ) is the outward realization of that manifestation.

16

Chapter 1

“The poem is where one’s intent goes,” says the writer of the “Da xu,” or “Greater Preface” to the Shijing. “In the mind it is one’s intent; coming forth in words it is a poem.”24 That suggests, as Stephen Owen puts it, “a perfectly adequate correlation between a prelinguistic interior state and a poem,”25 where the written words, or the zi, were considered not merely signs of language per se but schematizations of aesthetic patterns, “outlines or impressions, evidences of what lies outside of language.”26 Therefore, a faithful adherence to natural language, the living language as used by people in the society, was considered not only unnecessary but also undesirable in literary composition. Rather, poets and writers were encouraged to array written characters into “patterns” that deviate far from the living expressions of contemporary Chinese. They believed that, as Confucius is alleged to have admonished them, “If words lack patterning, they will not go far.”27

The Role of Classical Chinese The “patterned words” went very far indeed. One testament to the versatility of wenyan was the almost limitless variety of poetic forms with dazzlingly different tonal patterns and rhyming schemes: the siyan shi (tetrasyllabic) as exemplified by some pieces in Shijing, the Sao ti as initiated by Qu Yuan’s (340?278? b.c.) Lisao, the wuyan shi (pentasyllabic) and qiyan shi (heptasyllabic) in either jueju (stopped lines; quatrains) or lüshi (regulated verse) types, and the more musically orientated ci (long and short verse) and qu (lyric song), to name only a few. In prose writing, the generic variety was equally exuberant. As early as the third century, Lu Ji found the genres in prose writing so vigorously multiplied that he saw the necessity to define each of them in his Wen fu, in which he listed altogether nine prose genres, in addition to that of lyric poetry (shi). Merely two or three centuries later, with further developments in prose writing, Liu Xie saw the need to classify the genres more accurately and increased the number of genres to thirty-four in his Wenxin diaolong, most of which were in prose.28 In addition to this remarkable adaptability to different genres, wenyan demonstrated a great capacity for stylistic variation as well. For example, the acute contention between different schools of prose writing in the Tang (618– 907) and Ming (1368–1644) periods clearly testifies to the stylistic elasticity of the literary language.29 On the other hand, insulated from the living orality of contemporary society, wenyan was confined to the largely homogeneous community of the cultural elite. For the most part through China’s dynastic history, it was a written language of the literati, by the literati, and for the literati. Yet what might appear to be a cultural monopolism was a cultural isolationism as well. As wenyan sealed itself off from the protean linguistic changes diachronically as

Vernacularization before Shuihu zhuan

17

well as from the different social dialects synchronically, its referential capacity in relation to the physical world was severely hampered, and the development of narrative literature—which among all literary genres is by far the most closely tied to external reality—was hindered. Happily, however, wenyan did not always succeed in keeping itself out of touch with speech. As no writer of wenyan lived in a social vacuum and was completely immune from the influence of the circumambient orality, wenyan was never monolithic but constantly infiltrated by vernacular elements. While wenyan and baihua can be considered in general terms as two different written languages, they should by no means be polarized against each other as mutually exclusive. Just as no work in the entire history of Chinese literature is in a “pure” vernacular, a “pure” type of wenyan is equally nonexistent. What we have instead is a whole spectrum of different mixtures of wenyan and baihua ingredients. For instance, Bao Zhao’s (414?-466) letter to his sister, “Deng Dalei an yu mei shu,”30 is clearly much closer to the wenyan end of the spectrum than the contemporary work Shishuo xinyu, a collection of entertaining anecdotes and conversations compiled by Bao’s patron Prince Liu Yiqing (403–444), where vernacular elements are often evoked in order to catch some of the piquancy of the dialogues.31 Similarly, not all poets were equally willing to confine their versification to “patterned words” only; some of them were quite eager to absorb elements from folk songs. For instance, Bai Juyi’s (772–846) poems, especially those in imitation of the yuefu, would sound more understandable to the contemporary ears than those by most other Tang poets. Such vernacular elements reflected the writers’ creativity and contributed to the diversity of wenyan literature, but they were after all merely auxiliary and expedient and never constituted any real challenge to the dominance of wenyan as the sole literary language.

Chan (Zen) Buddhist Yulu and Statutes of Yuan Rulers While the Chinese had long been aware of the phonetic aspect of their written symbols (the rhymes in the Shijing may serve as an example), that awareness was greatly enhanced by the stimulation from Buddhism. As the Indian and central Asian Buddhists newly arriving in China used Sanskrit letters to represent Chinese sounds, the Chinese learned to indicate the pronunciation of any given character with two others, the first having the same consonant as the character in question and the second having the same vowel and, in some cases, the same tone. This system, known as fanqie, was initiated by the thirdcentury exegetist Sun Yan and further developed by Shen Yue (441–513). A greater impact on the written presentation of speech sounds was introduced from the early translation of Buddhist sutras. As most of the early translators

18

Chapter 1

of Buddhist texts were foreigners who were reasonably familiar with spoken Chinese of the time but whose command of wenyan was insufficient, the products of their translation tended to be infiltrated by oral expressions, resulting in texts conspicuously more vernacular than their non-Buddhist counterparts of the same time.32 In particular, the translators’ practice of transcribing Buddhist names and terms in Chinese characters could have been a catalyst for the later efforts to register Chinese spoken sounds in written forms, especially in the cases of the suffixation of the r sound and onomatopoeias. What was probably among the earliest uses of the written vernacular was associated with Chan Buddhism. With the conviction that true knowledge was never expressible verbally, early Chan masters had a strong distrust toward writing as a vehicle for their teachings, as their precept bu li wenzi (no establishment of written words) clearly indicates. The mastery of truth, as they believed, was based on intuitive cognizance (wu) rather than linguistic comprehension. To comply with that precept, however, became increasingly difficult for the later Chan Buddhists, as both the charm and the power of written communication were increasingly apparent. The written genre of yulu, which records—or rather recreates—conversations between a Chan master and his disciples, came into being in the Tang (618–907) and became more popular in the Song.33 For the Chan Buddhists, the use of a language medium that was vernacular to some extent was perhaps a much-needed compromise: While they could now write, they also wanted to write differently from the Confucians and Daoists in order to maintain the distinction of Buddhism from the indigenous ideologies. The immediate impact of the Chan writings on the rise of vernacular fiction may have been rather limited, for at least two reasons. Early Chan yulu were predominantly in prose, but after Master Shanzhao (947–1024) invented songgu—a form of rhymed verse in exegesis of gong’an, exemplary sayings and doings of early Chan sages—the role of prose was gradually diminished in the genre. This shift from prose to verse, as we will see, went in a direction diametrically contrary to the general process of vernacularization in literature, which went gradually from verse to prose. Another reason, perhaps a more fundamental one, was the metaphysical nature of the discourse. Chan yulu texts are more often than not abstruse and esoteric, even though they can be outrageously colloquial at times. To illustrate this, here is a short dialogue between Master Dongshan and a disciple, perhaps one of the best-known dialogues in the entire genre of Chan yulu: The disciple asked Dongshan, “How is it to be a Buddha?” Dongshan answered, “Three pounds of hemp.”34

Vernacularization before Shuihu zhuan

19

No wonder Keqin, a later Chan Buddhist, exclaimed in perplexity over this dialogue: “How many people went astray trying to understand this gong’an! You want to give it a bite, but you don’t know where to set your mouth.”35 While the Chan yulu testified to the possibility to recapture spoken words in characters, the abstruseness of the dialogues may have limited the readership of the genre to Chan Buddhists themselves and, later, interested literati. During the Song period (960–1279), yulu became a popular genre of writing for Neo-Confucians. Just as Chan Buddhists had done, disciples in NeoConfucianism would jot down their masters’ sayings, which were later collected and edited for publication.36 Their adoption of the yulu form and colloquial language in disseminating their masters’ teachings reflected an ambivalent attitude toward the authority of the classic texts and an increasingly felt need for a new form, autonomous from the traditional commentary, to express the new school’s understanding of the Confucian canon.37 In comparison with the Chan writings, the language in the Neo-Confucian yulu is more accessible. Yet in assessing its impact on the vernacularization in literature, we should take two facts into account. First, the use of the vernacular in the yulu was quite different from its use in fictional literature, as in yulu the prestige of the high learning in the content would easily offset the presumed crudeness of the language medium. If it was possible to record someone’s speech in writing in Zhu Xi’s (1130–1200) time, still, to record the great master’s words was one thing, and to record a storyteller’s would be quite another, as our later discussion of the oral and performative genres will indicate. Secondly, the publication of a NeoConfucian yulu was usually considerably later than the master’s lifetime, and for that reason the written vernacular in the yulu was not as early as the master’s name would suggest. The best-known exemplar of the genre, Zhuzi yulei, is supposedly based on Zhu Xi’s dialogues with his disciples during the last thirty years of his life, but the book, compiled by Li Jingde (fl.1263), did not appear in print until 1270.38 That was only nine years before the Mongols established their rule over all of China and was probably contemporaneous with some plays by Guan Hanqing, Bai Pu (1226–?), and other early Yuan zaju playwrights. If the early attempt at vernacularization represented by the Chan yulu was in a sense brought about by the cultural invasion of Indian Buddhism, the temporary erosion of wenyan in the government statutes of the Yuan period (1279–1368) was because of a more forceful and less negotiable foreign influence—namely, military conquest. As the Mongol conquerors suspended the civil service examinations, wenyan lost its most important institutional buttress. And this was compounded by the nescience of the greenhorn Yuan rulers with the classical language. Consequently, the language in some bureaucratic documents during the Yuan period features a curious hybrid of simple wenyan,

20

Chapter 1

Mandarin, and Mongolian terms and expressions transliterated into Chinese characters, sometimes in a distorted syntax. This linguistic aberration, as registered in the imperial edicts and other official statutes—some of which were inscribed on tablets—of course could not survive the particular social conditions of the historical moment.39 But what did have a far-reaching impact on the development of vernacular literature was the fact that, for the first time in the history of wenyan, the classical language lost, at least partially, its blessing from the supreme ruler and support from the bureaucracy. This might have been one of the reasons for the prosperity of the oral and performative genres, especially zaju, during the Yuan period. To meet the need of the Mongol rulers to sinicize themselves, works were written to explicate classical texts in a more colloquial language. The best known of such works were Xu Heng’s (1209–1281) Zhishuo Daxue yaolue (Direct Explication of the Essentials of the Great Learning), Daxue zhijie (The Great Learning Directly Explained), and Zhong yong zhijie (The Doctrine of the Mean Directly Explained). A later exemplar, Xiaojing zhijie (The Classic of Filial Piety Directly Explained), was compiled by the sinicized Uigur scholar Guan Yunshi (1286–1324). The temporary weakening of wenyan during the Yuan period brought about a more favorable climate for the development of baihua, and such works in the Yuan bureaucracy were a result of that changed climate. Targeting a relatively small readership and serving mainly practical and topical purposes, however, these writings did not mark the rise of the written vernacular as a new literary language. Like the early translations of Buddhist sutras, these works are good examples to show that vernacularization of Chinese written prose became much more likely when either the writers or the audience were on the periphery of the native culture. Similar examples are Pak t’ongsa ònhae and Nogoltae ònhae, textbooks compiled in the fourteenth century by Chinesespeaking Koreans to teach conversational Chinese to their fellow countrymen, where both the compilers and readers were foreigners.40 In the heartland of Chinese culture itself, where the traditional notion of writing was firmly rooted and the divergence of writing from speech was obstinately taken for granted, the written vernacular could only emerge from the massive and sustained contact and interaction between writing and orality.

Works Associated with Popular Performative Genres For contact and interaction between writing and orality to take place, an interface was needed, and the popular genres of public entertainment that flourished since the Tang times provided such an interface. The diverse forms of public storytelling and popular drama not only enriched the cultural life of the people but also played a decisive role in the vernacularization movement. It is

Vernacularization before Shuihu zhuan

21

therefore both necessary and worthwhile to review briefly some works that arose from popular genres prior to Shuihu zhuan. Obviously this review is more illustrative here than exhaustive, as it is impractical to examine all such works one after another. In our assessment of the vernacularity of a work that came into existence several centuries ago, we must always bear in mind that the vernacular is not an absolute concept but has to be determined by the linguistic context diachronically and synchronically. We should not, obviously, underestimate the vernacularity of a past work only because its language is different from modern Mandarin. The written vernacular, especially vernacular prose, is mutable diachronically in the wake of the changes in the spoken language. Synchronically, the written vernacular covers a spectrum of different language mixes, resulting in different stylistic nuances. It is, therefore, not always equivalent to written colloquialism, although in its early stage the written vernacular must always develop in close contact with living orality, typically appearing in either a notation of or a composition for an oral performance. The situation was further complicated by the dialectal heterogeneity in various parts of the nation. Most of the early vernacular works, regardless of their places of origin, seem to be based on a lingua franca with the northern dialect as the main ingredient—but that does not mean that linguistic dominance met no resistance from other dialects whose influence may have left their marks in vernacular works as well.41 Even the northern dialect itself was by no means monolithic, and the differences among the subdialects were often more than negligible. The formidable complexity of the situation simply means that there was no such thing as a norm against which we can measure the vernacularity of an early vernacular work. Fortunately, our purpose here is not to gauge accurately the vernacularity of any individual work but to map out a general course of vernacularization before the advent of Shuihu zhuan. This goal is achievable, as we can approach the issue in comparative terms, for comparisons can be made between different parts of the same work, between works of an identical subject matter, and between a work’s earlier textual exemplar(s) and its later edition(s). Among the manuscripts discovered in the early twentieth century in a cave near the northwestern city of Dunhuang were the bianwen, or transformation texts, from the Tang period.42 As “written descendants of oral narrative performed by popular entertainers” dealing “with secular or religious subjects,” the bianwen texts appear in a language that approaches the written vernacular.43 To determine the specific ways in which the bianwen texts were tied to orality is a complicated issue, but there is evidence, especially the verse-introductory formula, for some kind of relationship to spoken presentations,44 al-

22

Chapter 1

though the majority of the bianwen texts could be, according to Victor Mair, “several generations removed from the seminal oral performances that led to their birth.”45 Public storytelling of the Tang period may have had an impact on the professional raconteurs of subsequent times, and the prosimetric form as an essential characteristic of bianwen, a form that alternates prose with verse lines, was to be inherited in most of the later performative genres.46 During the Song and Yuan periods, various forms of performative art were flourishing. In association with such popular genres, texts were produced, either compositionally or notationally, which are generally accepted as vernacular works. Particularly interesting are the different works based on the same love story of Zhang Junrui, a young scholar, and Cui Yingying, a maiden from an aristocratic family. The story originated in “Yingying zhuan,” also titled “Huizhen ji,” a short tale written in wenyan by the Tang poet Yuan Zhen (779–831). Later it became a favorite subject for public storytelling and performance. Since the evolution of the Zhang-Cui story involved a number of popular genres in different periods, we may consider the different versions of the story as an epitome of the development of early vernacular literature generated from performing art. In “Cui Yingying Shangdiao dielianhua ci,” the drum-song ( guzici) version of the story attributed to the Song prince Zhao Lingzhi (fl. 1110), arias alternate with short passages in prose. The arias, which were to be sung to a particular tune, are basically in wenyan but interspersed with some colloquial expressions. All the passages in prose, however, are simply taken verbatim from Yuan Zhen’s tale in wenyan,47 each ending with the formulaic expression “Fenglao geban, zaihuo qiansheng,” which in turn introduces a song.48 A later version of the Zhang-Cui story is in zhugongdiao (“medley” or “all keys and modes,” as often called in English), which was a popular form of storytelling in northern China during the twelfth and thirteenth centuries.49 Only three texts that arose from zhugongdiao are extant today, among which both Liu Zhiyuan zhugongdiao (Medley on Liu Zhiyuan) and Tianbao yishi zhugongdiao (Medley on Anecdotes from the Tianbao Reign) are incomplete.50 Fragments of the anonymous Liu Zhiyuan zhugongdiao, based on the story of the Later Han (946–950) emperor Liu Zhiyuan and his wife Li Sanniang, were unearthed in 1907 by a group of Russian explorers. The songs in what remains of Liu Zhiyuan zhugongdiao appear in a language that is basically vernacular, but the prose sections were written in a mixture of simplified wenyan with some rustic colloquial expressions. Ascribed to the late-thirteenth-century writer Wang Bocheng, the other incomplete work in the genre, Tianbao yishi zhugongdiao, tells the story of the Tang emperor Li Longji and his beautiful consort Yang

Vernacularization before Shuihu zhuan

23

Yuhuan. All the extant fragments are comprised only of songs but no prose passages, similar to some of the Yuan-edition zaju plays. The only complete work in the genre, Xixiang ji zhugongdiao (Medley on the Romance of the Western Wing), is attributed to Dong Jieyuan, who lived during the reign of Emperor Zhangzong (1190–1208) of the Jin (1115–1234).51 This text expands the Zhang-Cui story to a formidable length of over fifty thousand characters. It comprises mainly verses to be sung in a large variety of tunes, interspersed with brief passages in prose. A prose passage either retells the content of the songs immediately preceding it or takes over the narrative line temporarily. Unlike the drum-song version, the medley text features little verbal correspondence to Yuan Zhen’s original wenyan tale; but, while all the songs are basically vernacular, the prose passages are still distinctly more wenyan than baihua. As in the cases of the two incomplete zhugongdiao, the extant texts of Xixiang ji zhugongdiao were considerably later than the heyday of the oral genre. There is a chance that even the rather limited use of the written vernacular is not entirely from the original version. In Ming Jiajing ben Dong Jieyuan Xixiang ji, for instance, the sentence-final particle zege coexists with its earlier equivalent zan, which suggests editing of the text in later times—although this is not proven. To see the difference between the stylistic registers in the verse and the prose in Xixiang ji zhugongdiao, let us consider the scene of the parley between Zhang Junrui and the monk Facong, an incident narrated in a set of verses (one hundred characters) followed by a prose passage of a comparable length (eighty-seven characters). Since the register in the songs and that in the prose are largely consistent, respectively, throughout the work, this selection, although chosen randomly, can be considered a fair sample. As the song goes, Zhang Junrui, infatuated with Yingying’s beauty, wants to get into that part of the monastery where the girl and her mother have taken up temporary residence. The monk, trying to protect the ladies from the intruder, blocks Zhang’s way, but only strengthens the latter’s determination to see his “Bodhisattva” (Shuiyue Guanyin). At this point the conversation shifts into prose. Facong reveals to Zhang the true identity of the “Bodhisattva.” As the scholar wonders at her family’s decision to stay at a monastery, the monk explains their choice in terms of their aversion to the mess and disorder of a more public place.52 Table 1 lists some linguistic features in the song and the prose passage respectively, with the number of appearances in parentheses. As we can see in the table, the two groups of linguistic features are mutually exclusive: No feature in either group appears both in the song and in the prose passage. More significantly, while all the markers from the prose are typical wenyan elements,

24

Chapter 1

Table 1 Incidence of linguistic features in the prose sample and verse sample from Xixiang ji Zhugongdiao Verse Prose _______________________________________________________________________________ Verb “to be” shi 4 nai 1 Speech introducing verb “to say” dao 2 yue 3 Pronoun “this” zhe/zheli 3 ci 2 Personal pronoun “you” ni 1 zi 1 Particles le 1 zhi 1 zhe 1 Ye 2 Yi 1 _______________________________________________________________________________

all their counterparts from the song have become part of the core of modern Mandarin, with the sole exception of dao, which was regularly used in MingQing vernacular fiction as the speech-introducing verb but has been replaced by shuo more recently.53 In the Yuan zaju (variety play), a dramatic genre developed on the basis of the Jin yuanben, zhugongdiao, and other earlier forms of popular performance, versification was brought even closer to spoken language, as the rhyming schemes were now regulated in accord with the contemporary pronunciation of the northern dialect, codified in Zhou Deqing’s Zhong yuan yinyun (Sounds and Rhymes of the Central Plain). But more relevant to our discussion here is the development in the prose. In Xixiang ji (Romance of the Western Wing), a zaju attributed to Wang Shifu (fl. 1280), the prose binbai (lines in the script to be spoken on the stage) is strikingly different from that in Xixiang ji zhugongdiao, and the disparity is definitely more than any natural evolution of the spoken language over the few decades’ separation could possibly account for. Table 2 compares the linguistic features in the prose passage from the zhugongdiao, which I discussed above, with those in the corresponding prose passage (ninety-five characters long) from the 1498 Hongzhi edition of the Xixiang ji zaju.54 Again, the number of occurrences of each feature is given in the parentheses. As the table shows, the prose passage in the zaju play is distinctly more vernacular. The two occurrences of wenyan particles in the passage do not necessarily weaken this conclusion, since they occasionally appear even in colloquial expressions in modern Mandarin as well. In addition, there are in the zaju passage three occurrences of the suffixation of the nonsyllabic r sound, as in moyangr (appearance/looks), xiao jiaor (little small feet), and chang qunr (long skirt). As the suffixation of the r sound has been typical of the pronunciation in northern dialect that makes little semantic difference, the textual reg-

Vernacularization before Shuihu zhuan

25

Table 2 Incidence of linguistic features in the prose samples from Xixiang ji zhugongdiao and Xixiang ji zaju zhugongdiao zaju ________________________________________________________________________________ Verb “to be” Speech introducing verb “to say” Pronoun “this” Personal pronoun “you” Particles

nai 1 shi 1 yue 3 Not used* ci 2 zhe 2 zi 1 ni 1 zhi 1 zhi 1 zhe 1 hu 1 ye 2 yi 1 ________________________________________________________________________________ * The word “yun” is not counted. Although a speech introducer, it appears only in stage instructions.

istration of that sound may represent a sedulous effort to align characters with speech. The guzici, zhugongdiao, and zaju texts based on the Zhang-Cui story span over two centuries, with almost equal intervals between them. From the discussion above, two conclusions can be tentatively reached. First, texts that arose from oral and popular genres are not equally vernacular, even though they are generally accepted as vernacular works. In general terms, the three texts are progressively vernacularized, which may suggest a gradual course of maturation of the written vernacular. Second, vernacularization in prose lagged considerably behind that in verse. In the actual performance, the spoken portions of the guzici or the zhugongdiao were necessarily colloquial (recited wenyan would hardly be comprehensible to the audience), but they appear in both texts as prose basically in wenyan, while the songs in the zhugongdiao text display a considerable degree of vernacularity. We will see whether these conclusions can be confirmed by our discussion of texts from a wider range of genres. Since the Xixiang ji zaju, which features extensive vernacular prose, is attributed to a playwright who wrote in the early Yuan period, it might seem that by the end of the thirteenth century the vernacularization of literary prose could have reached its culmination.55 That conclusion, however, is undercut by the fact that no extant texts of the play are from any period earlier than the Ming.56 We cannot be absolutely certain how much of the prose binbai was written by Wang Shifu himself and how much was added by later writers/editors, but at least we can approach that issue by comparing the text of Xixiang ji with some extant Yuan-edition zaju plays. Among the thirty extant Yuan texts collected in Yuankanben zaju sanshi zhong (Thirty Yuan-Edition Zaju Plays), some, such

26

Chapter 1

as Guan Zhang shuang fu Xishu meng, Chu Zhao-wang shuzhe xiachuan, and Yuan bao yuan Zhaoshi guer, are comprised only of arias, with no prose lines at all for binbai. Most of the other Yuan-edition zaju plays contain some but very brief— and sometimes unclear—binbai.57 It was in the Ming editions of the zaju plays that the binbai prose became amplified and expanded. A comparison of the Yuan-edition plays with their later versions in Zang Maoxun’s (?-1621) Yuanqu xuan (Selected Yuan Plays), which appeared in 1616, will immediately show a significant and in some cases even multifold expansion of the prose portions in the Ming texts. Against all the thirty extant Yuan-edition zaju plays, the text of the Xixiang ji, with its developed prose binbai and its extended length, would stand in striking contrast.58 It seems very likely that the vernacular prose in the Xixiang ji zaju as we have it today was not all written by Wang Shifu himself but amplified through a subsequent process of textual evolution. I will return briefly to this issue in chapter 4, where the occurrences of the adverb “jiu” in the binbai prose in the 1498 edition of Xixiang ji zaju will be cited as evidence that much of the vernacular prose in the text is likely to be of later origins than that of the prose in the Yuan-edition zaju plays, in which only “bian,” the predecessor of “jiu,” is used. Zheng Zhenduo also believed that “most of the binbai prose in Xixiang ji must have been added by people in later times,” as he considered the vernacular prose in the Ming editions of the play so inferior to the arias that they are “totally unmatchable with each other.”59 If the binbai prose in the Ming editions still leaves much to be desired, that would suggest that in Wang Shifu’s original version, the binbai, if there was any, could have been even more crude and inadequate. Indeed, this was evidenced by the late Ming pseudonymous writer Panguo Shuoren’s comment as he tried to justify his effort in Xixiang dingben to expand and revise the binbai prose from the Yuan version: “In the original text the lines for speech [shuobai ] were inadequate and paltry [wuzuguan].” Therefore he ventured to “change the wording while retaining the meaning.”60 That, of course, does not mean that binbai was underdeveloped in the staged zaju performances during the Yuan times. What it does suggest is that in Xixiang ji, as in so many other Yuan zaju plays, the lines spoken onstage had probably not been fully registered (where the writing was notational) or regulated (where the writing was compositional) textually until considerably later. The same can also be said of a particular kind of zaju, Shuihu zaju—namely, zaju plays based on episodes from the Shuihu complex, which will be discussed in more detail in the next chapter. Some of the extant Shuihu zaju boast binbai dialogues in extraordinarily fluent vernacular prose, but again, none of the texts is from any Yuan editions.61 The best-known playwright of Shuihu zaju dur-

Vernacularization before Shuihu zhuan

27

ing the Ming period was Zhu Youdun (1379–1439), a grandson of the first Ming emperor Zhu Yuanzhang (r. 1368–1398). The prince was certainly at home with what is outspokenly baihua, and in his Heixuanfeng zhang yi shucai there is even a large array of onomatopoeias, such as gututu and hulala, adding to the vividness of both the verses and spoken lines. But when Zhu Youdun published his Shuihu plays in 1433, the textualization of Shuihu zhuan may have long been under way, even though the fullest form of the narrative had not yet appeared in print.62 If we compare Zhu Youdun’s plays with those Yuan-edition zaju scripts, it seems reasonable to consider the much more developed vernacular prose in Zhu’s plays as the result of the vernacularization movement since the Yuan, of which the gradual textualization of the Shuihu complex must have been a most crucial part. Yet even Zhu Youdun’s plays were subject to further vernacularization. Zhu claimed that each of his plays contained “complete” spoken lines, as the title of each play is followed by the legend that reads “quanbin” (with complete spoken lines). W. L. Idema has pointed out that this legend “is not always accurate,” for there are “many conventional episodes for which no dialogue is provided—apparently the performers were trusted to provide the words themselves.”63 A comparison of the original version of Zhu’s Zhang yi shucai (collected in Shemotashi qucong) with its late Ming manuscript version from the Maiwangguan Library will immediately show that the prose binbai is significantly amplified in the later text. The situation with another dramatic genre, xiwen (or nanxi ), which was popular in the South during the Southern Song and Yuan periods, is strikingly similar to that of zaju; but unlike zaju, of which thirty Yuan exemplars are extant, no extant xiwen script can be safely dated to the Yuan period.64 Many xiwen plays were once included in the early Ming colletanea Yongle dadian, but only three are extant—namely, Xiao Sun tu (The Little Butcher Sun), Zhang Xie zhuangyuan (The Number One Scholar Zhang Xie), and Huanmen zidi cuo lishen (The Son of an Official’s Family Chooses a Wrong Way of Life). Apart from this trio, there is a group of five, generally considered the masterpieces of the genre: Jingchai ji (Romance of the Thorn Hairpin), Baitu ji (Story of the White Rabbit), Baiyue ting (The Moon-Worshipping Pavilion), Sha gou ji (The Slaughter of the Dog), and Pipa ji (Romance of the Lute). With the exception of Pipa ji, which is by general consensus attributed to the late Yuan playwright Gao Ming (fl. 1345),65 authors of all plays in both groups remain unknown.66 The plays in the latter group appear only in late Ming editions, with the exception of Baitu ji, which had a Chenghua (1465–1487) exemplar with the full title Xinbian Liu Zhiyuan huanxiang baitu ji, unearthed in 1967 and collected in Ming Chenghua shuochang cihua congkan. In the 1988 edition of Song-Yuan sida xiwen duben, which includes all the plays in the latter group except Pipa ji,

28

Chapter 1

Table 3 Percentage of binbai in the xiwen texts ___________________________________________________________ Yongle dadian texts

Xiao Sun tu Huanmen zidi cuo lishen Zhang Xie zhuang yuan

34.7381% 39.06% 50.57%

___________________________________________________________ Baiyue ting 58.69% Jingchai ji 61.45% Jiguge texts Baitu ji 61.54% Sha gou ji 65.31% Pipa ji 69.41% __________________________________________________________________

the texts are all based on the Jiguge editions of the late Ming, and the version of Sha gou ji could even have been revised by Feng Menglong (1574–1646).67 A comparison of the plays in Yongle dadian and the ones in the Jiguge editions will immediately show a significantly larger proportion of prose binbai in the later texts. This is corroborated by the statistics that are presented in a study by Liu Xiaopeng.68 Table 3, based on Liu Xiaopeng’s data, shows the percentage of the number of words for binbai in each of the three plays in Yongle dadian and the other five in Jiguge editions. The table displays a conspicuously larger proportion of binbai in the late Ming texts, and the percentage for binbai in the Jiguge version of Pipa ji is approximately twice as high as that in Xiao Sun tu. A similar result will be produced in a comparison between an earlier version and a later one of the same play. For Pipa ji, the Qing scholar Cai Yidian had a manuscript copy, allegedly a replica of a nonextant late Yuan or early Ming edition. Given the fact that the handwritten version carries no act division, a formal feature it shares with the three xiwen plays in Yongle dadian, it may indeed be based on an early text, even though we cannot certify its date. If we read that manuscript version of Pipa ji side-by-side with the late Ming Rongyutang edition of the play, we will again see in the latter numerous additions of binbai lines, especially those in colloquial prose.69 The fact that none of the pre-Ming editions of xiwen plays are extant may be telling in itself. If those editions had contained vernacular prose as extensive as in the late Ming texts, each play would have been a thick volume or even multiple volumes in traditional block print.70 It is highly unlikely that those many volumes would have disappeared altogether. Given the disparity between the early Ming editions of xiwen and their late Ming successors, it is reasonable to assume that those nonextant pre-Ming xiwen plays may have been similar to the Yuan-edition zaju, shorter and with less vernacular prose even than the early Ming texts. Like what happened with the different Xixiang texts and the Yuan zaju, the textual evolution of nanxi involved a gradual amplification

Vernacularization before Shuihu zhuan

29

and maturation of vernacular prose, until the late Ming chuanqi plays reversed the trend and turned to a more affected and ornate style.71

Works Associated with Storytelling Predominantly in Oral Prose If indeed oral prose in the popular performative genres was not reflected textually in any developed form until considerably later, similar “slippage” should be found as well with oral narratives where prose was the predominant form of discourse. Of course, we simply have no firsthand knowledge of the forms of storytelling in premodern China, but the conjecture about prose-dominated oral genres is based on the assumption that the texts arising from or associated with an oral genre more or less reflect the latter’s formal features. For instance, from the zhugongdiao texts we can infer that zhugongdiao may have been an oral genre that alternated spoken passages with songs set in different musical modes. Similarly, jiangshi (telling of histories) could be an oral genre in which a storyteller spoke more than sang, because all texts that could have possibly been associated with jiangshi are dominated by prose instead of verse. In Dong jing menghua lu, compiled by Meng Yuanlao and published in the Southern Song, there are accounts of jiangshi, a form of storytelling based on popular histories, especially of the periods of the Three Kingdoms and Five Dynasties.72 These oral narratives that had flourished in the Song did not appear in any texts that can possibly be dated to the Song period. During the Yuan, however, five popular histories, each containing the term “pinghua” in the title, were published in Jian’an. The five pinghua texts are: Xinkan quanxiang pinghua Wuwang fa Zhou shu, Xinkan quanxiang pinghua Yue Yi tu Qi qi guo chunqiu houji, Xinkan quanxiang Qin bing liu guo pinghua, Xinkan quanxiang pinghua Qian-Han shu xuji, and Zhizhi xinkan quanxiang pinghua Sanguo zhi. Only Pinghua Sanguo zhi is explicitly dated in the Zhizhi period (1321–1323), as we can see from the titles, but since all five pinghua appear in the same format, it is reasonable to assume that they were all published approximately at the same time. There are a few other popular histories, including Xinbian Wudai shi pinghua, Xuanhe yishi, which contains a very sketchy account of the Shuihu story, Lianggong jiu jian, and Da Tang Sanzang qujing shihua. These works are more difficult to date, but scholars tend to agree that most of them were also published in the Yuan.73 As results of compilation based on various sources, these texts feature an irregular alternation of sections in wenyan with ones leaning toward baihua. In Wudai shi pinghua, for instance, the sections recounting the youthful days of Huang Chao, Zhu Wen, Li Keyong, Shi Jingtang, Liu Zhiyuan, and Guo Wei are clearly more vernacular than the rest of the work. But even in these sec-

30

Chapter 1

tions the vernacular is limited, discontinuous, “often jerky and abrupt,” as Wilt L. Idema has put it.74 The sources for most of the wenyan portions have been identified, as they were incorporated, in some cases even copied verbatim, from classical texts, such as Shiji, Zizhi tong jian, and Tong jian gangmu.75 Significantly, however, no textual sources for the vernacular segments have been discovered. It may not be an unreasonable surmise that these vernacular portions could have been based on texts in manuscript form originally related to jiangshi. Otherwise it would be difficult to explain the neat correspondence between the subject matter of the pinghua texts and that in jiangshi as recorded in Menghua lu. But since the use of the vernacular in the pinghua texts is rather limited and awkward, one can hardly consider these vernacular segments faithful written records of the oral prose in jiangshi. At any rate, we can safely say that the oral narratives of jiangshi that had flourished in the Southern Song period did not assume any full textual form even one dynasty later—or at least no such texts have been found. Thanks to scholarship in recent decades, especially Patrick Hanan’s relentless and meticulous efforts, the dating of early vernacular short stories has now become much clearer. Many texts, previously labeled ambiguously as Song huaben, turn out to be not as hoary as previously supposed. The past confusion, I think, largely arose from the failure to differentiate the date of a story from that of the text. The picture can still be enlarged to include two more dates: the date of the setting in the story (most of the stories that could have flourished in the Southern Song are set in the Northern Song period) and the date of publication of the text, which could be different from that of the textual composition. For example, the story “Yang Wen ‘Lanluhu’ zhuan” is set in the Northern Song. With the subject matter fitting perfectly in the oralstory typology of podao as classified in Zuiweng tanlu, it could have had an oral existence in the Southern Song and therefore could well have been a “Song huaben.” But the text of the story as we have it in Qingping shantang huaben may have been composed sometime in the Yuan, and its earliest known version was published around 1550 in Liushi jia xiaoshuo. According to Hanan, among the 149 extant texts of short stories,76 127 are most likely from the Ming period and only 14 are probably from the Yuan, with the other 8 presenting no clear evidence, but possibly either from the late Yuan or the early Ming.77 This simply means that not a single early vernacular story in its textual form can be dated earlier than the Yuan, and that only about one-tenth of these stories were composed textually any time before the Ming. On the other hand, in the Southern Song and early Yuan accounts of popular storytelling during the Song period, the term “xiaoshuo” (literally, “small talk”), distinct from other subgenres of shuohua (literally, “talking”), seems to

Vernacularization before Shuihu zhuan

31

refer to shorter pieces of oral narrative.78 In Naide Weng’s Ducheng jisheng, for instance, it is stated that among the different schools of storytelling, “the tellers of xiaoshuo are the most formidable, for they can tell the sum and substance of a story about an entire dynasty in a short moment.”79 If indeed telling of short stories was popular during the Song period, and if indeed only a small number of texts of vernacular short stories can be dated earlier than the Ming and none before the Yuan, the situation is then very similar to that of the oral jiangshi and the written pinghua: The emergence of vernacular texts trailed notably behind their oral precursors. To be sure, we do not have to subscribe to the old theory that what is called a “huaben” was necessarily once a storyteller’s script. That theory should be abandoned, however, not because it erroneously relates two things that were actually unrelated, but only because it oversimplifies that relationship. Not all of the 149 stories were related to oral storytelling of xiaoshuo; there were, as Patrick Hanan has informed us, different sources for the stories, including classical tales, drama, and folklore.80 Some other stories are actually written in wenyan, and many of the later pieces could have been composed without an oral model because they emerged at a time when the written vernacular prose became securely entrenched and readily available as a literary language. But many of the earlier stories, especially those with the setting in the Northern Song or earlier, do fit neatly into the typology of subject matter for the oral genre of xiaoshuo as described in Ducheng jisheng and Zuiweng tanlu.81 This may suggest that the relationship between the texts of these short stories and the oral xiaoshuo could indeed have been very close—that is, close derivatively but by no means temporally, for it could have taken the oral stories many decades, in some cases even over a century, to get anchored textually in the written vernacular.82 Hu Shi was once disappointed by the immaturity of the written vernacular in the Yuan period as he saw it in the pinghua texts. Later on, he was amazed to read the much more developed and more spontaneous vernacular prose in what he called “Southern-Song huaben stories,” and he began to reproach himself for having wrongly underestimated the Song-Yuan vernacular literature.83 He did not recognize, as we do now, the “dual identity” of many of the early huaben: As stories, they may well have had sources in oral literature during the Southern Song, while as written texts they belong mostly to the post-Yuan times. This duality of the huaben tricked another scholar, Yan Dunyi. Yan argued that the written vernacular in Shuihu zhuan must have matured in the early Yuan on the ground that Shuihu zhuan could have originated with independent huaben stories that should have contained developed vernacular already, as evidenced by “the sophisticated language style” in the extant “Song huaben stories.”84 Like Hu Shi, Yan failed to realize that, in the earliest stage of ver-

32

Chapter 1

nacular literature, the transition from oral prose to text could not have been an instant switch but more likely a slow and strenuous process.

Vernacularization in Prose Patrick Hanan’s dating of the early vernacular stories is indeed consistent with what we have seen in the works associated with the performative genres and with the pinghua narratives. Just as an overwhelming majority of the vernacular stories were composed or textualized in the Ming, plays that contain developed vernacular prose usually appear in Ming editions, even if such plays were associated with dramatic genres flourishing in earlier periods or attributed to playwrights of earlier times. The texts of pinghua from the Yuan, as we have learned, feature an odd mixture of wenyan and baihua. Of course, some of the few stories that were probably written down in the Yuan, such as “Jiantie heshang,” appear in fairly mature vernacular prose. But in general terms, there seemed to be, as the discussion here suggests, a progressive course in the development of vernacular prose from the Southern Song to the Ming. Given the sharply limited vernacular prose in Yuan-edition zaju, the uneven vernacularity in pinghua narratives, the small number of vernacular short stories that can be dated to the Yuan, and the sudden proliferation of texts in vernacular prose that appeared in the Ming, especially after 1500, it seems reasonable to assume that the vernacular prose was written only tentatively and experimentally until it began to be used with a greater amount of ease and confidence perhaps around the middle of the Ming. This would put the development of vernacular prose on a track chronologically parallel to the process of the textualization of Shuihu zhuan. As the first half of the Ming might have witnessed the maturation of written vernacular prose, it was also during that time that the textualization of Shuihu zhuan was completed. This overlapping was by no means a coincidence, for the two events could well be both mutually defining and mutually dependent. The textualization of Shuihu zhuan, at any point in its long process, must have contributed enormously to the momentum of the vernacularization of the written prose in general, and the maturing vernacular prose elsewhere might also lend support to the textual evolution of Shuihu zhuan, until the process culminated in the fanben editions. Generally speaking, in post-bianwen times vernacular verse developed notably ahead of vernacular prose, as we have seen in the previous discussion. Songs in most of the popular performative genres in the Song-Jin-Yuan periods were set in particular musical modes, which often determined the metrical and tonal patterns of the verses. This formal rigidity could have defied oral improvisation. As a result, verses for the performance were most likely written compositionally rather than notationally. But for the spoken lines, there

Vernacularization before Shuihu zhuan

33

was little pressure for such textual regulations. With a little outline, the performers might well be given a free hand to compose onstage impromptu. That is precisely what Zang Maoxun suggests while trying to explain the scarcity of binbai in the Yuan-edition zaju, although he does not explicitly give his endorsement to the hypothesis.85 Wang Jide also suggests the performers’ composition of the spoken lines by arguing that the binbai in the Yuan zaju is “vulgar and obscene” and “not like the tone of the literati writers.”86 In his Xian qing ou ji, Li Yu (1611–1680) makes similar observations: “For the spoken lines in the northern drama, there are no more than a few sentences in an act. Even if you delete all the spoken lines and read only the arias, it still makes smooth reading. Therefore even those few sentences seem dispensable. Judging from this, probably there had been only the arias at the beginning, and the spoken lines might have been added later on.”87 Wang Guowei (1877–1927), however, thought otherwise. He argues that “arias and spoken lines are mutually generating” (qu bai xiangsheng ): “It is most obvious that there would be no way to compose the arias if one did not compose the spoken lines all the way along.” As for the scarcity of prose binbai in Yuan-edition zaju, Wang suggests publishers’ deletion as a hypothetical reason.88 Wu Mei (1884–1939), another prominent theorist on Chinese drama, explains the paucity and seeming dispensability of binbai in Yuan zaju in terms of a formal feature of the genre: The leading male or female role is the singing role, while the minor roles are speaking ones.89 Most contemporary Chinese critics tend to compromise between the conflicting views. Gu Xuejie, for instance, states that “It is an indisputable fact that the binbai in Yuan zaju, after the extended revisions by performers and commentators of the Ming, has lost its original features; yet it would be going too far to say that the original playwrights did not write any binbai at all.”90 Yet there may have been more fundamental reasons for the tardiness in the vernacularization of prose. One of the reasons could be technical: It was difficult to use characters to represent many colloquial expressions, as they had never had corresponding written forms. This was especially so with those “irregulars” in the spoken tongue such as sentence-final modal particles, suffixation of the r sound, onomatopoeias, and so on. Most of such irregulars could be easily trimmed off in versification, which was supposed to be concise and condensed even in the vernacular. Another reason, perhaps a more significant one, was conceptual. Although the vernacular in general was not an esteemed medium for writing, verses in the vernacular were not downright colloquial and therefore might still claim some respect, whereas vernacular prose that represented “unpatterned” and “unrefined” oral speech would be considered by writers too much beneath their dignity to write. This probably explains the fact that the authorship in the Yuan-Ming drama was in general not so much a thorny and perplexing issue as in the cases of the early vernacular prose nar-

34

Chapter 1

ratives.91 Given the lowly status of vernacular prose, it was possible for someone to write in it only when he not only could but also would. We can imagine that even when many could, those who would probably remained few. Because of such technical and conceptual barriers, especially the latter, vernacularization in prose had to be a slow and prolonged process. Despite the constant stimulation from orality, full-fledged vernacular prose could not be taken for granted and was to emerge only as the result of painful negotiations between different cultural forces. The discussion here of the popular genres and the texts resulting therefrom—especially the laggard and sometimes abortive transition from oral discourse to vernacular prose—sheds some light on the formation and evolution of Shuihu zhuan, with the general development of vernacular prose serving as a background to the discussion of the textualization of the narrative. If we disregard what happened in pre-Tang times and consider the Tang bianwen as the earliest effort to bring about a written vernacular, by the early Ming the vernacularization movement had been going on for several centuries. The length of time itself suggests the arduousness of the process. From oral genres constantly emerged texts that were to different degrees vernacular, but the texts, like the oral genres themselves, were constantly buried in oblivion with few exceptions, leaving later vernacular writers as timid and tentative as their predecessors. The experiments with vernacular prose in the Chan yulu could have had a much larger repercussion, but—perhaps because of the cryptic nature of the religious sect itself and its jargon-loaded language—the vernacular prose in yulu remained an isolated and transient phenomenon, failing to gain momentum and grow into a permeating cultural trend. Such was the fate of vernacularization until the more progressive and accumulative process started after the Southern Song period. At the bottom of the repeated relapses in vernacularization was the lack of a work of great influence that was both vernacular enough and “literary” enough to lead the written vernacular out of the doghouse of subculture and establish it securely as a literary language. That work finally came into existence. Shuihu zhuan, with its great length, appeal, and popularity, culminated the long process of vernacularization and entrenched the status of the written vernacular as a new linguistic medium for narrative literature. Qian Xuantong, one of the foremost Chinese scholars of the twentieth century, addresses the pivotal role of Shuihu zhuan in the history of Chinese vernacular literature, and his words deserve to be quoted at length: The embryo of China’s baihua literature can be traced back to the mid-Tang period, when poets like Bai Juyi wrote several baihua poems. During the Song times, some poems in the ci tradition by Liu

Vernacularization before Shuihu zhuan

35

Yong and Xin Qiji as well as some essays and correspondence by Cheng Hao, Cheng Yi, Zhang Zai, Zhu Xi, and Lu Jiuyuan were also written in the baihua. But the resort to the baihua at that time was more expedient than thorough-going. When the writers found archaism [guwen] not very effective in expressing their emotions or manifesting their theories, they compensated for it with the colloquial expressions of the time, which they mingled with the guwen in their writings. The baihua was therefore only auxiliary. . . . Not until the emergence of the Yuan drama, did the great playwrights such as Guan Hanqing, Ma Zhiyuan, Bai Renfu, and Zheng Dehui break the shackles of established genres and bring into being a new type of literature in the contemporary Northern dialect. . . . The Yuan drama can therefore be counted as vernacular literature, only that its lines were meant to be sung and for that reason it does not sound like natural language even though written in baihua. It was with Shuihu zhuan that literature of natural vernacular finally became fullfledged.92

The vernacular works before Shuihu zhuan may represent different stages in the course of vernacularization and in that sense, they may have paved the way for the advent of Shuihu zhuan. But it was with Shuihu zhuan that the long history of sporadic and piecemeal experiments with the written vernacular finally came to an end. After gradual acclimatization to spoken words, now the written vernacular, with full-fledged prose, took root in Chinese culture and was never again to retreat from the foreground of the literary scene. Exerting a strong influence unmatched by that of any vernacular work before it, Shuihu zhuan set the pace for the development of the new literary genre, the vernacular novel, which was to completely change the landscape of China’s narrative literature for the centuries to come.

2 Told or Written That Is the Question The field of early Chinese vernacular fiction has long been haunted by questions concerning the origins of the genre. How was each of the earliest fulllength vernacular novels—Shuihu zhuan, Sanguo yanyi, and Xiyou ji—related to the long oral tradition that preceded it? Did the popular story-cycles only provide the subject matter for the composition of the narrative, or did the oral model exert a shaping influence on the work in print on the level of narrative discourse as well? These questions are so hard to answer simply because we know so little about those popular traditions and about the textual evolution of the narratives themselves. Indeed, no words summarize our quandary better than these by W. L. Idema: “The prevailing uncertainty in these matters means that any view on the origin and role of Chinese colloquial fiction can only be advanced with diffidence.”1 In the case of Shuihu zhuan, a consensus has long been reached that there had been, before the narrative appeared in print, an oral complex of Shuihu stories. About that long tradition that started probably as early as the thirteenth century, the “hard facts” that we know are very few in number. And we know even less about the transition from the oral cycles to the book form of the narrative. With most of the historical information irrevocably lost, an investigation of the formative stage of Shuihu zhuan has to be, in a sense, a comparative study of different hypotheses. Yet if what Idema calls the “prevailing uncertainty” may daunt any attempt to probe the issue, it is also the very justification and motivation for doing so. This chapter revisits the oral Shuihu tradition, attempting to reclaim at least part of the evolutionary course of the Shuihu complex by examining the few ex-

36

Told or Written

37

tant texts or textual remnants that could be representative of different stages of that course. Next the chapter reviews different critical opinions on Shuihu zhuan, largely based on divergent assessments of the novel’s relationship to the oral tradition. My own belief is that the synthesis of the Shuihu materials from various popular genres may have culminated in a long narrative predominantly in oral prose, which did not merely become the source for the subject matter in Shuihu zhuan but bestowed on the novel much of the narrative discourse itself. This hypothesis is put forward here rather tentatively, before it is supported by the textual, philological, and historical studies of the narrative in later chapters.

Early Stage of the Shuihu Tradition If the development of traditional Chinese narrative literature was characterized by a general shift from historicity to fictionality, Shuihu zhuan holds a pivotal position in the course of that transition.2 While Sanguo yanyi, which appeared in print probably slightly earlier, depends heavily on historiography in both the official (zhengshi ) and the popular versions ( yeshi), Shuihu zhuan’s relationship to history is rather tenuous. In Song shi (History of the Song), there are brief references to the historical Song Jiang and his band in the early twelfth century. In his memorial to the throne, Hou Meng, prefect of Dongping, referred to the bandits as “Song Jiang’s thirty-six men, for whom the thousands of official troops were no match.” He therefore suggested to the emperor that an amnesty be offered to Song Jiang in order to enlist his service in the campaign against another rebellion led by Fang La.3 In “Zhang Shuye zhuan,” it is recorded that Song Jiang and his band surrendered after being ambushed and defeated by the forces led by General Zhang Shuye.4 In Li Zhi’s Huang Song shi chao gang yao, Song Jiang’s name is listed among those who led the attack on Fang La, but judging from the accounts in Song shi of the defeat of Fang La, Song Jiang may not have been the main force in the historical campaign as is the case in Shuihu zhuan.5 Around such historical references, the novel agglomerates a large number of fictional tales that probably had first developed orally as separate short pieces before becoming assembled together. Luo Ye’s Zuiweng tanlu contains references to eight different types of oral stories under the rubric of xiaoshuo that were circulating during the Southern Song.6 For each type, a number of story titles are given as examples. Four stories, judging by their titles, may well have belonged to an oral complex centering on the Liangshan bandits, although none of them is now extant. “Shitou Sun Li” (Sun Li the Stony Man) is listed as a gong’an (court case) story. Most likely it was a story about the Liangshan chieftain Sun Li, and its classification as a gong’an story is consistent with Sun Li’s status in Shuihu zhuan as a police captain before joining the rebellion. “Qing-

38

Chapter 2

mian Shou” (Blue-faced Beast) is labeled as a podao (broadsword) tale. It may well have been a story about Yang Zhi, a Liangshan rebel nicknamed Qingmian Shou who was a skilled user of the broadsword. Among the ganbang (staff ) tales there are “Wu Xingzhe” (Wu, the Untonsured Monk) and “Hua Heshang” (The Tattooed Monk), which were almost certainly stories of the two monk-turned-bandit heroes who used staffs as their weapons: Wu Song and Lu Zhishen, who is nicknamed Hua Heshang.7 No other titles listed seem to suggest a similar relationship to the Shuihu complex. That, however, does not necessarily mean that these four were the only Shuihu stories circulating at the time, for the titles are listed only as examples. In fact, not only these four figures but most of the thirty-six who were to become major chieftains of the band in Shuihu zhuan were already popular with storytellers during the thirteenth century. This is evidenced by Gong Shengyu’s (1222–1304) “Song Jiang sanshiliu zan” (Encomiums to Song Jiang and His Thirty-Six), in which Gong celebrates each of the bandit heroes in a short piece of verse, which perhaps was originally attached to a portrait of each rebel either drawn by himself or by Li Song (fl. 1240), another painter one generation earlier.8 The thirty-six names in Gong Shengyu’s list are not completely identical with those of the thirty-six major chieftains, or tiangang xing (Stars of Heavenly Spirits), in Shuihu zhuan. Gongsun Sheng and Lin Chong, two prominent names among the thirty-six in the novel, are absent from Gong Shengyu’s list. On the other hand, among the names on Gong’s list, Chao Gai is not counted as one of the thirty-six in Shuihu zhuan, and Sun Li, despite his important role in the early days of the Shuihu complex, is relegated to the status of a minor chieftain, or a disha xing (Star of Earthly Fiends), in the novel. Although short, Gong’s prefatory note before the verses is remarkably informative about the state of the oral Shuihu complex at the time. It tells us at least two things. First, such figures as Yang Zhi, Sun Li, Wu Song, and Lu Zhishen, who had previously been protagonists in separate and individual stories, became by Gong Shengyu’s time regarded as among the thirty-six fellow members of the same group. This may mean that the process in which the short stories of individual bandits became strung together into a longer narrative may have started as early as the thirteenth century. Second, while he claims that “stories of Song Jiang are heard on the streets,” Gong Shengyu clearly states that he had not seen any such stories in written form apart from the brief notes about Song Jiang in wenyan historiographies.9 Since the artist had such a keen interest in the bandit heroes, any written literature about them, if there were any at the time, would probably not escape his notice. What Gong Shengyu says in the encomiums, therefore, makes it very likely that by the artist’s time no Shuihu texts had yet been produced from the oral storytelling—at least not for the general reading public.

Told or Written

39

One text, a unified but very sketchy version of the nucleus of the Shuihu complex, finally appears in Xuanhe yishi, compiled anonymously probably in the early Yuan period.10 Part of the work is a chronicle of the last years of the Northern Song, with its military defeat by the Jin invaders, the abduction of its last two emperors, and the transfer of the nation’s capital from Kaifeng to the southern city of Lin’an. At other places, such as the section on the Shuihu rebels and that on Emperor Huizong’s dalliance with the courtesan Li Shishi, the tone of the narrator shifts from that of a historiographer to that of a storyteller. The embryonic Shuihu narrative in Xuanhe yishi consists of four episodes: (1) Yang Zhi, stranded in a small town, has to sell his broadsword and accidentally kills a rascal. He is sentenced to exile but is rescued by his friend Sun Li and others. The group of twelve afterward goes to become rebels in Taihangshan (Taihang Mountains). (2) Liang Shicheng, the prefect of Daming, sends an enormous amount of valuables to his father-in-law, the prime minister Cai Jing, as birthday gifts. Chao Gai and his cohorts drug the escorts and seize the gifts. Before they would have been arrested, Song Jiang, a county clerk, sends them a message that allows them to escape and join Yang Zhi and other outlaws. (3) To thank Song Jiang, Chao Gai and the band send him gold, which betrays Song Jiang’s association with the bandits to his adulterous mistress Yan Poxi. Song Jiang kills the woman and, while hiding in a monastery, acquires a “heavenly writ” that lists the names of the thirty-six rebels. Song and thirteen others join the band at Liangshan and, with the death of Chao Gai, bring the number of the chieftains to thirty-three. The number is afterward brought up to thirty-six with the arrivals of Lu Zhishen, Li Heng, and Huyan Chuo. (4) The court offers amnesty to the band, and Song Jiang is made a garrison commander ( jiedushi) after he helps quell the rebellion led by Fang La. In comparison with the immense length of Shuihu zhuan, the Shuihu segment in Xuanhe yishi is very short, occupying merely ten pages in the Shiliju congshu edition.11 It is incomplete even considered as an outline of the narrative, for it leaves out many elements that must have become part of the Shuihu complex by then. Stories about Lu Zhishen and Wu Song, for instance, seem to have already been in circulation during the Southern Song, judging from the titles of stories in Zuiweng tanlu, but neither story found its way into Xuanhe yishi. Even those episodes that are included are very different from their counterparts in Shuihu zhuan. Yet since the segment appears in a work that was intended to be a popular history, as the form of chronicle and the title of the book both indicate,12 the Shuihu complex could by then have started its evolutionary process from individual short tales into jiangshi, a different oral genre specializing in long narratives that “tell about history.” In general, Xuanhe yishi, like those texts of the Yuan under the rubric of pinghua, appears in a language medium that mixes wenyan with ingredients from

40

Chapter 2

baihua. There is no consistent stylistic register throughout the work; while elsewhere the language is often unadulterated wenyan, in the Shuihu segment it leans closer to baihua. As William O. Hennessey informs us, most of the wenyan sections in Xuanhe yishi have word-for-word parallels in seven classical sources that date from the Southern Song. Significantly, none of the vernacular sections, including the Shuihu segment, is found to have any textual parallels.13 On the other hand, it is quite obvious that the Shuihu segment was not written ab initio. In addition to its stylistic distinction from most other portions of the work, the Shuihu section starts and ends abruptly, with little or no connection with what proceeds or follows it. The entire Shuihu section is put under the heading of “Xuanhe sinian” (The Fourth Year of the Xuanhe Reign), even though the only thing in the section that could possibly have happened in that year was Song Jiang’s surrender to the throne, which the narrative mentions only in passing, while everything else is said to be in the second year of Xuanhe.14 This seems to indicate that the Shuihu segment was incorporated wholesale, with no attempt on the part of the compiler to assimilate it properly into the structure of Xuanhe yishi. Since there appear to be no written sources of Shuihu stories apart from some brief historical notes by Gong Shengyu’s time, which was perhaps only slightly earlier,15 we are led to believe that the source for the Shuihu segment in Xuanhe yishi may have been in some way associated to oral storytelling. This seems even more likely if we take into account the fact that the language in the Shuihu segment is notably more colloquial than elsewhere in Xuanhe yishi.16

Shuihu Plays in the Yuan and Early Ming Periods Unfortunately, the segment in Xuanhe yishi, which is the earliest extant text of a Shuihu narrative before Shuihu zhuan, happens to be the only such text; there are simply no extant texts of any Shuihu narratives that can be dated between Xuanhe yishi and the earliest known editions of the novel. Meanwhile, however, the Shuihu complex must have been vigorously active during that time, as evidenced by a large number of Shuihu zaju from the Yuan and the early Ming. Sixteen titles of Shuihu plays are listed in Zhong Sicheng’s (fl. 1321) Lu gui bu, all by early Yuan playwrights: eight by Gao Wenxiu, two by Li Wenwei, one by Yang Xianzhi, three by “Hongzi” Li Er, and two by Kang Jinzhi. Lu gui bu xubian, another bibliographic source attributed to Jia Zhongming (1343–?) on the late Yuan and early Ming drama, lists four titles of Shuihu plays without giving the authors’ names. Out of these twenty plays, five are extant and collected in Shuihu xiqu ji, which also includes another five extant Shuihu plays based on the seventeenth-century manuscripts from the Maiwangguan Library, whose titles do not appear in either Lu gui bu or Lu gui bu xubian. Apart

Told or Written

41

from these, the two fifteenth-century Shuihu plays by Zhu Youdun, Hei Xuanfeng zhang yi shucai and Baozi heshang zi huansu, are also included in Shuihu xiqu ji. For three of these plays—Nao tongtai, Dongping fu, and Jiugong bagua zhen— the dating is problematic. Fu Xihua and Du Yingao designate them as plays from the late Yuan or early Ming.17 Y. W. Ma, noticing in them the emergence of minor chieftains of the rebellion and a closer parallel in their dramatic actions to Shuihu zhuan, believes them to be later than the novel. Actually, even for those plays attributed unequivocally to Yuan playwrights, the extant texts are all from late Ming editions. George A. Hayden has convincingly suggested that all the Yuan plays of the Shuihu zaju may have undergone extensive revisions in the Ming.18 Even for Zhu Youdun’s two plays, of which we know the exact date of the original publications, the situation is complicated by the existence of the late Ming manuscripts from the Maiwangguan Library, which feature a larger amount of dialogue and spoken verse than Zhu’s original versions. In Shuihu xiqu ji, the text of Zhang yi shucai is based on the late Ming manuscript, while that of the other play, Baozi heshang, appears in its original version. As I am discussing here the formation of the Shuihu saga, the focus of my interest is primarily on the dramatic plot of the plays. For that purpose one may still consider the plays attributed to Yuan or early Ming authors as early Shuihu plays, assuming that the basic structure of the dramatic action is not likely to have been radically reshaped by the subsequent revisions. Judging by the structure of these early Shuihu plays, the existence of a popular Shuihu complex during that period is beyond any doubt, as the audience’s background knowledge of the half-historical and half-legendary Liangshan rebellion is simply taken for granted. In most of the plays, usually at the very beginning, there is a monologue by Song Jiang summarizing how he has come to Liangshan and become the leader of the band, and the monologue mentions so many events in the band’s history in such a sketchy way that it would certainly confuse any viewer of the play who was not pre-informed of the tradition. It is striking that the monologue in different plays follows a largely standard line, which may suggest that some portions of the Shuihu complex had become established enough by Yuan times as to have such a strong centripetal impact on the plays.19 Apart from that, however, a Shuihu zaju is an autonomous work in itself, with its plot not necessarily consistent with the ongoing tradition of the Shuihu complex. An early Shuihu zaju typically focuses on an episode involving one or a few bandits, for a very limited period of time, in most cases a holiday. Except for Li Kui fujing,20 each of the extant Shuihu plays from the Yuan and the early Ming features a dramatic action that is not to be found in Shuihu zhuan.21

42

Chapter 2

The disparity between the plots of the early Shuihu plays and that of Shuihu zhuan may seem to suggest that the plays represent an evolutionary stage of the Shuihu complex that was vastly different from its final form. Sun Kaidi has even suggested that the Shuihu zaju and Shuihu zhuan may have belonged to two separate though not mutually insulated traditions, one in the north and the other in the south.22 If that were the case, one would expect the plots of the early Shuihu plays to be basically consistent among themselves. This is, however, not true. On the ranking of the Liangshan chieftains, an issue that carried particular significance throughout the entire history of the Shuihu tradition, the plays are irreconcilably contradictory to one another. The same position of Number 13, for instance, is assigned to three different characters in as many plays: to Li Kui in Shuang xian gong, to Hua Rong in Sanhu xiashan, and to Ruan Jin in Baozi heshang. An early Shuihu play, as mentioned earlier, is typically about a short excursion of one or a few Liangshan figures from the lair—after the assemblage of the major chieftains has reached the plenary number of thirty-six,23 but before the band’s surrender to the throne.24 The dramatic actions, however divergent they might appear to us today, did not really encroach on the core of the Shuihu complex—namely, the aggregation of the heroes under the banner of rebellion. In Yan Dunyi’s words, such dramatic plots are “amplifications of the story” but not the “making of the blood and bones of the story itself.”25 By the time of the early Shuihu plays, one of the elements that did make “the blood and bones” of the Shuihu tradition was Song Jiang’s killing of his mistress Yan Poxi, as the event is uniformly mentioned in Song Jiang’s monologue in several different plays. Yet even though the episode is obviously very dramatic and therefore easily adaptable into a play, none of the early Shuihu zaju, extant or not, was based on it.26 Indeed, the playwrights might have to strike a balance: On the one hand the dramatic plot had to be kept within a tradition familiar to the audience, while on the other hand commercial competition among theatrical troupes would necessitate innovations. They might be consciously shunning the materials from the Shuihu complex where the authority of tradition was too intimidating and prohibitive of any modification, turning instead to a zone where the tradition was less dominant and therefore more promising for further experiments. The bandits’ activities after their final ranking (pai zuoci) and before the imperial amnesty may well have been such a zone, which allowed a compromise between tradition and novelty.27 To some extent, the Shuihu plays can be considered a few more examples of those numerous literary works that— driven by an anxiety of the influence from a powerful tradition—have to struggle for their own territory on the flanks of that tradition. Indeed, many of the early Xiyou zaju and Sanguo zaju are also character-

Told or Written

43

ized by a similar deviation from the established narrative cycles. Judging by the titles listed in Lu gui bu, some of the nonextant Sanguo plays attributed to Yuan authors, such as Sima Zhao fuduo shoushantai (Sima Zhao Retakes the Rostrum for Receiving the Abdication of the Crown), Guan Dawang san zhuo hongyi guai (General Guan Thrice Captures the Demon in Red), and Mang Zhang Fei da’nao xiang fu yuan (The Reckless Zhang Fei Creates an Uproar at the Prime Minister’s Residence), might contain plots that have no parallels in either Sanguo zhi pinghua or Sanguo yanyi.28 The plots of the two extant Sanguo plays collected in Yuanqu xuan—Gejiang douzhi (The Battle of Wits across the River) and Lianhuan ji (The Interlocking Ruses)—are closer to Sanguo zhi pinghua than to Sanguo yanyi. The former play is about Zhou Yu’s unsuccessful strategy to retake Jingzhou by marrying Sun Quan’s younger sister to Liu Bei. In the play, the bride goes to Jingzhou for the wedding, rather than the groom going to the Wu as is the case in the novel. The latter play is based on the episode in which Wang Yun promises the beautiful girl Diao Chan to both the powerful and treacherous minister Dong Zhuo and Dong’s adopted son Lü Bu and, by doing so, has Dong killed by Lü. In the play, Diao Chan turns out to be Lü Bu’s separated wife, a twist it shares with Sanguozhi pinghua but not Sanguo yanyi. As for the Xiyou plays, Lu gui bu lists the title Tang Sanzang xitian qujing (Tripitaka of the Tang Travels to the West for the Sutras), a nonextant zaju attributed to the Yuan playwright Wu Changling. The Tianyige edition of Lu gui bu gives the subtitle of the play, Lao Huihui donglou jiaofo (The Old Hui on the Eastern Tower Invokes the Buddha),29 which suggests that at least part of the plot of the play was quite alien to the novel Xiyou ji. The same can be said of an extant play entitled Xiyou ji, which may or may not be Yang Jingxian’s Xiyou ji listed in Lu gui bu xubian.30 While the title of the play is identical to that of the novel, Sun Wukong in that play is actually more like the ape spirit in the story “Chen Xunjian meilin shiqi ji” (in Qingping shantang huaben) than the heroic monkey in the full-length fiction.31 We therefore need not have too many doubts about the continuity of the Shuihu complex only because of the disparity between the early Shuihu plays on the one hand and the two narratives, Xuanhe yishi and Shuihu zhuan, on the other. This is, of course, by no means to downplay the pertinence of the Shuihu zaju to the study here. Even those dramatic plots that are apparently different from the plot in Shuihu zhuan could have a latent kinship with the novel. In several early extant Shuihu plays, part of the plot follows a stereotyped pattern: An honest man is framed by a treacherous wife/concubine and her lover, usually a yanei (man with powerful connections), when one or a number of Liangshan bandit heroes, who are the husband’s friends or sworn brothers, come to rescue the wronged man and mete out justice to the adulterous couple. Plays

44

Chapter 2

whose plots can at least partly fit this pattern include Hei Xuanfeng shuang xiangong, Yan Qing puyu, Huan laomo, and Zheng bao’en. In the last play, however, it is the wife who is framed by an adulterous concubine and who is subsequently rescued by three Liangshan chieftains. This pattern bears a strong resemblance to two long narrative stretches in Shuihu zhuan, in each of which the hero, Wu Song and Shi Xiu respectively, rebuffs the sexual advances of a sister-in-law and subsequently kills the adulterous woman and her lover in order to avenge a wronged/murdered brother (chapters 24–26, 44–45). Another Yuan zaju, although not a Shuihu play, features a plot that parallels the two stretches in Shuihu zhuan even more closely, especially the one about Wu Song. In that play, entitled Gengzhi Zhang Qian ti sha qi (The Righteous Zhang Qian Kills a Sister-in-law on His Brother’s Behalf ),32 the protagonist Zhang Qian relentlessly resists seduction by his sworn brother’s wife in order to honor his fraternal loyalty. In the end, he has to kill the woman to stop her attempted murder of her husband, and he is consequently thrown into jail. It is, however, hard to say whether the two stretches in Shuihu zhuan were already part of the Shuihu complex during the Yuan and served as models for the zaju, or whether they represent the narrative’s adaptations of those Yuan zaju plays. Whichever was the case, it seems fair to say that the Wu Song and Shi Xiu stories might have once belonged to the same repertoire in popular orality as those Yuan plays featuring similar plots.

The Hypothetical Shuihu cihua Despite the contributions the Shuihu zaju may have made to the Shuihu complex, their plots could have mostly belonged to the periphery rather than the nucleus of the tradition. As the actions in most of the plays are paralleled neither in Xuanhe yishi nor in Shuihu zhuan, we can hardly consider the Shuihu zaju the only link between the two narratives of the Shuihu saga. The major part of the tradition must have been carried on in other genres of popular orality, of which, unfortunately, no textual relics of any extended length are now available. Zheng Zhenduo asserts that in the Yuan period, side-by-side with the Shuihu zaju, “there must have been a Shuihu zhuan,” by which he means a possible earlier version of the narrative not necessarily sharing that title.33 Sun Kaidi, who puts forward the hypothesis that the Yuan antecedent of Shuihu zhuan might be in the form of a Shuihu zhuan cihua, attempts to explain some of the contradictions in Shuihu zhuan in terms of its duplicate incorporation of similar incidents from two different versions of the cihua, one from the south and the other from the north.34 Our knowledge of cihua as a form of oral narrative was deficient until very recently. One extant text that contains the word “cihua” in its title is Da Tang

Told or Written

45

Qin Wang cihua, a narrative about the founding of the Tang dynasty compiled in the late Ming by Zhu Shenglin.35 We have reason to believe that by the late Ming that form of storytelling might have long passed its heyday, for the word “cihua” had by then lost its stringency as a generic label. Jin ping mei cihua, which also contains the word in its title, does not have the formal features that would distinguish it from other vernacular narratives of the time, even though cihua might well have been one of the sources for the work. Again, in the story “Jiang Xingge chonghui zhenzhu shan” ( Jiang Xingge Reencounters the Pearl Shirt), collected in Gujin xiaoshuo, the narrator claims to be telling a cihua.36 But the story, according to Patrick Hanan, may have been written by Feng Menglong himself, and indeed it bears little formal distinction from Feng’s other stories.37 If by the late Ming the conception of cihua had become so murky, it is reasonable to surmise that the genre must have belonged to a period considerably earlier, possibly the Yuan and the early Ming.38 The sixteen texts of shuochang cihua printed in the Chenghua period (1465–1487) and unearthed in 1967,39 which appeared in 1973 in the volume Ming Chenghua shuochang cihua congkan, have substantially enhanced our knowledge of cihua as a form of storytelling.40 The woodblock editions of the shuochang cihua texts date from the 1470s, and some of them may be reprints of Yuan editions over one hundred years earlier.41 The dating of the prints and the possibility of even earlier textual prototypes agree with our conjecture on cihua’s prime time. Except for “Bao Longtu duan baihu jing zhuan” (Lord Bao Judges the Case of the White Tiger Demon), which consists entirely of verse, all the cihua texts are in prosimetric form, alternating between prose and verse. The absence of prose in that particular cihua, however, does not necessarily suggest that it was performed in a different manner. Most likely, as in so many other texts in early vernacular literature, vernacular prose was not registered or regulated textually but left for the performers to improvise. Despite the word “ci” in the name of the genre, which often suggests a diversity of metric patterns, the verses in these cihua appear overwhelmingly in heptasyllabic lines, which occasionally vary into decasyllabic ones. Unlike zhugongdiao, these shuochang cihua texts do not indicate the tunes for the songs; in fact, no names of any tunes are given anywhere.42 Additionally, in contrast to bianwen, there are no verse-introducing phrases. The prose passages are rather parsimonious, in some cases presenting merely one or two exchanges in the characters’ dialogue. The story is thus narrated mainly by the verses, which are heavily formulary and therefore could be orally improvisable.43 The dearth of prose in the shuochang cihua texts contrasts strikingly with what we see in Da Tang Qin Wang cihua, where prose, although interspersed with verses, is the more important vehicle of narration.

46

Chapter 2

To these shuochang cihua, a stanza of verse in Shuihu zhuan seems quite akin in nature. In chapter 48 of the Rongyutang edition, when Song Jiang and his forces are launching a second assault on Zhujia Zhuang (Zhu Family Village), the narrative shifts into these verse lines: Lone Dragon Cliff before Lone Dragon Mount, Atop Lone Dragon Mount is Zhu Family Village. Surrounding the mount is a flowing stream, Circling the village are lines of trailing willows. Within the walls bristle swords and halberds, Before the gate spears and lances are in array. For a foe, people here are all valiant soldiers, In a battle, every man is in his prime vigor. Zhu Long, on the battlefield, is hard to rival, Zhu Hu, in a fight, nobody can equal. And Zhu Biao is an even greater warrior, Bellowing in anger like Xiang Yu, the Hegemon. Lord Zhu is a man full of wily strategies, And his treasures fill a thousand chests. Before the gate stand two white banners in a pair, Strikingly writ on them are these two lines: “Level up the Marsh and seize Chao Gai, Trample flat Liangshan and capture Song Jiang.” (3: 1587–1588)44

In the Chinese text these are all heptasyllabic lines, which are the standard metric form for shuochang cihua. The phrases describing the military array and the martial prowess of the Zhu brothers are clearly formulaic, which can be easily applied to any number of other battle scenes. More important, the stanza forwards the action of the story, rather than merely commenting on the characters or events in the narrative as the verses do elsewhere in Shuihu zhuan. Indeed, these lines may well be a residue of an earlier cihua, as Sun Kaidi has suggested.45 Sun Kaidi, however, obviously considered the textual inheritance to be on a much larger scale. He suggests in the same article that the long stretch in Shuihu zhuan on Song Jiang’s exile to Jiangzhou (chapters 36 to 41), the episodes of Dame Wang’s instigation of Ximen Qing’s lust for Pan Jinlian (chapter 25), and that of Lu Zhishen’s encounter with the evil Daoist priest Qiu Xiaoyi may all have been taken over from the cihua.46 Additionally, Sun Kaidi cites from Shuihu zhuan words and phrases from the southern dialects as evidence of his hypothetical southern Shuihu zhuan cihua.47 But since most of those southern

Told or Written

47

expressions do not occur in sections that Sun identifies as derived from the cihua, Richard Irwin notes the dilemma: Either the scope of the cihua needs to “be enlarged to include them,” or they suggest instead “the locale in which the novel itself was compiled.”48 Indeed, Sun Kaidi has largely disregarded the formal differences between the verse-dominated cihua and the prose narrative of the novel and proposed a relationship of textual derivation so immediate and so extensive that the text of Shuihu zhuan and his hypothesized Shuihu zhuan cihua almost become identical. Irwin must have noted this as a possible problem with Sun Kaidi’s hypothesis, for he tried to straighten out the contradiction with a hypothesis of his own: “As a matter of fact, such a short time elapsed between the writing of the missing cihua and the drafting of the novel that they may have been done by the same group of men.”49 Our knowledge of another formal feature of the shuochang cihua also contradicts Sun Kaidi’s theory of one single southern cihua as Shuihu zhuan’s antecedent, even though he acknowledges possible supplements from another cihua from the north. The cihua, as evidenced by the chenghua texts, was a form of rather limited length,50 which tells one coherent and integrated story.51 Given the episodicity and exuberance of narrative actions in Shuihu zhuan, one Shuihu zhuan cihua as the bedrock for the voluminous and panoramic narrative is hardly conceivable. Nevertheless, the cihua hypothesis remains useful for us. Among the sixteen shuochang cihua, twelve belong respectively to two clusters: four on the life of Guan Suo, allegedly son of Guan Yu, a prominent general of the ThreeKingdoms period; and eight about Judge Bao, who is called Bao Longtu or Bao Shizhi in these cihua, a legendary upright official of the eleventh century. The cluster of Judge Bao cihua particularly deserves our attention. Each of the Judge Bao cihua tells a story of the incorruptible official, a story that is self-contained and noncontingent on any other cihua in the same cluster. Except for the plot in “Bao Shizhi chushen zhuan” (The Family Origins of Lord Bao), which tells of Judge Bao’s entrance into officialdom and for that reason has to precede all other Judge Bao stories, there is no established temporal sequence for all the stories in the cihua cluster. Yet clearly all the Judge Bao cihua belong to the same story complex, as there are frequent cross-references in one cihua to the events narrated in another. In “Bao Longtu Chenzhou tiaomi ji” (Lord Bao Sells Rice in Chenzhou), Judge Bao, immediately before he leaves for Chenzhou on a famine-relief mission, fines Empress Cao one hundred ounces of gold for illegally lending her chariot to another imperial consort of a lower rank.52 That incident, along with Lord Bao’s executions of four of the emperor’s corrupt kinsmen during his mission in Chenzhou, is referred to in “Bao Longtu duan Cao Guojiu gong’an zhuan” (Lord Bao Judges the Case of Empress Cao’s

48

Chapter 2

Brothers).53 And the executions of the unscrupulous Cao brothers, along with Judge Bao’s many other feats, are in turn enumerated in the introductory lines of “Renzong renmu zhuan” (Emperor Renzong Acknowledges His Mother), just before Judge Bao meets the banished empress dowager in a broken kiln on his way back from Chenzhou to the capital.54 As we can see, no unified temporal order can possibly be established for all the events in these different stories—or rather, such a temporal order may have been considered totally irrelevant. That implies that these Judge Bao cihua, while belonging to the same story complex, remained as separate and independent short pieces, not ready to be synthesized into a long Judge Bao narrative. The implication of such a cihua cluster is relevant to our discussion of the role that cihua possibly played in the evolution of the Shuihu narrative. If there had been any Shuihu cihua—which is likely, as the textual vestiges in Shuihu zhuan suggest—it must have existed in the form of multiple and relatively short pieces, each telling a self-contained Shuihu story. Since the stanza of verses that could be a remnant from a cihua occurs amid the narrative of Song Jiang’s campaign against Zhujia Zhuang, one might hypothesize a cihua about Liangshan bandits’ three assaults on the village. Given the immense base of the Shuihu complex, however, it might not be the only Shuihu cihua; rather, a cluster of such cihua is highly possible, although textual traces from other Shuihu cihua have been lost. For instance, there might be—to speculate a little further—a cihua on Lin Chong, another one on Lu Zhishen, and still another one on Wu Song. Among these cihua there could be a strong affinity, yet—as in the case of the cluster of Judge Bao cihua—a mechanism was lacking that could coordinate between different cihua and then integrate them into a longer narrative of epic dimensions. If indeed such a cluster of Shuihu cihua existed, the stories had to belong to a relatively early stage in the evolution of the Shuihu cycle, most likely in the Yuan, before they were later assimilated into a much longer Shuihu narrative where prose instead of verse was the primary form of storytelling. This conjecture tallies with the rough dating of the cihua’s heyday and corroborates the delineation in chapter 1 of the general process in which the vernacularization of written prose lagged considerably behind that of verse. In addition, it helps explain the narrative structure in some portions of Shuihu zhuan, especially in the first seventy chapters, where the narrative looks rather like a collage of the personal sagas of individual bandit heroes. At this point, a couple of clarifications are in order. The discussion of a possible cluster of Shuihu cihua does not mean that the cihua was the only form of storytelling that carried the tradition. The seemingly focused attention on the cihua is only because it was, apart from the Shuihu zaju, the only genre in popular orality with which some textual relations can be traced in Shuihu zhuan.

Told or Written

49

Indeed, even those textual vestiges are rather tenuous. In any case, the evolution of the Shuihu complex should not be taken as a unilinear process. There might very well be other genres of oral performance existing contemporaneously with the Shuihu cihua and Shuihu zaju, all contributing to the final form of the narrative in Shuihu zhuan. Furthermore, although I suggested that the Shuihu stories told in the cihua had to be prosified later in a different form of storytelling, oral narratives that were primarily in prose did not necessarily belong only to the later stages of the tradition. Some of them might have existed early. The Southern Song stories of Yang Zhi, Wu Song, and Sun Li in the genre of xiaoshuo, as one may recall, might well be prose narratives, as the word “xiaoshuo” (small talk) would suggest. The Shuihu segment in Xuanhe yishi, possibly based on storytellers’ notes, points to some Shuihu narratives of more extended length in the late thirteenth century. Since different forms of shuoshu where prose is the primary medium—especially those devoted to historical themes such as jiangshi and pinghua55—have continued to prosper since the Song times down to the present day, it would be highly improbable that no such genres played any role in the Shuihu complex when Shuihu zaju and Shuihu cihua were flourishing.56 We may well believe that the different oral forms in the Shuihu story-cycles conditioned and supplemented each other as well as contended with each other for a niche, all enriching and enlarging the Shuihu complex as a result. In this conglomeration, zaju and cihua, as determined simply by their own formal features, had to remain in pieces of limited length, each independent and self-contained. For them there was never such a need—nor did they ever have the capacity—to absorb materials from different sources and synthesize them into a single work of a mammoth size. That task had to be fulfilled by storytelling in prose, which enjoyed the elasticity to incorporate subplots into a basically unified temporal-spatial scale and prosify materials from theatrical and prosimetric sources to fit in its own narrative discourse.57 The development of that long prose narrative itself had to be a gradual process. Each appropriation of new material would inevitably necessitate an adjustment of the preexisting narrative structure. And the augmentation of the narrative was not simply an indiscriminate quantitative expansion, as materials that would potentially jeopardize the integrity of the narrative had to be barred out. Other factors for the selection and remaking of the narrative materials, as will be considered later, could include the dynamic interaction between the raconteurs and the audiences and the particular interests of those men of letters who helped transmit the story-cycles. Eventually the prose-dominated oral narrative reached its mature form, ready to be edited and compiled for its debut in print.

50

Chapter 2

Premodern Chinese Views on Shuihu zhuan and Early Vernacular Fiction The conviction that the rise of Chinese vernacular fiction was historically related to the tradition of oral storytelling started early among Chinese scholars, although many aspects of that relationship have remained murky up to the present day. In his preface (dated 1589) to Zhong yi Shuihu zhuan, the pseudonymous writer Tiandu Waichen explicitly associates the emergence of the new narrative genre with the flourishing tradition of storytelling during the Song period:58 The ascendance of xiaoshuo [fiction] started during the reign of the Renzong Emperor of the Song Dynasty. That was an age when people enjoyed a peaceful and affluent life, with the security of the national boundaries unchallenged. After taking care of state affairs, the emperor had the leisure to have the officials in charge of court music [ jiaofang yuebu] collect and compile popular stories, set them to music, and have them performed alternately with court dramas. Since then, it has flourished both at the court and among the commoners.59

Tiandu Waichen suggests the existence of a certain kind of texts to be used for the performance. We do not know whether they were detailed scripts or sketchy promptbooks, but since they were first “collected” and then “compiled,” they could be both notational and compositional. In any case these texts were intimately tied to orality, although since they were “set to music” they were probably some species of chantefables, like guzici, taozhen, or changzhuan, and did not contain extended stretches of vernacular prose. Such observations by a scholar not too remote from the formative period of vernacular fiction seem to lend considerable weight to the theory on the relationship of the incipience of vernacular fiction to oral storytelling. Since they appear in a preface to an edition of Shuihu zhuan, obviously the writer must have believed that Shuihu zhuan exemplified that relationship. Hu Yinglin (1551–1602), who also lived early enough to witness some of the earliest known editions of Shuihu zhuan, made similar remarks that seem to confirm the origins of vernacular fiction, especially those of Shuihu zhuan, in popular orality:60 “Nowadays the talks of the streets [ jietan xiangyu] are in circulation. What are called yanyi are usually inferior to chuanqi and zaju, but Shuihu zhuan, compiled by someone with the name of Shi, a native of Wulin [present-day Hangzhou] in the Yuan times, is extraordinarily popular.”61 Yet such remarks, on the other hand, were largely responsible for an elitist contempt for ver-

Told or Written

51

nacular fiction in general and Shuihu zhuan in particular, a critical attitude prevailing among conservative literati scholars of the Ming-Qing times. The new narrative genre was considered a literary form congenitally deficient, being inextricably involved with the lower strata of the culture and contaminated by the vulgarity of public entertaining. While poetry and classical prose were acclaimed as imparting Confucian values and convictions and therefore indispensably beneficial in one’s education and cultivation—paving the way in some cases ultimately to officialdom—vernacular fiction was usually dismissed as something morally uncouth, aesthetically crude, and therefore a mere pastime for vulgar tastes. “It was most painful to have nothing to do in the spare time, especially when one was satiated with wine and tea,” Lu Xun wrote sarcastically about the prevailing contempt and condescension toward vernacular fiction during the Ming and Qing periods. “And this was made worse by the fact that there were no dancing halls at that time. So people needed something to while the time away.”62 To many literati, vernacular fiction was an improper form of literature, deviating from the rules long established in the literary tradition in classical Chinese. As a result, early vernacular fiction was for a long time kept at the bottom of the hierarchy of letters. For Shuihu zhuan, the disdain for the new literary form was compounded by a moral condemnation. The Ming scholar Tian Rucheng (1503–?), for instance, claims that Luo Guanzhong, the purported compiler of Shuihu zhuan, had his offspring of three generations turned dumb as a punishment for “all the treachery and deception” in the narrative.63 In an imperial edict by the Qianlong Emperor (r. 1735–1796) of the Qing, Shuihu zhuan, along with Xixiang ji, was banned both because of its “vulgar and indecent language” ( bili zhi ci) and because of its possible “pernicious” influence on readers.64 At odds with the general disparagement of vernacular fiction, a group of scholars in the Ming and Qing periods stepped forward as champions of the new genre. Prominent among them were Li Zhi (1527–1602), Jin Shengtan (1608–1661), Ye Zhou (?–1625), Feng Menglong (1574–1645), Mao Zonggang (1632–1709+), and Zhang Zhupo (1670–1698). Their common ground for arguing for the literary excellence of vernacular fiction was a readiness to welcome a pivotal shift in the course of Chinese literature, a shift away from the unchallenged dominance of expressive genres toward a generic diversity with a considerable portion of the nation’s literature becoming more mimetic oriented.65 After the long estrangement of narrative literature from the terrestrial daily life of the common people, these critics hailed the advent of vernacular fiction, in which narrative was finally brought into much closer contact with external reality.

52

Chapter 2

Li Zhi’s philosophy, which laid the foundation for a new critical approach to narrative literature, argues in favor of truth from direct life experience as against abstract Confucian principles. According to Li Zhi, what governs a person’s relations with others is the most fundamental activities of material life. “Eating and clothing are all human ethics are about. Apart from eating and clothing, there can be no talk of ethics.”66 Naturally, this philosophical stance led to a new critical view that considered narrative literature as a representation of material reality, including all the commonplaces of the day. This new mimetic view found a most vigorous expression in the criticism on Shuihu zhuan by the commentary ( pingdian) critics.67 In a preface to the Rongyutang edition of Shuihu zhuan (1610), for instance, the writer argues that the narrative, which he attributes to Shi Nai’an and Luo Guanzhong, was modeled upon a larger Shuihu zhuan, which was nothing other than life itself:68 There was a Shuihu zhuan in life before Shi Nai’an and Luo Guanzhong made it appear in ink. . . . There were adulteresses in life, then they were instanced by Yang Xiong’s wife and Wu Song’s sisterin-law; there were procuresses in life, then they were instanced by the Old Woman Wang; and there were liaisons between mistresses and their man-servants in life, then they were instanced by Lu Junyi’s wife and Li Gu. . . . If there were no such things in real life in the first place, how could the writer have accomplished all this even if he shut himself in his study for years and worked his heart out?69

Naturally, for the critics who were mindful of literary imitation of real-life models, verisimilitude became a major principle for critical assessment. When Jin Shengtan in his commentaries on Shuihu zhuan comes to a passage that he considers a successful description of the action or character portrayal, he often bursts in exclamation: “Just like it! Very much like it!” ( Xiang! Ji xiang!) The mission of a narrative as seen by Jin Shengtan is not to record a historical or allegedly historical event, but to create a varisemblance to real life. If narrative is supposed to present a vivid and lifelike world, then, among other things, its characters should speak as people do in real life. Jin Shengtan recognized the role of individualized speech of the characters in Shuihu zhuan. To him, one of the greatest merits of the novel was its use of colloquial language, getting rid of those particles from classical Chinese such as zhi, hu, zhe, and ye. “Each individual character is made to speak in his own individual way.”70 In a similar manner, the commentator in the Rongyutang edition observes that after reading Shuihu zhuan one sees the images of various characters and hears

Told or Written

53

their various voices, while “forgetting there is any mediation of language and writing.”71 This representational view of narrative literature with its emphasis on lifelike character portrayal led these commentators to a hospitable acceptance of Shuihu zhuan and its vernacular prose. The Rongyutang commentator made a good point in noting the spontaneity of the language of Shuihu zhuan. As writing approached the living voice, its mediation naturally became less noticeable than that in wenyan, which had long been divorced from the living tongue. However, the commentators usually did not associate this merit of Shuihu zhuan with the narrative’s origins in popular orality. Instead, they tended to attribute every achievement of the vernacular narrative solely to the creative mind of a writer. Jin Shengtan, for instance, considered all the effects of the narrative’s colloquial language entirely derived from the pen at the command of a literary genius: If one has in his bosom unusual talents, he must have an unusually effective pen; and if he has an unusually effective pen, he must have unusual vigor. Without unusual talents, there is no way to conceive the plot; without an unusually effective pen, there is no way to give full play to the talents; and without unusual vigor, there is no way to support the pen.72

Shuihu zhuan, according to Jin Shengtan, was a piece of writing based on a topic of the writer’s own choice, like an eight-legged essay ( bagu wen).73 He even envisaged the scene of the composition: The writer, whom he chose to believe to be Shi Nai’an, “spread out the paper and picked up a brush, selected a topic [timu], and then wrote out his fine thoughts and polished phrases [ jinxin xiukou].”74 Like writing an examination essay of the eight-legged style, the selection of a topic was said to be most important: “As long as the topic is a good one, the book will be well done.”75 The various events and divergent voices of the characters in the book were all said to center around the chosen topic. They “are used as methods to begin [qi ], continue [cheng], change direction [zhuan], and sum up [he] the composition,” in order to constitute a structural pattern that was well conceived in the first place.76 And, as “each chapter has its own principle of organization [zhang fa], so does each sentence [ jufa] and each word [zifa].”77 This eulogy suggests that everything in the narrative, including the use of colloquial speech, is governed by an ingeniously conceived writing plan. Little wonder that Jin Shengtan calls Shuihu zhuan the “fifth book of genius” and ranks it on a par with Sima Qian’s Shiji.78 Thus there were among Chinese scholars of the Ming and Qing periods

54

Chapter 2

two conflicting critical attitudes toward Shuihu zhuan. While many literati disdained it as something aesthetically inferior and morally pernicious, especially because of its association with oral storytelling, the commentators represented by Jin Shengtan endeavored to establish it as a respectable literary work by severing its ties to popular orality and explaining everything in terms of the literary imagination of a writing genius. To put Jin Shengtan’s hermeneutic strategy in perspective, the following passage by Fredric Jameson, although on a different cultural tradition, may be helpful: The great traditional systems of hermeneutic . . . sprang from cultural need and from the desperate attempt of the society in question to assimilate monuments of other times and places, whose original impulses were quite foreign to them, and which required a kind of rewriting—through elaborate commentary, and by means of the theory of figures—to take their place in the new scheme of things. Thus Homer was allegorized, and both pagan texts and the Old Testament itself refashioned to bring them into consonance with the New.79

As a champion for vernacular fiction, Jin Shengtan knew perfectly well that the only way to surmount the orthodox literati’s disdain and to elevate the status of the new genre was to distance it from its low and humble origins. The oral-derived Shuihu zhuan would not be accepted as such by the established literary world, to which its “original impulses were quite foreign.” With his “elaborate commentary,” Jin Shengtan wanted to introduce Shuihu zhuan into the pantheon of great literature, but not before he tried to strip the novel of all the traces left by its oral antecedents.

Modern Western Disparagement of Shuihu zhuan The “Literary Reform” movement launched by Hu Shi (1891–1962) and other Chinese intellectuals in the early twentieth century brought Ming-Qing vernacular fiction to the foreground of literary studies. Since then, much has been written by both Chinese and Western scholars on Shuihu zhuan and other early Chinese vernacular novels.80 In the last few decades Western scholars have offered many critical insights that have vastly enhanced our understanding of the genre. When it comes to the relationship of the rise of vernacular fiction to the oral and popular traditions, however, their opinions bifurcate in a way similar to what happened among the Chinese literati scholars of the Ming-Qing periods. A slight oversimplification helps make it clear. Modern critical views on the rise of Chinese vernacular fiction have developed along two divergent

Told or Written

55

lines: (1) The close ties to popular storytelling were largely responsible for both the moral and aesthetic “limitations” of early vernacular fiction; and (2) vernacular fiction is a refined art form, for it is a phenomenon of the literati culture, and its connections to oral traditions are not essential to its nature. While prevalent in its own time, the first trend of critical thinking on early Chinese vernacular fiction is now largely out of vogue, but a brief engagement with that trend is pertinent to our discussion of the incipience of the genre in general and the formative period of Shuihu zhuan in particular. Claiming to be disappointed by the formal as well as moral limitations of the early Chinese vernacular narratives,81 scholars of this camp compare these works unfavorably with their counterparts in the Western novel.82 The formal limitations of the Chinese works are attributed to what was considered as the detrimental influence from popular storytelling.83 The devices and conventions of oral performance, as the theory goes, survived the evolutionary process from orality to writing, persisted in “versions designed to be read,” and became undesirable “literary clichés” as a result.84 Such formal flaws are found to be particularly annoying in Shuihu zhuan, where those formulary phrases usually associated with the raconteur’s art are dismissed as “conventional hyperbole that add nothing to our actual apprehension,” each being “a storyteller’s cliché that could as well be omitted.”85 The “most disturbing” influence of the raconteurs on the vernacular narratives is believed to be the “heterogeneous and episodic quality of plot,” resulting in the vernacular narratives’ preoccupation with “surface reality” and failure to offer a coherent pattern of meaning.86 Again, this conviction of Chinese vernacular fiction being episodic in plot would be particularly applicable to Shuihu zhuan, a narrative that may be, as we have seen, an agglomeration of materials originally from different popular sources. Early Chinese vernacular fiction is also believed faulty in its moral standing, for which popular orality is found once again to be responsible. With most of their thematic material inherited from oral traditions, the “moral ambiguity” in some vernacular narratives is considered almost as a symptom of a congenital malformation, as much of the material from the popular sources is said to be “frankly pornographic or immoral in nature.”87 Again, it is Shuihu zhuan that has to bear the brunt of the censure, for it is called a book “full of scenes of brutality and sadism, representing only one extremity of Chinese culture.”88 Admittedly, the criticism on the formal features of the early Chinese vernacular narratives is basically true, for the narrative structure is often episodic in nature and the formulary phraseology does at its extreme appear irritating to those allergic to “literary clichés.” What one may wish to contradict is therefore not the descriptions of the characteristics of the narrative form but the declaration or suggestion that such a narrative form represents artistic inferi-

56

Chapter 2

ority. The Homeric epics, crystallized from a long oral tradition, and the numerous literary epics in Western literature from the Aeneid on were no longer meant to be sung but to be read in their book forms. They can all be called— to different degrees—“episodic” in their narrative structure, and they are full of traditional phraseology in their versification, but few people would challenge their artistic excellence. The reason is that literary critics always have to maintain a generic differentiation, by which the epics are distinguished from later forms of narrative literature and therefore not subject to the critical principles governing later narrative art. Even within that more recent genre of Western narrative literature—the novel—generic differentiation has never ceased to function as a critical principle. Works such as Gil Blas, Moll Flanders, and Roderick Random are usually accepted as good literature, and their episodicity, under the protection of the generic label of the “picaresque novel,” has not suffered much from unfavorable comparisons with a Clarissa or a Liaisons Dangereuses. If generic diversity is taken for granted and no all-governing norms can be established even within Western narrative tradition itself, then there should be no justification for imposing features of some of the Western narrative works as standards of quality upon a form of narrative literature in a totally different culture. W. L. Idema’s criticism of this type of Eurocentrism is forceful and pertinent: “We should not compare the totality of Chinese ‘fiction’ with some selected summits of western ‘fiction,’ without going into their different places in their perspective literatures, and without a thorough awareness of the difference in the concepts of literature involved. This is like comparing champagne to Shaoxing without being aware of the difference between wine and jiu.”89 Calling for an enhancement of our awareness of literary multiplicity, Eugene Eoyang adopts a different but equally vivid analogy: While we eat the peach, we should “taste the apricot” as well.90 As for the moral censure of the early Chinese vernacular fiction, especially of Shuihu zhuan, one may, once again, concede that those reprehensions are not totally unjustified. A modern reader can hardly read without a shudder the passage in Shuihu zhuan where Wu Song, in an act of vengeance, struggles to cut off a woman’s head with a dagger that has been “blunted by too much killing” (chapter 31; 2: 974). Similarly, the penalty for the adulteress Pan Qiaoyun, a brutal dismemberment executed by a cuckolded husband and his wronged friend, is to the modern sensibility definitely out of proportion to her crime (chapter 46; 3: 1527–1528). Even Song Jiang himself, who is often depicted as a man of generosity and magnanimity, can be so carried away by the malicious desire for vengeance that he has his enemy dissected alive and orders that his heart be prepared as an hors d’oeuvre (chapter 41; 3: 1342–1343). We might, however, compare this spirit of vengeance with that in the

Told or Written

57

Iliad, where part of the narrative is about Achilles’ terrible anger ignited by Hektor’s killing of his beloved companion Patroklos. When Achilles has finally taken his revenge by killing Hektor, he and his comrades-in-arms ruthlessly mutilate the Trojan warrior’s body before tying it to a chariot and letting the running horses drag it with the head tumbling in the dust. To get a sense of the spirit of vengeance that is as irrational and violent as anything in Shuihu zhuan, let us quote the verses at some length: He spoke, and pulled the brazen spear from the body, and laid it on one side, and stripped away from the shoulders the bloody armor. And the other sons of the Achaians came running about him, and gazed upon the stature and on the imposing beauty of Hektor; and none stood beside him who did not stab him; and thus they would speak one to another, each looking at his neighbor: “See now, Hektor is much softer to handle than he was when he set the ships ablaze with the burning firebrand.” . . . In both of his [Hektor’s] feet at the back he [Achilles] made holes by the tendons in the space between ankle and heel, and drew thongs of ox-hide through them, and fastened them to the chariot so as to let the head drag, and mounted the chariot, and lifted the glorious armor inside it, then whipped the horses to a run, and they winged their way unreluctant. A cloud of dust rose where Hektor was dragged, his dark hair was falling about him, and all that head that was once so handsome was tumbled in the dust. (XXII, ll. 367–403)91

It is significant to note that such details did not go unnoticed in subsequent ages. In eighteenth-century England, when Henry Fielding was trying to create a new type of fiction modeled on the classical epic, some of his fellow writers found the moral aspect of the Homeric poems unacceptable. Alexander Pope, despite the homage he paid to the Greek minstrel in the famous line that “Nature and Homer were . . . the same,” found a “most shocking” thing in the Homeric poems, “that spirit of cruelty which appears too manifestly in the Iliad.”92 Samuel Richardson even held the Iliad responsible for the belligerent spirit of subsequent times: “I am afraid this poem, noble as it truly is, had done some infinite mischief for a series of ages; since to it, and its copy the Eneid, is owing, in a general measure, the savage spirit that has actuated, from the earliest ages to this time, the fighting fellows, that, worse than lions and tigers, have ravaged the earth, and made it a field of blood.”93 I am not suggesting that Shuihu zhuan is a work derived from a kind of orality similar to that which produced the Homeric verses. As I stated earlier,

58

Chapter 2

popular orality in China that led to the rise of vernacular fiction was enormously different from what Walter Ong terms the “primary orality” of Homeric Greece. The comparison between the violent spirit in Shuihu zhuan with the vengeful wrath in the Iliad is meant to show that, if the early Chinese vernacular fiction can sometimes appear wanting in civility to our modern sensibility, it is in that respect not very different from the Iliad, which has been a major source of inspiration for Western narrative tradition. This is not to suggest that oral people are inevitably more violent and brutal than the literate. Yet as we have learned from scientific research, people in an oral culture are more susceptible to “states of confused excitement” due to “a lack of systematized fancy or delusions acting as ego defenses.”94 Achilles’ wrath may have derived from such a “blind frenzy,” which, due to “confused cultural memories,” became a persistent convention in literary epics all the way down to Orlando furioso, even when it “had grown less understandable and more palpably histrionic as social conditions and personality structures changed with the growing effects of literacy.”95 In a similar way, it is possible that the violent revenge in Shuihu zhuan, often presented in hyperbolic words, is a convention that can be traced very far back in the oral tradition. Like the furious violence of Achilles or of Orlando, Wu Song’s wrathful killing gives him an epic stature. Such characters are products of oral noetic processes, which operate most effectively with outsize figures or “ ‘heavy’ characters, persons whose deeds are monumental [and] memorable.”96 Hyperbolic rendering of the violence would help create a type of such “heavy” characters and achieve an effect of “sensationalism” in the delivery of the story. So, just as the vengeful savagery in the Western epics should be viewed in the context of the conventions they inherited from their previous oral mode of existence, so should that in Shuihu zhuan.

Recent Western Reassessment of Shuihu zhuan In the last two or three decades, the study of Chinese vernacular fiction in the West has reached an unprecedented level of sophistication. Indeed, all students of the genre today, in the West as well as in China itself, have to be grateful for the brilliant scholarship of a new generation of Western sinologists who have promoted Chinese vernacular fiction to the status of respectable literature that it fully deserves, washing clean the stigma of “limitations” and revealing instead its limitless artistic possibilities. Amid the new boom of the study of Chinese vernacular fiction, there emerged another trend of critical thinking on the origins of the genre, which started with the much-needed redress for the failure to consider the best achievements in vernacular fiction as part of the mainstream culture. The scholars have challenged, forcefully and rightly, the older view that regarded vernacular fiction indiscriminately as a form of popular entertainment and put the

Told or Written

59

genre virtually in the category of subculture. Andrew Plaks, for instance, complains that scholars of Chinese vernacular fiction in the past overemphasized its connection with the popular tradition to such an extent that “the more crucial role” of historiography has been neglected. From there he argues that the great works of Ming vernacular fiction can “lend themselves to most meaningful interpretation when they are treated not as examples of a ‘popular’ counter-culture, but rather as major documents in the mainstream of Ming and Qing literati culture.”97 In his monumental Four Masterworks of the Ming Novel, Plaks has formally established the concept of the “literati novel” (wenren xiaoshuo) with the argument that the masterworks in vernacular fiction, like the “literati painting” (wenren hua), display many of the presumptions, aesthetic expectations, and a sense of self-realization that were typical of the literati culture.98 W. L. Idema also calls for a more discriminate treatment of the works in vernacular fiction. He argues for a division of the colloquial fictional narratives in the Ming and Qing periods into two different categories, which he designates as “chapbooks” and “literary novels” respectively. While putting works like Sanxia wuyi and Shuo Tang in the first category, Idema uses the second label for those more esteemed works in the genre.99 In this critical context, the role of popular orality in the formative period of vernacular fiction has often to be downplayed. Indeed, to the older trend of disparagement of Chinese vernacular fiction, the discount of the pertinence of orality may be an essential and necessary antidote. Yet one has to wonder whether the repudiation or disregard of its pretextual life is indeed the right price that early vernacular fiction should have to pay in order to earn due respectability. To be sure, nobody really denies the connection of early vernacular fiction to popular orality, but there seems to be a reluctance to acknowledge that connection as a shaping force on the new literary genre, as the ties of a text to its possible oral predecessors cannot always be proven by tangible evidence. Understandably, people feel compelled to focus instead on the text-reader relationship in literary reception, attenuating the relevance of an early vernacular narrative’s possible preprint existence. The argument is plausible: Since the text was meant to be read, it should be considered as written literature. And that argument seems to be corroborated by the rightful abrogation of the traditional theory that huaben and pinghua were scripts or promptbooks used in actual storytelling.100 Although Shuihu zhuan is not called a pinghua, this argument can easily be applied to the novel as well: Not only was the work meant for reading, the different recensions of the work—the so-called fanben (full editions) and jianben (simplified editions)—might be catering to different types of readership! It is indisputable that the printed texts of early vernacular narratives were

60

Chapter 2

meant for a reading public from the very start. But to reject the old belief that a huaben or a pinghua was a storyteller’s scenario is one thing; to repudiate the kinship of early huaben or vernacular sections in pinghua to the oral tradition is quite another. A huaben or a pinghua does not have to be a storyteller’s promptbook to be related to an oral tradition. The text could be situated at any point along the line of transition from the oral to the written without being directly used as a promptbook in an actual oral presentation.101 Indeed, the text of an orally derived narrative is in itself a paradox, from which no modern student of any oral literature from an age predating tape recording would be able to find an escape. For however deeply the object of one’s study was once immersed in an oral tradition, one would still have to have some written or printed text on the desk that has evolved one way or another from what was once told by word of mouth, and that text is, of course, meant to be read. Still, while it is not only justified but also necessary to keep in mind the difference between the text and the voice in terms of the mode of transmission and the psychological dynamics of reception, to sever the ties of the text from its oral precursor on the ground that the text was meant to be read is simply to annihilate altogether any orality studies that predate acoustic recording. The printed texts of the Iliad and the Odyssey—let us return to the most convenient examples—have been for many centuries meant for a reading public, of course; but if we stress the distinction between the text and the singing of the minstrel to an undue extent, we will deprive ourselves of the only access to the Homeric oral world. That is, the original oral mode of “production” of the poems cannot be denied simply on the grounds of the subsequent literate/literary mode of “consumption.” Some early vernacular texts, especially the vernacular sections in the pinghua, feature a prose that is obviously uneven and awkward, which means that they are not quite textual equivalents to the stories once told in oral prose. One important reason was the predominance of wenyan. In the tentative writing of the vernacular prose, the influence of wenyan would loom large and its interference could sometimes be irresistible. In the case of pinghua, the limited vernacular prose is often outweighed by the textual incorporations from classical sources. Yet while such borrowings contribute to the pinghua, which are generally considered vernacular works, they actually remain in the classical language and contribute nothing to those sections where the language is more vernacular. In other words, while most of the wenyan segments of the pinghua texts were compiled on the basis of a variety of textual sources, the same cannot be said of the vernacular sections, which are much more likely to have had some kind of kinship to orality. Scholars have argued, quite justly and properly, that Shuihu zhuan and other Ming masterworks in vernacular fiction brought Chinese narrative literature

Told or Written

61

to an unprecedented artistic level. Recent critical analyses, especially those by Andrew Plaks in his Four Masterworks of the Ming Novel, have greatly enhanced our knowledge of the textual intricacy and artistic sophistication of these narratives. We have come to realize that these works, once dismissed by some literati readers as coarse and crude, are in fact pregnant with ironic and allegorical meanings.102 Yet the exposition of the textual richness, especially the rhetorical stance of irony, sometimes carries with it the implication—if not explicit proposition—that the possible pretextual life of the early vernacular narrative no longer matters. The textual sophistication of Shuihu zhuan, for instance, can be taken as a sign that the work, by thoroughly transforming the “source material,” had far distanced itself from its oral past, which was too primitive to bear much relevance to the work’s textual existence.103 Since what is interpreted as an ironical rhetorical stance is taken as counterevidence for the kinship to popular orality, one may wish to raise the question whether a narrative of oral or popular origins is categorically incapable of being interpreted ironically or allegorically. A case in Western literature that may come to mind at this point is the Neo-Platonist Porphyry’s reading of the Odyssey, in which Homer’s verse describing a cave in Ithaca with one entrance for men and the other for gods104 is interpreted as the poet’s allegorical presentation of the cosmos itself.105 But a weightier example will be the tradition of the Christian exegeses following the Paulian aphorism that “The letter kills, but the spirit gives life.”106 Origen, for instance, admonishes the readers of The Song of Songs that it will be detrimental to understand the text literally in terms of the flesh: For if he does not know how to listen to the names of love purely and with chaste ears, he may twist everything he has heard from the inner man to the outer and fleshly man and be turned away from the Spirit to the flesh. Then he will nourish in himself fleshly desires, and it will seem because of the divine Scriptures that he is impelled and moved to the lusts of the flesh.107

Similarly, St. Augustine draws distinctions between things and signs and between the things to be used and the things to be enjoyed,108 on the basis of which he proceeds to exhort: “He who follows the letter takes figurative expressions as though they were literal and does not refer to the things signified to anything else.”109 In other words, the Scriptures, according to the early Church fathers, must not be read straightforwardly. Rather, they are also “roundabout writing[s] with deep seated meanings” (shenwen qubi), to adopt Jin Shengtan’s terminology.110

62

Chapter 2

If we accept the well-established theories that both the Homeric epics and much of the Christian Scriptures are of oral origin (although the modes in which they were tied to the oral traditions differed from each other), it will seem that the ironical and allegorical interpretations are by no means proofs of the literary nature of the origin of any text. Nothing prevents an orally derived text from being interpreted into something different from its most obvious meaning. Inversely, a text’s capacity for being interpreted ironically or allegorically does not preclude the possibility of its oral provenance. We should not take for granted any connection between the type of the textual meaning and the mode of the textual origination. If a work, at a certain point in the history of its reception, acquires an elite status in literature, it is made so by the intellectual climate of the time. Many a work of oral or popular origins in Western literature has found its way into what is called the “literary canon.” In Chinese literature, the best-known example is, of course, the Shijing, a collection of folk lyrics and ceremonial songs of the court during the Spring and Autumn period, which was later appropriated by the mainstream culture and became a classic for the literati. After that, some of the poems received many rounds of allegorical interpretations. That, however, does not contradict the fact that many poems in the Shijing are of oral provenance.

On the Bifurcated Critical Attitudes toward Shuihu zhuan and Its Oral Provenance This chapter has presented an overview of the formative stage of Shuihu zhuan in some genres of oral performance and discussed divergent views on the oral provenance of early vernacular fiction in general and that of Shuihu zhuan in particular. I have noted that the two different attitudes toward Shuihu zhuan’s preprint life that were prevalent during the Ming-Qing periods have largely been taken over by our own age. While some scholars denigrate the work and other early vernacular narratives for their inceptions in popular orality, others praise them for being able to break away from oral sources or for being less involved with popular materials than generally supposed. Despite their different assessments of the narrative art, both groups seem to agree that the oral tradition belonged to a humble and primitive past, a mire from which Shuihu zhuan would have to lift itself before reaching the solid ground of fine literature. This bifurcation of the critical views on Shuihu zhuan coincides with such a binary logic: The work should be considered either a work derived from a tradition of popular storytelling or an artifact produced by sophisticated literary sensibility. It cannot be both. Like the theory of orality once held by A. B. Lord, which refuses to see any intermediate zone between oral and literate

Told or Written

63

modes of composition and transmission, this binary logic puts Chinese literary tradition and the tradition of popular orality from the Southern Song to the Ming period into two mutually exclusive categories. Yet in the social conditions of premodern China, different cultural forces were constantly reshaped and reshuffled owing to the drastic social mobility. Especially, when Chinese narrative literature needed to explore new possibilities beyond wenyan, the forces of popular orality happened to be firmly allied with the forces for change within the literary culture. Under these historical circumstances, Shuihu zhuan as the earliest vernacular novel was a great literary innovation launched by men of letters, and it could be such precisely because much of its narrative discourse had originated in the spoken words of the storytellers. It is therefore both an orally derived narrative to a great extent and a work conceived by a literary sensibility, and it was this dual nature of its origins that made the narrative both truly vernacular and truly “literary,” marking the inauguration of vernacular fiction as a new literary genre. As I noted earlier, in more recent scholarship on early Chinese vernacular fiction, the downplaying of the impact from the oral tradition was often meant to be a rectifying measure for an earlier tendency to slight the genre. To correct a wrong, overshooting the target is perhaps inevitable. The seeming negligence of the role of the oral tradition is therefore quite understandable. It is even more so when we take into account the fact mentioned earlier: Whatever we say about the oral antecedents of the earliest vernacular narratives, the facts that we know about the evolution of those narratives in the popular genres remain meager. In the case of Shuihu zhuan, that the narrative developed from an oral story complex seems beyond any doubt, but many of the details about that development—without discovery of further evidence—will remain hypothetical. The brief discussion of the relevant oral genres earlier in this chapter is a start, but if the study of the impact from orality is based exclusively on the tenuous evidence of those few historical facts, one might end up making much ado about nothing. To certify the role of oral storytelling in the making of the narrative, and especially to substantiate my conjecture of a process of synthesis of the Shuihu stories in an oral narrative mostly in prose, I now have to turn to the text of Shuihu zhuan itself.

3 The Narrative Pattern The Uniform versus the Multiform If indeed the narrative discourse in Shuihu zhuan is to a large extent of oral provenance, what is the most convincing textual evidence of the ties to its oral antecedents? Scholars in the past focused on some formal features, in Shuihu zhuan and in other vernacular narratives, that were considered devices in storytelling before they survived the transition from the oral to the written, or, in C. T. Hsia’s words, “a storyteller’s clichés.”1 These features include the frequent use of formulary phrases such as “huashuo,” “queshuo,” “qieshuo,” “buzai huaxia,” and so on, which could have been part of the storyteller’s jargon; suspense at the end of each chapter, supposedly copied from the storyteller’s strategy to lure the audience back to the next session; and the narrator’s occasional addresses to the reader, in the same manner as a storyteller would address his audience. The conviction that such formal characteristics in vernacular fiction constitute evidence of a strong kinship to oral storytelling has been called in question. W. L. Idema argues forcefully that these formal features in the “storyteller’s manner” should be considered not as residues from popular orality of the past but as a deliberate choice by the writers in the new literary genre: “If these compilers or authors would choose to adopt some of the storyteller’s conventions, we cannot treat such elements as leftovers from the past, but we have to consider them as deliberate artistic inventions.”2 What Idema calls “deliberate artistic inventions” are defined by Patrick Hanan in more narratological terms. The invented elements, according to Hanan, are not merely some isolated and sporadic narrative expediencies. Put together, they create a narrative frame resembling the milieu for oral communication, or a “simulated context,” as Hanan terms it:

64

The Narrative Pattern

65

“Simulated context” means the context of situation in which a piece of fiction claims to be transmitted. In Chinese vernacular fiction, of course, the simulacrum is that of the oral storyteller addressing his audience, a pretense in which the author and reader happily acquiesce in order that the fiction can be communicated. It is not only a “mimesis of direct address,” it is also a mimesis of direct reception. In fact, it imitates a complete linguistic situation.3

Hanan’s comment appears, most judiciously, in an article on Ling Mengchu’s stories. For the later writers of vernacular fiction like Ling Mengchu, written vernacular prose was already fully established. They could write simply by following the models of earlier works in the genre, without having any direct contact with storytelling themselves. In fact, the generic term “nihuaben” (imitative huaben), which refers to the vernacular stories written by Feng Menglong, Ling Mengchu, and others, clearly points to such a stage of literary development.4 Yet if these writers all felt the need to achieve a simulacrum of oral communication in their writings, it can only mean that they were under the pressure of a well-established generic convention. By definition, a literary convention takes repeated practice to establish and to enforce, and therefore it necessarily indicates the maturity of the genre. In the early stage of vernacular fiction, however, when it was too young a genre to have any conventions of its own, it could still be possible that such formal features were copied from oral storytelling. Nevertheless, the theory of simulated orality in vernacular fiction as the writer’s craft is extremely illuminating and deserves tremendous credit. Since the storyteller-audience relationship in a vernacular narrative could be a genuine inheritance from an oral model in some cases and a deliberate choice by the writer as a simulated narrative frame in others, the nature of those formal characteristics in the “storyteller’s manner” is ambiguous. They are, therefore, not reliable indicators of the narrative’s relationship to orality. In this chapter I will look into Shuihu zhuan to see how the narrative’s connection to the oral tradition is reflected textually in the work. I will simply disregard those moments of the “storyteller’s manner,” for the reason just discussed. Instead, I will investigate the narrative discourse to see whether the economy of story making can be better explained in terms of oral composition and oral performance. My study will be on the fanben text, not only because it is earlier than any known editions in the jianben recension but also because it best demonstrates the possible influence from the oral tradition, as I believe. In any case, Shuihu zhuan should not be considered as based on an oral narrative “in the absolute,” as the narrative must have undergone extensive editing and compilation before it appeared in print. But a proper measurement of its

66

Chapter 3

“degree of orality” can be vital to a better understanding of its aesthetic qualities and a more appropriate appraisal of its place in the vernacularization of written prose and the development of premodern vernacular fiction.5 As I noted earlier, Shuihu zhuan came into being as a result of a long process consisting of repeated and cumulative efforts to align words on paper with words of mouth. Writing, or rather, “rewriting” as Henry Y. H. Zhao would call it, played a crucial part in the textualization of the narrative.6 To highlight orality, as I will do in this chapter, is by no means to neglect the role of writing and the textual evolution of the narrative, which will be considered later in the book. If Shuihu zhuan was the culmination of the vernacularization movement that had lasted for several centuries, that culmination could not have been achieved in any writer’s closed studio. For the “writers” of the earliest vernacular fiction had to “write” while listening attentively to the voices from the living orality.

A Thematic Study and Definitions of Terms Since Milman Parry’s early twentieth century epoch-making work on the classical epic, scholarly interest in oral and orally derived literature in various cultures has been getting stronger. The best-known part of Parry’s theory has been his discovery that the Homeric verses are made overwhelmingly of formulas and formulaic phrases, which made it possible for the epic singer to compose continually. But in the last years of his scholarly life, Parry proposed a radical shift of emphasis from the “formula” to what he termed the “theme.” Analogous to formulas on the level of verse making, themes are those regularly recurring units in the process of narration—either units of action or units of description. In his review of Walter Arend’s “Die typischen Scenen bei Homer,” Parry clearly considers the “theme” as another distinguishing feature of traditional oral poetry in addition to the “formula,” for the singers, according to Parry, “tended also to keep only a single set of details for a given action.”7 Parry’s theory on the patterning of narrative units in traditional oral poetry was later inherited and further developed by Albert B. Lord. Following Parry’s definition of formula, Lord defines themes as “groups of ideas regularly used in telling a tale in the formulaic style of traditional song.”8 In the introduction to his Singer of the Tales, Lord defines the words oral and oral epic in terms of both formulary diction in verse making and the use of themes in story making: Stated briefly, oral epic song is narrative poetry composed in a manner evolved over many generations by singers of tales who did not know how to write; it consists of the building of metrical lines and

The Narrative Pattern

67

half lines by means of formulas and formulaic expressions and of the building of songs by the use of themes. This is the technical sense in which I shall use the word “oral” and “oral epic” in this book.9

If formulary diction and the use of themes are two major principles in the composition of traditional oral epics, one may wish to raise some questions concerning the composition of an oral narrative that is mostly in prose. Can what has been said of the singer of tales be equally said of the teller of stories? Or do we have to produce another set of theoretical tenets for the study of oral prose narratives? Let us first of all consider formulary diction. Oral prose stories do often feature a kind of diction that is, in a broad sense, formulaic.10 But unlike the case with stories sung in verse, the language medium in prose stories permits more freedom in rhetoric and diction, for the syntactical elasticity of prose allows many more options for phrasing than the rigorous metrical pattern of verse. Although a formulary expression sometimes helps the teller of prose stories, there are many alternative phrases, and the teller does not feel any compelling need to resort to the formulary one. Therefore, if formulas are vital to singers of tales, they are only auxiliary to storytellers. For this reason, formulary diction probably cannot be taken as a starting point in the study of the oral narrative in prose. Admittedly, our knowledge of the actual storytelling in different periods of Chinese history is deficient. Yet if the extant texts that resulted from or were associated with the various oral genres reflect those forms of storytelling with reasonable faithfulness, we can divide the forms of storytelling roughly into two categories. The first category includes all the chantefables, such as bianwen, guzici, taozhen, changzhuan, zhugongdiao, baojuan, tanci, and many others. In these genres, verses to be sung alternated with prose to be spoken, but singing of verses to some kind of music was the dominant mode of presentation. In the other category are those genres more immediately under the rubric of shuohua, such as xiaoshuo and jiangshi. As the word “shuohua” would suggest, these genres gave a much larger role to speaking than singing. Accordingly they were for the most part in prose, and the role of verse was usually limited to some stock elements or summary of the narrative. The teller of a prose narrative, free from the pressure of the metrical rigidity of the verse form, enjoys a greater flexibility in diction than the singer of tales, but he shares the same problems with the singer on the level of poiesis, or composition. A short story lasting half an hour causes few problems, but for a long narrative running several days or even several weeks, as in the case of the story of Wu Song told by the famous contemporary raconteur Wang

68

Chapter 3

Shaotang, a strategy is needed to keep the narrative going. The storyteller must have everything ready at hand in order to keep the audience engaged and not have to stop and think of the next episode, which would prove embarrassing. Since the issue of poiesis in oral prose narration is similar to that in the singing of an oral epic, we have reason to believe that the principle of composition by themes in the case of oral epics may also be applicable to oral prose narratives. Shuihu zhuan’s immediate oral antecedent, as I hypothesized earlier, could have been a long narrative predominantly in prose. If a large portion of Shuihu zhuan is composed of regular patterns of recurring thematic units and such recurring units distribute over a long span in the narrative, it might suggest an oral mode of story making. In a pioneering article, Patrick Hanan examines some thematic units in his investigation of the narrative structure of Ping yao zhuan and makes it clear that the narrative is largely composed of stereotyped sequences that recur in other works of early vernacular fiction. Hanan considers the recurrence of such narrative material more likely to have taken place in oral fiction than an author’s writing: “On the whole, the kind of literary relations we have described would be consistent with development in oral narrative, although it would not prove that point.”11 In his discussion of the narrative structure in Shuihu zhuan, Peter Li remarks that “the fundamental pattern . . . is a cyclical chain or, more accurately, a sequence of cycles in each of which a different hero is featured.” Li considers the first seventy chapters dominated by eight such cycles that follow a similar “general pattern.”12 In what follows I will make a thematic study of the narrative discourse in Shuihu zhuan. I will call those more or less typical scenes and stock elements of action “segments.” A segment usually has no autonomous thematic meaning of its own and should not be confused with a “motif,” which usually carries a moral or philosophical connotation. Following Lord’s definition of themes, by segments I refer to “the groups of ideas regularly used” in the telling of a traditional tale.13 However, different from Lord’s theme in oral epics, which suggests a formulaic style of diction and therefore a verbal correspondence between its recurrences, a segment is a stock element in oral or orally derived tales that frequently recurs—but not necessarily with verbal repetition—due to the casual and elastic diction of prose. In Shuihu zhuan, such recurrent segments are abundant. I will not, however, isolate these segments out of context and deal with them individually. Instead, I will discuss them in larger stereotyped patterns of action, which I call “sequences.” A small detail, as Bernard Fenik puts it, becomes significant only “when it is found to be part of a larger cluster of items, or if it is always or usually found at a certain specific, definable juncture of action.”14 Accordingly, segments will be examined not as independent entities but as constituents of sequences. This is by no means a new method. William Hansen, in his study of the narrative patterning and incon-

The Narrative Pattern

69

sistency in the Odyssey, focuses not on those isolated “small narrative units for comparative analysis” but on “a group of closely related narrative patterns,” and narrative details are examined in terms of their “arrangements” in the “larger narrative sequence.”15 In the following analysis of the narrative patterning in Shuihu zhuan, blocks of narrative that are modeled on the same sequence will be termed “narrative stretches,” and different expressions of the same segment in parallel narrative stretches will be called “variants.” I will consider four narrative sequences in Shuihu zhuan, which are named respectively the “exile-imprisonment sequence,” the “tragicomic survival sequence,” the “sequence of beating the bully and defending the injured,” and finally the “sequence of slaying a sister-in-law.” Each of these sequences provides a more or less fixed pattern of story making followed in a number of narrative stretches. These sequences do not cover the whole narrative. There are, of course, many other patterned narrative units in Shuihu zhuan. But since each of these four sequences is responsible for the story making in a number of narrative stretches of significant length, they should be considered as good samples for the thematic patterning of the narrative.

The Exile-Imprisonment Sequence The “exile-imprisonment sequence” is one of the most obvious sequences in Shuihu zhuan. Its four major variant stretches tell stories respectively of four major Liangshan figures—namely, Lin Chong, Wu Song, Song Jiang, and Lu Junyi—each of whom undergoes exile and imprisonment in a similar pattern. The earliest of the four stretches is the story of Lin Chong (chapters 7–10). Framed by the powerful minister Gao Qiu and his son, Lin Chong is arrested. Lin’s father-in-law bribes the officials in hopes of having the case extenuated. Lin is sentenced into exile. He is tattooed, and two policemen, Dong Chao and Xue Ba by name, are ordered to escort him to his destination. Lin’s fatherin-law and wife come to see him off. Before they set out, the two policemen are invited to drink in a tavern by Lu Qian, Gao Qiu’s confidant, who bribes the two men into promising to kill Lin Chong during the journey. Lin Chong is badly tortured on the trip, and when he is about to be murdered, Lu Zhishen, a friend of Lin’s, jumps from a tree and comes to Lin’s rescue. After his arrival at the destination, Lin Chong sends money to the head jailer and the warden and is spared the beating of the “discipline club.” The Lin Chong stretch follows a scheme that runs like this: a. Bribery by a friendly party as a rescue effort b. Bribery by a hostile party to have prisoner killed c. Dear ones seeing prisoner off d. Two escorts’ good/bad treatment of prisoner on the way

70

Chapter 3

e. Escorts’ attempt to kill prisoner f. Prisoner saved by a rescuer g. Prisoner bribes captors to be spared a beating

Each of these elements, which we call “segments,” recurs in other stretches following the same “exile-imprisonment” sequence, as we are going to see. To be sure, not every stretch contains all seven segments. The Lin Chong stretch, which embraces all the segments, is the only one that reaches the plenum of the sequence, but the general pattern remains more or less stable in all four stretches. In the stretch that tells the story of Wu Song (chapters 27–30), segments a, b, c, d, e, and g appear, but in a different order as d, g, b, a, c, d, and e. Segment d appears twice, as Wu Song is twice exiled. He is first exiled to Mengzhou for having killed his adulterous sister-in-law and her lover, and later on in Mengzhou he is again exiled after General Zhang falsely accuses him of theft. Segment g somewhat deviates from its variants in other stories, for Wu Song refuses to offer bribes to the officials after his arrival in Mengzhou but, despite his defiance, the warden and his son Shi En not only spare him from a beating but also treat him kindly. Segment b follows, as when Wu Song is sent to jail by General Zhang, Jiang the Door-God, whom Wu Song has beaten earlier, bribes officials in the hope of having Wu Song killed in jail. After that, segment a appears, for Shi En also sends money to the officials, trying to save his friend’s life. After Wu Song is for the second time sentenced into exile, Shi En, like Lin Chong’s father-in-law and wife in the Lin Chong story, sees the prisoner off (segment c). Again, like Dong Chao and Xue Ba in the Lin Chong stretch, the two escorting policemen, with the assistance of two men dispatched by Jiang the Door-God, plan to kill Wu Song on the way (segment e). Segment f is absent from the Wu Song stretch, for Wu Song does not have a rescuer. But he may well be considered a dual agent in the action, as both prisoner and rescuer, for he saves his own life by killing the escorts one after another. The stretch that tells Song Jiang’s “exile-imprisonment” story (chapters 36–38) is somewhat less complicated, for it features only segments a, c, d, and g. Since killing his mistress Yan Poxi, Song Jiang has been away from home. On returning, he is arrested and put in jail. Song’s father bribes the officials to seek leniency (segment a). Song is tattooed and sent into exile to Jiangzhou. The father and the brother see him off (segment c). Song is escorted by two policemen, Zhang Qian and Li Wan, who treat him well on the way (segment d ). Segment g develops into a complication. Song Jiang, in offering bribes after his arrival in Jiangzhou, deliberately leaves out one official, Dai Zong, in order to provoke Dai into coming to see him in person.

The Narrative Pattern

71

Finally, we come to the stretch about Lu Junyi (chapters 61–62). Betrayed by his unfaithful wife and her lover, Li Gu, who is none other than his steward, Lu Junyi is arrested and imprisoned. From that point on, the stretch follows a route almost identical to that of Lin Chong. Both the friendly party (Chai Jin from Liangshan) and the hostile party (Li Gu) send bribes to the officials (segments a and b). Segment b adopts an additional variant in the familiar tavern scene, as it does in the Lin Chong story: Before they set out on the way to exile, Li Gu invites the two policemen for a drink in a tavern. He gives them gold in exchange for their promise to kill Lu during the trip. Lu Junyi is tortured on the way (segment d), and when he is about to be killed (segment e), Yan Qing, Lu’s faithful servant, comes to his master’s rescue (segment f ). Segment g is absent from the Lu Junyi story, for he does not arrive at the destination and therefore does not have to bribe the officials to be spared from a beating. The stretches of Lin Chong and of Lu Junyi are the closest parallels to each other and the closest to a full development of the sequence. In fact, there are many correspondences between the two stretches on the level of narrative details, too small to even qualify as segments. One conspicuous instance is that in both stories the two escorts are invariably Dong Chao and Xue Ba.16 In both stretches it is Xue Ba who appears to be the more unscrupulous, when the prisoner’s enemy (Lu Qian/Li Gu) bribes them in a tavern and asks them to kill the prisoner. In both stretches the two escorts torture the prisoner by scalding his feet with boiling water while pretending to help him wash. In both stretches the escorts tie the prisoner to a tree with the excuse of preventing him from escaping, before they prepare to kill him. Even the two rescuers, Lu Zhishen and Yan Qing, make their appearance in the same manner by jumping from the tree where they have been hiding.

The Tragicomic Survival Sequence The “tragicomic survival sequence” refers to the recurring narrative pattern in which the hero’s life is jeopardized in a critical situation before things take a dramatic turn: The fiendish captors turn out to be friends who have long admired the hero’s repute, although they have never met him before. This sequence is embodied in many stretches in the narrative, but only four major examples will be examined here, three of which concern Song Jiang, the other Wu Song. Song Jiang I (chapter 32): Song Jiang, on his way to visit his friend Hua Rong, stumbles over a trip cord and is seized by robbers from Qingfengshan who have been lying in ambush. Taken to the bandits’ lair, Song is tied to a pillar before the robber chieftains come and give the order to cut out his heart and liver. Song Jiang, muttering to himself and bemoaning his fate, utters his

72

Chapter 3

own name. He is overheard by Yan Shun, the big chieftain. Having interrogated the captive to confirm his identity, the astonished Yan invites the other chieftains to join him in kowtowing to Song Jiang. Song Jiang II (chapter 36): Song Jiang, escorted by two policemen, is on his way to exile in Jiangzhou. Feeling thirsty, they decide to have a drink in a tavern. While covetously eyeing Song’s full pack, the host drugs the wine, and shortly after drinking it the three travelers collapse. The host awaits his helpers before butchering the three, when Li Jun and two companions arrive. They check Song Jiang’s documents. When they find out Song Jiang’s identity, they pour an antidote down his throat. When Song finally comes to himself, everyone, including Li Li, the host of the tavern, kneels and kowtows four times. Song Jiang III (chapter 37): An old squire lets Song Jiang and the two escorts stay at his manor for the night. At midnight Song overhears people talking and finds out that the squire is the father of a man whom he has unknowingly offended before. Realizing the danger, they flee, and the young man and his brother chase them. As the pursuers approach, Song Jiang is alarmed to see his way barred by a river. A boat suddenly appears, and Song and his escorts are allowed to get on board. The boatman turns out to be a robber, and when the boat reaches the middle of the river, he offers the three men the choice either to be killed with a blade or jump into the water to drown. At that moment a small craft reaches them. On the craft stands Li Jun, who recognizes Song Jiang. At this juncture the stupefied boatman Zhang Heng drops to his knees before Song and kowtows. When they get back to the bank, Li Jun reveals the prisoner’s identity to the Mu brothers, who chased Song Jiang to the riverbank a moment ago, and the two brothers also pay their respect by kowtowing to Song. Wu Song (chapter 31): Wu Song, weary after long travel, puts up for the night in an old temple. He has scarcely fallen asleep when he finds himself tied up, carried to a nearby village, and pushed into a slaughterhouse. The captors wait for their chieftain to start, and a moment later the man and his wife come in; they turn out to be Wu Song’s sworn brother Zhang Qing and his wife Sun the Witch. The woman recognizes Wu Song and she relates Wu’s heroic deeds to the ruffians, who fall to their knees and kowtow to their captive. Although it can become a narrative stretch of considerable length, the “tragicomic survival sequence” is not in itself complicated. The scheme of the sequence runs as follows: a. Hero caught or held under control b. Hero’s life jeopardized c. Hero’s identity revealed d. Hero saluted by captors

The Narrative Pattern

73

All four stretches examined above invariably contain these segments, and the order of the segments remains fixed. Segment d is not only a constant in this sequence but frequently makes its appearance in other sequences or in shorter narrative passages as well. The captured generals of the government troops, for instance, would without exception kowtow to Song Jiang when overwhelmed by the latter’s profuse display of kindness. The other three segments are, however, less stable. They diverge into somewhat different shapes in different situations. Segment a, for instance, finds a variant in each of the four stretches: Song Jiang I: Hero caught when stumbling over a trip cord Song Jiang II: Hero drugged and rendered powerless Song Jiang III: Hero hijacked on the river Wu Song: Hero seized and tied while sleeping

Yet these variants are in themselves stock elements that are readily used elsewhere in the narrative. The variant in Song Jiang I certainly looks familiar to the reader of Shuihu zhuan, for a number of officers of government troops are caught by the Liangshan bandits when they are tripped by a cord. The variant of the segment in Song Jiang II recurs frequently in the whole work. Here are some of the major scenes involving drugged wine: Yang Zhi and his soldiers are drugged, and the birthday gifts that have been entrusted under their convoy are taken away (chapter 16); Wu Song is served drugged wine at the Crossroads Rise Tavern and pretends to be drinking it (chapter 27); Dai Zong, while serving as Prefect Cai’s envoy to his father, Prime Minister Cai, in the capital, is drugged at Zhu Gui’s tavern (chapter 39); and Zhu Fu and Zhu Gui serve Li Yun, the police chief, drugged wine and meat in order to rescue the arrested Li Kui (chapter 43). Similarly, the variant of the segment in Song Jiang III— hero being hijacked on the river—reappears in a parallel scene where Zhang Shun is tied up and thrown into the water by a boatman when the ferryboat reaches the middle of the river (chapter 61).

The Sequence of Beating the Bully and Defending the Injured There are three major episodes in Shuihu zhuan in which the hero, outraged at an injustice, takes it upon himself to fight a bully on behalf of the wronged party. The first episode is about Lu Zhishen’s deadly beating of Zheng the Butcher, a rascal who has been pressing a poor singing girl and her father to pay him a debt they never owed (chapter 3). The second episode features Wu Song, and it is in conjunction with the hero’s “exile-imprisonment” story. Wu

74

Chapter 3

Song, as the story goes, conquers Jiang the Door-God and restores Shi En to the ownership of the tavern that Jiang seized from him earlier (chapter 29). The last episode tells a story of Li Kui, who, with the assistance of his friend Yan Qing, fights and kills two robbers and returns an abducted girl to her anxious parents (chapter 73). All these episodes follow a similar pattern, the “sequence of beating the bully and defending the injured.” Let us call these three episodes the Lu stretch, the Wu stretch, and the Li stretch, respectively. The scheme of the sequence runs as follows: a. Hero learns of the grievances of the injured b. Hero provokes the bully c. Hero fights the bully d. Justice restored

Segment a appears in all three stories. In the Wu stretch, Wu Song’s curiosity leads him to learn about Shi En’s grievances. He has been treated by Shi En very kindly, and he presses Shi En for the reason. The latter tells him of what has happened to his tavern. In each of the other stretches, the Lu stretch and the Li stretch, the episode starts ironically: The hero is annoyed with the aggrieved party. In the Lu stretch, Lu Zhishen, while drinking in a tavern with two friends, hears people crying next door and becomes furious at the disturbance. He summons and interrogates them, and they turn out to be a singing girl and her father. In the Li stretch, Li Kui and his friend Yan Qing seek lodging at a manor and are told that the squire is having trouble. After quarreling with the vassals, they are given a place to stay for the night by the squire himself. That night, Li is irritated to hear the squire and his wife crying next door. He goes in the morning to question the old couple and learns about their grievances. The variants of segment a in the Lu stretch and the Li stretch appear very alike, for both are triggered by the hero’s annoyance at a disturbance next door. At the same time, however, the variant in the Li stretch itself contains a scene that frequently recurs in the narrative—the scene of a traveler seeking lodging at a manor. Lu Zhishen, for instance, comes to a manor to ask for a place to put up for the night, only to be refused by the vassals, before the squire finally intervenes and invites him to stay (chapter 5). The pattern is strikingly similar to that in the Li stretch. Segment b, the hero’s provocation of the bully, is conspicuously absent from the Li stretch, but it plays an important role in the other two stretches. In the Lu stretch, Lu Zhishen tells Zheng the Butcher to prepare for him ten catties of lean meat, ten catties of fat meat, and ten catties of gristle, all chopped fine,

The Narrative Pattern

75

a task too mischievous to be real business, which enrages Zheng and stirs up a fight. In the Wu stretch, the half-drunken Wu Song enters the tavern of Jiang the Door-God and, after demanding the best wine, asks Jiang’s concubine to come over and keep him company at the table, a blatant sexual overture. The girl is infuriated and charges at Wu Song, who throws her and two of the waiters into the open wine vats, which finally draws Jiang the Door-God into the clash. Sexual overtures to a wife/concubine of a tavern host as a means of provocation is in itself a recurring element in the narrative. In another of Wu Song’s stories (chapter 27), for instance, Wu Song repeatedly asks the hostess for the best wine and, being aware that the woman is harboring malicious designs against him, freely teases her on her “loneliness” with her husband away from home. Lu Zhishen’s fight with Zheng the Butcher and Wu Song’s fight with Jiang the Door-God are obviously based on the same pattern. In either case, an anticipated struggle between giants turns out to be a lopsided battle. While Lu Zhishen kills Zheng with three blows of his fist, Wu Song conquers Jiang with two kicks of his foot. In the Li stretch, Li Kui slays the two abductors with two whacks of his ax before finding the abducted girl weeping in the bedroom and restoring her to her parents. There are several parallel episodes in the novel that involve freeing a woman from abduction. At the Waguan Monastery, Lu Zhishen and Shi Jin encounter a priest and a monk who have ruined the monastery and abducted a girl. Lu and Shi kill the abductors, but the girl, when she is free, drowns herself in a well (chapter 6). A similar episode occurs in chapter 31: On passing a cemetery temple, Wu Song spots a priest embracing a girl by the window. Wu fights the priest, kills him, and releases the kidnapped girl. Again, in chapter 32, Wang the Stumpy Tiger captures an official’s wife and takes her to his bandits’ lair, where he intends to keep her as his woman, before Song Jiang finally persuades him to release her. A slightly more elaborate variant of this relatively fixed narrative element is found in another Lu Zhishen story, where Lu lies naked in the bridal bed and waits for the tyrannical groom Zhou Tong. Lu beats Zhou and later befriends him in order to persuade him to free the girl from the forced marriage (chapter 5).

The Sequence of Slaying an Adulterous Sister-in-Law There are in Shuihu zhuan two long narrative stretches, each telling of the killing of an adulterous sister-in-law. The first is a story about Wu Song (chapters 24–26), the second a story about Shi Xiu (chapters 44–46). There is one conspicuous difference. In the Wu Song story the hero kills the sister-in-law

76

Chapter 3

to avenge a murdered brother, while Shi Xiu’s slaughter of the woman is justified to a lesser extent—probably to defend the name of a sworn brother as well as his own. But both stretches are based on strikingly similar configurations of narrative segments.17 Some of the segments in the sequence and their variants in the two narrative stretches will be examined here, with W indicating a variant in the Wu Song stretch and S a variant in the Shi Xiu stretch: a. Meeting of brothers W. Wu Song, newly made a police officer, meets his brother Wu the Elder on the street. S. Yang Xiong, a warden of the local prison, has a scrimmage on the street with a group of soldiers, and Shi Xiu, a passer-by, comes to his assistance. The two become sworn brothers. (Both Wu’s and Yang’s professions are familiar elements in the narrative, for a good number of Liangshan heroes were police officers or petty officials before going to Liangshan.) b. Hero saluting sister-in-law and invited to live in the house W. Wu Song is brought home by Wu the Elder and, on seeing Pan Jinlian, his sister-in-law, Wu Song kowtows. Pan courteously returns the greeting. She invites Wu Song to live with them. S. Shi Xiu is brought home by Yang Xiong and introduced to the latter’s wife, Pan Qiaoyun, whom Shi calls “sister-in-law” and to whom he kowtows. Pan returns the greeting. She cleans a vacant room for Shi Xiu to live in. c. Hero becoming aware of the woman’s licentiousness W. Pan Jinlian makes sexual advances to Wu Song, who reprimands her. His displeasure with the woman grows gradually from “four- or five-tenths” (siwufen) to “eight- or nine-tenths” ( bajiufen). S. Pan Qiaoyun flirts with the monk Pei Ruhai. Shi Xiu secretly watches their flirtation. His awareness of what is brewing grows gradually from “one-tenth” ( yifen) to “ten-tenths” (shifen). d. Hero dismissed from the house W. In front of her husband, Pan Jinlian falsely accuses Wu Song of taking liberties with her. Wu Song moves out, and his absence is extended by a mission to send gifts for the magistrate to his relatives in the Eastern capital. (A gift-sending mission is a recurring element in the narrative. One may think of the similar mission for Yang Zhi in chapter 16, for instance.) S. In front of her husband, Pan Qiaoyun falsely accuses Shi Xiu of sexually assailing her. Shi Xiu moves out but lives nearby.

The Narrative Pattern

77

e. Removal of the husband W. Pan Jinlian poisons Wu the Elder to death. S. Yang Xiong has to spend most nights in his office. (A minor official’s bustling life is by no means unfamiliar to the reader of Shuihu zhuan. Examples are Dai Zong, Cai Fu, and Song Jiang himself.) f. Hero’s killing of the adulterer W. Wu Song fights with Ximen Qing in a tavern and follows him to the street to kill him. (Again, a tavern is a convenient place for a turn in the action, a point to which I will return later in the chapter.) S. Shi Xiu late at night kills the friar who is standing guard for the adulterer and then kills the lecherous monk himself. g. Hero forces a confession out of the sister-in-law before killing her W. Wu Song invites neighbors to dinner. At the dinner table he forces a confession out of Pan Jinlian and the procuress Old Woman Wang, an accomplice in the adultery. He kills Pan and takes the old woman to the magistrate. S. Shi Xiu has Yang Xiong trick Pan Qiaoyun and her maid into going to a secluded place, where the two men force confessions out of Qiaoyun and the maid, an accomplice. Shi Xiu persuades his sworn brother to kill both women.

Recurrences of Story-Making Material at Two Different Levels In each of the four groups, a number of parallel stretches share the same narrative sequence and follow a more or less stereotyped pattern of story making. While there are signs of flexibility and imaginative adaptation, the overall thematic structures in these narrative stretches are normally standardized or formalized. A comparison among the four groups of narrative stretches discussed above, however, indicates that they are not formalized in the same way. In the “exileimprisonment sequence,” segments are highly stable, with few radically divergent variants in the parallel narrative stretches. In other words, the narrative sequence is more or less self-sustaining, making little use of elements from outside to vary its morphological forms. For this reason, I call it a “closed sequence.” Parallel narrative stretches patterned on a closed sequence can be obviously similar. In the four “exile-imprisonment” stretches, for instance, what differentiates a bribery episode in one stretch from the corresponding episode in another is often merely the difference in the names of characters and places. A segment in the other three sequences, however, readily assumes diver-

78

Chapter 3

gent variants in the parallel narrative stretches. A segmental variant serves its function within the narrative stretch and more often than not is also an element repeatedly making its appearance outside the sequence. In geometrical terminology, such a segmental variant would be at the intersection of a horizontal and a vertical line—the horizontal line standing for its function in story making of the narrative stretch, and the vertical line being its kinship to its other occurrences elsewhere in the narrative, either within or out of the cycle of the parallel narrative stretches following the same sequence. Most of the narrative sequences in Shuihu zhuan contain a considerable number of scenes that recur not within but out of the parallel stretches, as the study above demonstrates. In one of the “tragicomic survival” stretches, for instance, Song Jiang is drugged and rendered powerless, and that episode fits into a largely fixed narrative pattern. While drinking drugged wine does not recur in any other parallel stretches modeled upon the same sequence, it is in itself a stock element that repeatedly appears elsewhere in the work. A narrative sequence like this can thus be called an “open sequence.” Most sequences in Shuihu zhuan belong to this type. Thus the pattern of story making in a great portion of Shuihu zhuan can be said to be recurrent on two different levels. On one level a stereotyped sequence occurs repeatedly in a number of parallel narrative stretches, while on the other level a segmental variant serving as a link along a narrative stretch can be a recurrent element, in most cases recurring not within the cycle of parallel stretches modeled on the same sequence but elsewhere in the narrative. That is to say, for a certain segmental slot in a narrative sequence, a number of different variants can be used; and on the other hand, the same variant can be used for different segmental slots in various places in the narrative. In his study of the classical epic, Lord has noticed recurrent patterns on these two levels. On the one hand, “the same basic incidents and descriptions are met with time and again”;18 on the other hand, the recurrence is found on the level of much larger blocks of narrative, which Lord calls alternately patterns, sequences, or complexes of themes. Lord points out that in these parallel narrative blocks, the overarching scheme remains the same while the details at corresponding positions may vary. “True,” Lord says at one point in his discussion of the parallel blocks in the Iliad, “the characters are different and the shift from the gods to Agamemnon is puzzling. But the sequence that Homer is following is a well established one.”19 On the following page he reaffirms the same idea: “This plan is different from the earlier one. . . . But the general sequence here is the same as that on the earlier passage.”20 Indeed, what we have seen in the narrative sequences in Shuihu zhuan is strikingly similar to the recurrent patterning of story-making material that Lord has discovered in the Iliad.

The Narrative Pattern

79

The Etic and the Emic in the Story-Making Economy The dynamic between what is stable and what is variable in the process of story making is certainly reminiscent of Vladimir Propp’s delineation of the morphology of Russian folktales. What Propp attempted to do was not only to isolate component parts of the tale but also—and more important—to observe each of these components in terms of its relationship to other components and to the whole tale. Different from the traditional “motifs” that are variable, Propp’s minimal narrative units—“functions,” as he calls them—“serve as stable, constant elements in folktales, independent of who performs them, and how they are fulfilled by the dramatis personae.”21 Alan Dundes has found this distinction between the motif and the function to be compatible with Kenneth Pike’s distinction between what he calls the “etic” and the “emic” approach in his Language in Relation to a Unified Theory of the Structure of Human Behavior, where the terms etic and emic are coined out of the last portions of the words phonetic and phonemic. The etic approach, according to Dundes, is “nonstructural but classificatory in that the analyst devises logical categories of systems, classes and units without attempting to make them reflect actual structure in particular data,” while in contrast, the emic approach is “a mono-contextual, structural one.”22 By incorporating Pike’s scheme into Propp’s theory, Dundes proposes to call the Proppian function emic motif or motifeme, and to use the term allomotif to refer to “those motifs which occur in any given motifemic context.” “Allomotifs would bear the same relationship to motifeme as do allophones to phonemes and allomorphs to morphemes.”23 In another article that deals with North American Indian folktales, Dundes recapitulates these ideas and proposes a “combined Proppian/Pike structural model” for folktales in general: In applying Propp’s morphological framework to American Indian folktales, I have adopted some of the terminology and theory of Kenneth L. Pike, as expressed in the latter’s Language in Relation to a Unified Theory of the Structure of Human Behavior. Propp’s function becomes thus a motifeme instead, which permits the associated notions of motif and allomotif. Folktales may thus be defined as sequences of motifemes. The motifemic slots may be filled with various motifs and the specific alternative motifs for any given motifemic slot may be labelled allomotifs.24

This “Proppian-Pike” model, if used as a supplement to the Parry-Lord theory of the theme, can be extremely helpful in delineating the story making in an orally derived narrative. Lord has noticed the difference and relationship

80

Chapter 3

between the stable and the variable and between the structural and the nonstructural, but he has failed to work out a theoretical paradigm for the storymaking operation. He has discussed the themes and parallel theme complexes as recurrent elements mostly in an etic sense, while the ways each of the recurrent elements fits into a larger narrative unit is not given much consideration. In this regard, the emphasis of Propp, Pike, and Dundes upon the emic approach and upon the dynamic in the building up of the narrative structure provides precisely what the Parry-Lord theme theory lacks. The story making in those sections of Shuihu zhuan discussed above will be best explained only when we combine the etic approach with the emic one. As has been shown, some of the segments—especially those in a closed sequence—are fairly stable, and the corresponding segmental slots they leave in the parallel narrative stretches are filled up with similar or even identical elements. Other segments—especially those in an open sequence—have their segmental slots in the parallel stretches filled by divergent variants, which may be themselves type scenes or action elements recurring elsewhere in the narrative. If we apply Dundes’ model and at the same time accommodate the ParryLord theory, we can call those alternative variants for any given segmental slot in a narrative sequence “allosegments,” each of which may be a recurring element in the narrative. In a closed sequence in Shuihu zhuan, an identical allosegment—or a group of very similar ones—is often used to fill the corresponding segmental slots in the parallel narrative stretches. Recall that in two stretches of the “exile-imprisonment sequence,” Lin Chong and Lu Junyi are escorted to exile by the same two policemen, who torture the prisoner and finally attempt to murder him in precisely the same way. In another stretch of the same sequence, the Song Jiang story, the names of the escorts are changed to Zhang Qian and Li Wan, which are just similarly stereotyped names. In most cases, corresponding allosegments in the parallel narrative stretches remain similar, while the agents of the action and small details may vary. In the Lin Chong stretch, the Song Jiang stretch, and the Wu Song stretch of the “exile-imprisonment” sequence, for instance, the segment of “dear ones seeing prisoner off ” varies into a group of allosegments that feature respectively Lin Chong’s wife and father-in-law, Song Jiang’s father and brother, and Wu Song’s sworn brother Shi En. Small details in these allosegments also have to vary in order to suit the different nature of the personal relationships—Lin Chong’s farewell speech to his wife and father-in-law is different from Song Jiang’s to his father and brother, for example—but basically these allosegments run in neat parallels. In an open sequence, however, corresponding allosegments in the parallel stretches may appear much less similar. For instance, in the “tragicomic sur-

The Narrative Pattern

81

vival sequence” discussed above, the segment “hero caught or held under control” leaves a segmental slot in each of the four parallel stretches, which is filled respectively by one of the following allosegments: 1. Hero caught while stumbling over a trip cord 2. Hero drugged and rendered powerless 3. Hero hijacked on a river 4. Hero seized and tied up while sleeping

These allosegments are specific expressions of the same segment in the sequence, but their contents are not simply recurrences of a shared pattern, as is the case in a closed sequence. This, however, does not amount to much inconvenience for the narrator in the story making, for most of these allosegments (1, 2, and 3) recur elsewhere in the narrative, as noted above, and they must therefore be stock elements as well. The above analyses allow us to see the network of thematic patterning in part of Shuihu zhuan. Along the line of a narrative stretch, the narrator did not have to worry about what to tell next, for that part of the narrative is modeled upon some standard narrative sequence with which he was already familiar from repeated telling. And across the lines of parallel stretches, corresponding segmental slots may be filled either with identical or similar allosegments—as in a closed sequence—or with divergent allosegments, each of which is readily available to the narrator as stock stuff—as in an open sequence. Therefore, in these sections of Shuihu zhuan, both the plot patterns and most of the narrative details are recurrent and therefore highly conventionalized.

Encoding and Retrieval in Oral Transmission In its early stage, the Shuihu complex had existed in multiple pieces comprised of separate stories. Even in a Shuihu zaju or a Shuihu cihua of the Yuan period, the scope of the action had to be rather limited. If those Shuihu stories from different genres of oral performance had to be synthesized in a long narrative predominantly in prose, as I proposed earlier, we have to raise questions concerning the overall structure of that long narrative. The thematic study here demonstrates that numerous narrative stretches in Shuihu zhuan simply follow stereotyped sequences. If within one of those stretches the narrator simply followed a conventional model in story making, how did the storyteller of the long, synthesized narrative keep the narration going beyond the stretch? How did the narrator manage to memorize where and how to make the shift into the next narrative block and therefore produce a narrative of such enormous length?

82

Chapter 3

The answer, at least in part, lies with memory, which is particularly an issue for oral narratives. A written narrative, even if its story may cover a span of hundreds of years and may take weeks to read, exists in a textual form in which all parts are concurrent, producing a spatial unity transcending temporal vicissitudes. A writer has the privilege of reviewing the foregoing parts of his story in order to decide what to write next, and a reader can thumb the pages back and forth to refresh his impression of what he has just read. But what an oral narrator has vocally uttered is irrevocably gone. He has to commit everything to memory—both his predecessors’ handling of the same story and the foregoing parts of his own performance. The issue of course involves the question about whether traditional storytellers used promptbooks (diben or jiaoben). Si Su has reported that some modern storytellers in Yangzhou did have a kind of jiaoben for their performance, although they were initially reluctant to admit that to an outsider. A jiaoben, as Si Su tells us, provided prompts mainly for the verses, while the prose part of the story was usually left to the raconteur’s improvisation, a point that might corroborate Zang Maoxun’s hypothesis on the paucity of prose binbai in the Yuan-edition zaju plays.25 Vena Hrdlickova suggests, by citing another Chinese source, that a jiaoben contained only “notes containing the simple skeleton of the story.” The major training of the storytellers had to be, according to Hrdlickova, learning “from their master’s lips.”26 While the use of promptbooks may differ from one oral genre to another and even from one storyteller to another, it would be, in general, quite embarrassing for a storyteller to have to refer to the jiaoben during the course of his performance. In the telling of a long story, a promptbook, no matter what it actually contains, would only alleviate the burden of one’s memory to a limited extent. Reporting on his fieldwork on storytelling in Kunming, Mark Bender informs us that a storyteller “claimed that he needed to review a written episode only once, then tell it, embellishing greatly on the tortuously twisted but often minimal plot line.”27 This seems to be confirmed by Vibeke Børdahl’s study of modern Yangzhou storytelling. Although some storytellers have the ability and interest to work on notations of their performances, a master in the trade “had his repertoire entirely committed to memory and taught only by telling and performing orally.”28 Modern psychology divides human memory into two categories: procedural and propositional memories. While procedural memory refers to the “automatic” memory of skillful procedures (“I remember how to ride a bicycle”), propositional memory “usually requires directed attention,”29 which can be further subdivided into semantic and episodic memories. The former is involved in knowledge independent of a person’s past experience (“The earth is round”),

The Narrative Pattern

83

and the latter, episodic memory, is in charge of mental retention of past happenings and doings. In trying to remember a story that had been told by his predecessors and by himself in the past, what functions in the storyteller’s mind is chiefly episodic memory. Oral storytelling is therefore a process in which considerations of “universal” aesthetic principles such as coherent unity in structure have to give way to a more personal and more urgent need: to tell whatever one remembers in order to keep the narrative going. According to Endel Tulving, the elements of episodic memory are of two types: elements of encoding and elements of retrieval. Encoding refers to the process that converts the information input into a memory trace,30 and retrieval refers to the process of activating the stored information that has been held in abeyance, a process facilitated by responses termed “retrieval cues”—namely, “those aspects of the individual’s physical and cognitive environment that initiate and influence the process of retrieval.”31 In the case of oral storytelling, retrieval cues could reside in the audience’s response as well as the physical surroundings of the performance, but the most important retrieval cues must be those that were built into the narrative itself, those “stimuli” or “memory hooks” that prompt for the storyteller the next unit of the narrative. In applying the psychological theory of memory to the study of oral narration, we can view the formation of the narrative as a series of repeated processes of encoding and retrieval. Each telling of the story is at the same time both a process of encoding and a process of retrieval, encoding for the later storytellers in the same tradition and retrieval of what had been encoded in memory by one’s predecessor’s performance. Since episodic memory is “context-dependent,”32 and context is always mutable, there is inevitably a discrepancy between what has been encoded earlier and what is retrieved later. This helps explain the fact that in oral transmission, verbatim repetition is rare if not impossible. Rather, each telling would tend to “forget” some of the less memorable elements from the previous telling and add something of its own that might be more “retrievable.” It is this selective memory in an oral tradition—selective in the sense of keeping and adding more retrievable narrative units while automatically dropping less retrievable ones—that helps to shape the structural features of an oral narrative, especially the typology of narrative sequence in parallel stretches. Standardized action patterns in narrative sequences are easily retrievable, for in the case of a fixed sequence, what the narrator has to memorize is not all the details in the verbal material but a large and familiar unit, which psychologists call a “chunk.”33 Probably the simplest example of a chunk is a word in an alphabetic language as opposed to its separate constituent letters. A proficient English speaker remembers the word united as a familiar unit and does

84

Chapter 3

not have to memorize all the letters one after another. But a multiword term such as United Nations Educational, Scientific, and Cultural Organization may cause difficulty in retrieving it from memory, even when one feels certain that he “knows” the term. To solve the problem the acronym UNESCO was invented, which helps retrieve that lengthy term in full. Experiments have demonstrated that initial letters probably function as retrieval cues in the recalling of the words.34 Although each word in the term United Nations Educational, Scientific, and Cultural Organization is a familiar unit, or a chunk, one needs the retrieval cues in U-N-E-S-C-O to recall all the constituent words and put them in their respective sequential positions. The situation for the storyteller of the long Shuihu narrative could be similar. While he had a good number of narrative stretches in standardized plot patterns, or chunks, he had still to invent a set of retrieval cues, or “memory hooks,” to help him recall each of those chunks at the right place in the narration. One such retrieval cue may be numerical, for several events in the narrative occur in sets of three. Here are some instances of the triplets: Shi En enters the jail three times to call on the imprisoned Wu Song (chapter 30); Song Jiang makes three attacks on the Zhu Family Village before the Liangshan bandits finally storm the place (chapters 47–50); and Marshal Gao, commanding the government troops in their attack on Liangshan, is defeated three times in succession (chapters 78–80).35 Indeed, the number three appears frequently in many folktales of various cultures, as has been noted by Axel Olrik. Dwelling on what he calls “the law of three,” Olrik considers the number such a hallmark of the great bulk of popular narrative that “When a folklorist comes upon a three he thinks, as does the Swiss who catches sight of his Alps again, ‘Now I am home.’ ”36 Although Olrik stops short of offering any explanation of this phenomenon, it seems that when a certain number is tied to a basically repetitive pattern of action, the number may serve as a prompt to the storyteller as to where he is in the narrative and what follows next. So, the immutable numerical sequence serves precisely the same function as the initial letters in UNESCO in retrieving a combination of chunks, not only in their entirety but also in their proper order. The best-known threefold story in Shuihu zhuan is that of the three attacks on the Zhu Village. If after a “first attack” ( yida) comes inevitably a “second attack” (erda) that is in turn followed by a “third attack” (sanda), the storyteller was not likely to lose his way in the narrative. What is involved here is a most simple methodology for facilitating memory retrieval, and the series of chunks are “numbered” in a way not unlike the numbering of data folders on a modern secretary’s desk. But this numerical methodology, effective as it may have been, could not have played a major role for the Shuihu storyteller’s memory organization, for those threefold events after all account for only a relatively small portion of

The Narrative Pattern

85

the lengthy narrative. If the shorter pieces of Shuihu stories were indeed strung into a longer narrative in the world of popular orality rather than in a writer’s closed studio, we should be able to find in the book indications of some overarching strategy by which the narrator could weave a bunch of otherwise fragmentary narrative units into an overall pattern of action that is easily retrievable, even though it is somewhat meandering.

Linkage Incident as a Mnemonic Device Such indications are not lacking. As is generally known, the first seventy chapters of Shuihu zhuan consist of a number of tales of heroic adventures, each of which features one of the major characters of the narrative who goes through various misfortunes and is finally forced to join the Liangshan rebellion and become one of the chieftains. Since each of these tales is only a part of the entire narrative, let us call these tales “sections.” To a large extent, these sections follow some more or less stereotyped patterns, as illustrated in the foregoing discussion of the narrative sequences and their respective morphological stretches. Each of these sections may be a memory chunk in itself—for the memory can certainly be facilitated by a structural repetitiveness—but each is a relatively selfcontained narrative unit with little or no causal connection with any other. There is no logical rationale that could automatically assign each of the sections into a determinate position in the narrative synthesis. For instance, there is no causal linkage between the section of Song Jiang’s murder of his mistress Yan Poxi and the one of Wu Song’s killing of his sister-in-law. Although the Wu Song section immediately follows the Song Jiang section in the narrative, it is by no means contingent upon the Song Jiang story. In other words, the shift of narrative focus from Song Jiang to Wu Song at this juncture is not determined by any inevitable law of the narrative plot but due to the vagaries of narration. It hinges upon a highly fortuitous incident, which has virtually nothing to do with the main action but merely serves the function of linking the two narrative sections: Song Jiang, self-exiled after killing Yan Poxi and seeking refuge at Chai Jin’s manor, meets Wu Song there and stays with Wu Song for a while before the latter leaves for Qinghe County and embarks on his adventures. Such “linkage incidents” are typical of Shuihu zhuan. When the narrative shifts from a section of hero A to one of hero B, usually A and B will meet and stay together for some time before B replaces A as the focal figure of the narrative. Here are some other scenarios of the “linkage incidents”: 1. Wang Jin and his mother seek lodging at the manor of the Shi family, where Wang Jin meets Shi Jin and for some time teaches the latter martial skills. Wang Jin leaves and the section of Shi Jin starts (chapter 2).

86

Chapter 3

2. Shi Jin inquires of an officer in a tavern about his master Wang Jin. The officer turns out to be Lu Da. They drink together with Li Zhong. Afterward Shi Jin fades out of the narrative, and the focus shifts to Lu Da (chapter 3). 3. Lu Da is demonstrating his military prowess to some vagabonds at the Xiangguo Monastery’s vegetable garden while Lin Chong watches his performance through the gap in the compound wall. The two become friends. The Lu Da section draws to an end and the Lin Chong section begins (chapter 7). 4. Seeking to join the band at Liangshan, Lin Chong is asked to kill a traveler to prove his mettle. He encounters Yang Zhi, and the two fight to a standoff. Lin Chong afterward leaves the spotlight and Yang Zhi comes to the narrative foreground (chapter 12). 5. Wu Song, while drinking in a tavern, abuses the waiter and beats up a fellow customer, who turns out to be Kong Liang. Wu Song gets drunk, falls into a brook and is captured. He is taken to the manor of the Kong family, where Song Jiang, who is there for a visit, recognizes Wu Song and rescues him. This puts an end to the long Wu Song section and leads to another section about Song Jiang (chapter 32). 6. Song Jiang, serving his term of exile in Jiangzhou, drinks with Dai Zong in a tavern, where Dai Zong invites Li Kui to join them. Song Jiang and Li Kui become friends. This suspends the Song Jiang section and leads to an interpolation of Li Kui’s skirmish with Zhang Shun (chapter 38). 7. Dai Zong and Yang Lin are on their way to search for Gongsun Sheng when they see Yang Xiong insulted by some rowdies, who are driven away by Shi Xiu. Dai Zong and Yang Lin invite Shi Xiu to drink in a tavern. When Yang Xiong, who is at that time a police officer, comes to the tavern to look for Shi Xiu, Dai Zong and Yang Lin slip away, leaving the scene to Shi Xiu and Yang Xiong, who become sworn brothers in the tavern. This initiates the Shi Xiu-Yang Xiong section of the narrative (chapter 44). 8. Having triumphantly returned from his attack on the Zhu Family Village, Song Jiang receives a brief visit from Lei Heng, who is passing Liangshanbo on an official mission. The narrative focus shifts afterward to Lei Heng (chapter 51).

I forbear to list more examples. Suffice it to say that the “linkage incidents” are a salient structural feature in Shuihu zhuan. They string up the otherwise

The Narrative Pattern

87

separate narrative chunks into a chain of storytelling, resulting in a “structural pattern of the ‘billiard ball’ shift in narrative focus,” as Andrew Plaks puts it.37 Each linkage incident occurs at a “narrative intersection,” where two neighboring narrative sections join. Since a linkage incident features the protagonists of both of the two neighboring narrative sections, it not only winds up the foregoing section but also channels the storytelling into the direction of the section that follows, like an arrow signal at an intersection. It thus serves as a retrieval cue for the next chunk of the narration. The memory-retrieving function of the link-chain structure can be illustrated by numerous children’s songs that are popular among Chinese kindergartners. In these songs, although each line can be meaningful in itself, the whole piece is usually nonsensical, for there is normally no logic that leads one line to the next. This means that children trying to memorize the song cannot resort to a sense of semantic coherence, even if they are sophisticated enough to have developed that sense. What facilitates children’s memories, however, is a formal device, where the last word of the foregoing line is identical with the first word of the following line, by which every two neighboring lines are “linked up.” One of these songs goes like this: Yi-er-san-si-wu, shangshan da laohu Laohu da budao, laile xiao songshu Songshu laihui pao, wo lai shuyishu Yi-er-san-si-wu, shangshan da laohu One-two-three-four-five, I go up the mountain to hunt tigers Tigers can’t be hunted, there come small squirrels Squirrels are running back and forth, let me count them One-two-three-four-five, I go up the mountain to hunt tigers)

The parallel of the pattern in such children’s songs to the link-chain structure in Shuihu zhuan is a meaningful one. If each line in the song corresponds to a section in Shuihu zhuan that tells the story of a major Liangshan hero, the linkage incidents in the narrative play precisely the same role as the “linkage words” in the song.

Tavern Scene as the Most Typical Linkage Incident Furthermore, the retrieving function of such linkage incidents is tremendously facilitated by their being more often than not stock material in themselves. In the eight linkage incidents listed above, the meeting between the character(s) of the preceding narrative section and the character(s) of the following one is

88

Chapter 3

in most cases based on somewhat formalized patterns. While seeking lodging at a manor (scenario 1) and fighting with an adversary who turns out to be a friend (scenario 4) are both frequently recurring scenes in the narrative, what deserves our particular attention is the scene of drinking in a tavern (scenarios 2, 5, 6, and 7), which is obviously the most typical linkage incident. When a narrative section is about to end, a tavern seems to be the most convenient setting for a shift in narrative focus. Actually, tavern scenes abound in the narrative, especially in the first seventy chapters. They are not only used as a transition from one narrative section to another with a different focal character but often appear within a section as well, when a change of locale, and/or a shake-up of the configuration of minor characters, and/or a major turn of the plot are involved, although the major character(s) may remain the same. In other words, a tavern scene can be not only introduced as a link between two narrative sections but also used as a link between two segments within the same section. In the Lin Chong story, for instance, almost all the major junctures are marked by a tavern scene. Soon after the story starts, when Lu Qian needs to get Lin Chong away from home in order to secure Lin’s wife for Gao Qiu’s adopted son, he takes Lin to drink in a tavern (chapter 7). Also, when Lin Chong is on his way to Cangzhou to serve his sentence of exile, he and the two escorts decide to drink in a tavern. He quarrels with the waiters for not serving them food and wine. The tavern keeper, informing Lin Chong that there is in their village a squire called Chai Jin who is supportive of bold men in exile, tells Lin that it would be ill-advised to go to Chai Jin for help while intoxicated on wine. This tavern scene immediately sends Lin Chong to Chai Jin’s manor and leads to the episode of Lin Chong’s bout with Arms Instructor Hong (chapter 9). After arriving in Cangzhou, Lin Chong meets his old acquaintance Li Xiaoer, now owner of a tavern. Having watched some suspicious customers in his tavern and overheard their dubious talk, Li senses that something is going on against Lin Chong and alerts Lin to the danger. When Lin Chong is put in charge of the army fodder depot, Lin goes that evening to drink in a tavern. Meanwhile, his shack collapses under the weight of the snow. Lin Chong has to put up for the night in a temple nearby, which saves his life, for his enemies set fire to the fodder depot that night. After taking his revenge, Lin Chong is about to escape when he comes to a house—not a tavern but clearly a variant of a tavern—where he abuses some villagers, from whom he confiscates some food and wine. Drunk, Lin Chong is captured by the villagers and taken to a squire who turns out to be Chai Jin (chapter 10). Later, under the guidance of Chai Jin, Lin Chong is on his way to join the outlaws on Liangshan. Again, he goes to drink in a tavern, where he meets the tavern keeper Zhu Gui, who takes him to see the brigand chieftains (chapter 11).

The Narrative Pattern

89

While tavern scenes are almost invariably adopted for the linkage incidents in the Lin Chong story, elsewhere in the narrative the linkage incidents also involve other scenes, but the tavern scene still seems to be the most frequent choice. Although it is neither practicable nor desirable to look into all the tavern scenes one after another, it is probably not redundant to give a few more instances to mark their features. For Yang Zhi, his pivotal transformation from an officer of government troops to a diehard outlaw occurs in a tavern scene. Having lost the birthday convoy, Yang Zhi deserts and enters a tavern, where he makes trouble by refusing to pay for his food and wine. The tavern keeper who fights with Yang Zhi turns out to be Cao Zheng, who befriends Yang Zhi and helps Yang and Lu Zhishen conquer the impregnable lair of Double-Dragon Mountain (chapter 17). In the Song Jiang story there are several tavern scenes, and we have encountered some of them in the foregoing discussion of the narrative sequences. Here is another instance illustrating the way a tavern scene serves as a link. Leading his friends to join the Liangshan bandits, Song Jiang stops to drink in a tavern, where his men clash with one of the customers who has refused to leave the large table. The quarrel is quickly over, for the man turns out to be Shi Yong, another admirer of Song Jiang, who brings a message from Song Jiang’s brother urging Song Jiang to go home for their father’s funeral. This immediately turns Song Jiang away from the route bound for Liangshan to hasten home, where he is caught up in another round of adventures (chapter 35). Whether used as a link between two narrative sections or between two segments within a section, these tavern scenes are highly stereotyped. Characters may meet before they go to drink in a tavern together, or they run into each other in a tavern, or someone makes trouble in a tavern, most often by picking a quarrel with the waiter or another customer, which finally leads to the appearance of another character who is to be involved in the following narrative sequence. In other cases, a tavern scene is introduced only because the turn of events necessitates a change of locale for the protagonist. In the Lin Chong story, for instance, Lin Chong goes to drink in a tavern while his shack collapses under snow, so that he can avoid an untimely death that would have cut the narrative short. But whatever the context and whatever the differences in detail, a tavern scene is in most cases a narrator’s expediency, for it pertains less to the “story,” or the narrative content, than to the act of “narrating,” or the way to tell the story.38 It is in a way like an interlude between two scenes in a drama, a short performance to entertain the audience while the dramatic action is suspended, as it were, to rearrange the stage sets. Similarly, a tavern scene is a device to keep the verbal flow of storytelling uninterrupted while the story

90

Chapter 3

comes to a point where the narrator must make an adjustment of the locale or the configuration of characters. Unlike an interlude that is usually a digression from the dramatic action, however, a tavern scene acts as a bridge joining one narrative unit to another. If linkage incidents facilitate the narrator’s story making, then linkage incidents featuring a highly stereotyped element such as a tavern scene certainly become a double blessing. As is often the case, when the narrative event comes to a turn, the major character enters a tavern where, one way or another, he is to meet someone who will lead the narrative to the next section. Since a tavern scene does not belong so much to the plot of the narrative as to the act of narrating, as mentioned above, the details in a tavern scene are most often interchangeable with those in other tavern scenes. In the scene where Song Jiang meets Shi Yong in a tavern, for example, Shi Yong appears in the narrative as a new figure. He comes to the spotlight as Song Jiang’s men ask him to leave the large table, which leads to a brawl. But the same purpose can be achieved if Song Jiang meets Shi Yong on the way and invites him to drink in a tavern, or if Song Jiang’s men pick a quarrel with a waiter and draw Shi Yong into the clash, or if Song Jiang gets into some kind of a trouble in a tavern and Shi Yong comes to his assistance, as happens in some other tavern scenes. Hence the narrator did not have to memorize all the verbal information about a tavern scene in a fixed relation to the narrative context, for he could easily vary the details and by doing so sustain the flow of storytelling. Song Jiang, in order to keep himself well informed, sets up four taverns around the bandits’ lair as lookout stations (chapter 60). Interestingly, the narrator, who designed this strategy for Song Jiang, used the same strategy for himself as well. Along the lines of the narrative he set up a large number of taverns—many times more than Song Jiang does around the lair—for the narrator, just as Song Jiang, needed these “lookout stations” to keep himself informed of the patterns of story making that might otherwise remain dormant in his memory.

Jin Shengtan on the Recurrent Narrative Patterns As we have seen, the stereotyped sequences and recurrent scenes constitute a salient feature of the narrative structure of Shuihu zhuan. In reviewing the work’s narrative patterns, the Ming-Qing commentators were strongly inclined to look for principles of coherency in terms of careful structural design on the part of the writer. They claimed that everything in the narrative had been properly woven into a structural unity, and thus “the entire work reads like a single sentence.”39 Jin Shengtan, for instance, insisted that the narrative of Shuihu zhuan

The Narrative Pattern

91

was well governed by its principles of organization, whether on the level of chapters or on the lower levels of sentences or words.40 But the recurrence of scenes and action patterns in Shuihu zhuan did not escape Jin Shengtan’s attention. Instead of dismissing them as inadvertent redundancy, he regarded them as the writer’s deliberate violation of the general rule against redundancy, for he thought that the writer needed to take that risk to flaunt his writing skills.41 In his “Du Diwu caizi shu fa,” Jin Shengtan discusses some of the recurrent scenes and action patterns in terms of two writing techniques: the technique of “direct repetition” (zheng fanfa) and the technique of “partial repetition” (luefanfa). Of the technique of “direct repetition,” he enumerates several examples from the narrative: After the account of Wu Song’s fight with the tiger, there are accounts of Li Kui’s killing of four tigers and of the quarrel between the two Xie brothers and the Mao family over a tiger. After the story of Pan Jinlian’s adultery with Ximen Qing, there is that of Pan Qiaoyun with Pei Ruhai. After the rescue of Song Jiang from the execution ground at Jiangzhou, there is that of Lu Junyi in Damingfu. After the account of He Tao’s attempt to capture the robbers, there is that of Huang An’s trying to capture the same robbers. After the story of Lin Chong’s exile, there is that of Lu Junyi’s exile. Also, after the release of Chao Gai by Zhu Tong and Lei Heng, there is the release of Song Jiang by the same pair. This means that the author, for his amusement, deliberately repeats [ fan] the topic [timu] and yet has the ability not to borrow any detail or stroke from the previous treatment. Truly, he is bursting with methods and techniques.42

To illustrate what he calls the technique of “partial repetition,” Jin Shengtan also cites details of the narrative action: Examples are Lin Chong’s purchase of a sword followed by Yang Zhi’s selling of a sword, the parts played by the urchins Tang Niuer and Yunge, the beating of Butcher Zheng in his butcher shop and the beating of “Door God” Jiang in Kuaihuo Forest, and Lu Zhishen’s trying out his monk’s staff at Waguan Monastery and Wu Song’s trying out his monk’s dagger on top of Centipede Ridge.43

In claiming the recurrent patterns of the narrative to be a result of the writer’s deliberate manipulation, Jin Shengtan’s position is hardly tenable. Avoidance of repetition was an inveterate literary convention in China as in the West, but

92

Chapter 3

it was only one of many conventions. Even if the “writer” of Shuihu zhuan was audacious enough to go against the trend, it would be unlikely that he could be so perverse as to violate this one rule against repetition so stubbornly while voluntarily—as Jin Shengtan strongly believed—following other conventions. In the narrative sequences and the tavern scenes discussed here, the recurrences of narrative patterns are so extensive that they are well beyond the reasonable scope of any writer’s experimental move, and therefore they can hardly be explained or explained away in terms of any “writing” technique. On the other hand, if we interpret such recurrences in terms of oral composition and performance, everything seems to fall into its proper place. The narrator was not going against any convention, but was precisely following the convention of oral storytelling, for while recurrence is something to be avoided in literary discourse, it is the very lifeline of an oral narrative.

A Web of Recurrent Narrative Patterns in Early Vernacular Literature Admittedly, recurrent narrative patterns alone are not always reliable indicators of orality, although much more reliable than those isolated formulaic phrases claiming, genuinely or ostensibly, an oral context of communication. It would be less likely for a writer to simulate an oral mode of composition by producing extensive recurrent narrative patterns, but still we cannot rule out such a possibility altogether, especially when oral-derived narratives were looked upon as models by writers of vernacular fiction. Yet when a high density of recurrences in a narrative is compounded with our knowledge of an oral tradition preceding the text, the recurrent patterns become more compelling evidence of an oral provenance of the narrative. Shuihu zhuan, I think, should be considered as such a case. Recurrent patterns can also be found in Jin ping mei, although not quite comparable in degree of density to those in Shuihu zhuan. Andrew Plaks has noted that in Jin ping mei, “the same things seem to happen over and over again in the Ximen household.”44 The recurrences, however, do not necessarily suggest direct oral contributions to the narrative discourse, simply for the reason that there were no recognized oral story cycles preceding the text. Despite many sporadic elements from orality, especially popular verses, the relationship of Jin ping mei to popular orality may have been more mediated.45 In Sanguo yanyi and Xiyou ji, we see extensive recurrences of narrative elements. It is particularly so in Xiyou ji, which demonstrates a significantly high density of recurrences. In addition to the frequent appearances of the same figures or types of figures, the account of the journey to the west is based on the pilgrims’ successive misfortunes and survivals that can all be considered

The Narrative Pattern

93

as variant stretches of a more or less fixed narrative sequence. Both Sanguo yanyi and Xiyou ji belong to the earliest stage of vernacular fiction, each being the culmination of a long oral tradition and a lineage of textual evolution, although the case of Sanguo yanyi is complicated by its incorporation of classical sources in Sanguo zhi and Tong jian gangmu. The duality of its origins in both oral storytelling and classical historiography may have been responsible for its type of linguistic medium being a mixture of baihua with simplified wenyan. In addition to the recurrent narrative patterns in Shuihu zhuan itself, some narrative sequences or scenes in the narrative occur in other works of early vernacular literature as well.46 Earlier in this chapter I discussed the sequence of slaying an adulterous sister-in-law, which serves as the trunk line in two long narrative stretches in Shuihu zhuan. That sequence, as noted in chapter 2, is also the pattern for the dramatic plots in several early Shuihu zaju and even in another Yuan zaju featuring no Shuihu figures. Similarly, what I have called the “exile-imprisonment sequence” in this chapter is actually, according to Patrick Hanan, “one of the hoariest sequences in early fiction.” Hanan points out that the sequence appears in chapter 8 of San Sui ping yao zhuan as well as the story “Zaojiao lin dawang jiaxing” in Jingshi tong yan. Both of these bear “a close similarity, even some verbal similarity” to the Lin Chong and Lu Junyi stretches, and even the two escorts have the same names in each case.47 The two famous escorts in both the Lin Chong and Lu Junyi stories, Dong Chao and Xue Ba, also appear as two of Judge Bao’s yamen runners in the cihua “Bao Shizhi chushen zhuan,” where their insolence to a farmer almost incurs a good beating for them as the humble farmer is none other than Lord Bao himself, the new jinshi.48 In another Judge Bao cihua, “Renzong renmu zhuan,” it is Dong Chao (whose given name is written in a homophonic substitute) and Xue Ba who come upon a destitute old woman in an abandoned kiln who turns out to be the banished empress dowager.49 The ubiquity of the duo does not even stop with the powerful Judge Bao. They make another appearance in the story “Jiantie heshang,” in which they are the constables detaining Huangfu Song’s wife who is wrongly accused of adultery, although their names are given a little twist to become Dong Ba and Xue Chao instead.50 Furthermore, their two fellow policemen in the story are called Zhang Qian and Li Wan, exactly the same names as those of the two escorts in Song Jiang’s exile-imprisonment story. If the shared names of Dong Chao and Xue Ba suggest a shared repository of stuff material between the oral antecedents of Shuihu zhuan and the cluster of Judge Bao cihua, they are certainly not the only evidence. Each of the Judge Bao cihua starts with a segment in verse serving as an introduction

94

Chapter 3

to the story. Despite verbal variations, this introduction in most Judge Bao cihua follows a standardized line. It eulogizes two emperors of the Song—Taizu, founder of the dynasty, and Renzong, ruler of its halcyon days—and then it goes on to extol two officials, Minister Bao Zheng and General Di Qing, as embodying the best civil and military services.51 One does not have to read Shuihu zhuan meticulously to notice that this is precisely the line followed in the yinshou (introduction) of the work, although the yinshou appears mostly in prose. Out of all emperors of the Song, Taizu and Renzong are presented as gods descended from Heaven; and out of all officials during Renzong’s reign, Bao Zheng and Di Qing are chosen to be incarnations of the “Civil Affairs Star” and “Military Affairs Star” respectively, who are said to be responsible for the peace and prosperity of the time. But while such an introduction fits properly in a Judge Bao cihua, it may seem misplaced in Shuihu zhuan, which is neither about Bao Zheng nor about Di Qing. In fact, Bao Zheng’s presence in the narrative does not go beyond the yinshou itself, where he is said to be distributing medicine to try to save people from the plague—but to no avail, adding nothing to the presumed wisdom and sagacity of the Civil Affairs Star. To reasonably explain the misplacement of such an introduction, we may have to consider it as having been a stock move for the raconteurs to start a story set in Renzong’s reign. In a shorter narrative—a cihua, for example—that focused on the plague during the Renzong period, the introduction may have fit well. But once it survived that short piece and was incorporated later into a long narrative that is set not in Renzong’s time but in the Xuanhe period, it became a textual Rip Van Winkle, so to speak.52 If we broaden our vista and look beyond the genre of cihua, we will find that Shuihu zhuan shares much of its stuff material with many other works in early vernacular literature. All the bandit heroes in Shuihu zhuan address each other not by name but as gege (older brother) or xiongdi (younger brother). Indeed, Pearl Buck titled her translation of the novel All Men Are Brothers for good reason. Sworn-brotherhood, which is a major theme of the work, is prevalent virtually everywhere in early vernacular fiction. One of the bestknown embodiments of this theme is of course in Sanguo yanyi, where the relationship between a prince (Liu Bei) and his two generals (Guan Yu and Zhang Fei) is metamorphosed into one between sworn brothers, a relationship to which the audience/readers of the narrative would find it much easier to relate. The bond among the four pilgrims in Xiyou ji, despite its religious veneer of a master-disciples relationship, is in my judgment more secular in nature and may be better characterized as one of sworn brothers. In the cihua text “Hua Guan Suo chushen zhuan,” Guan Suo’s relationship to his sworn brothers completely overshadows his relationship with his true brother, Guan

The Narrative Pattern

95

Ping. This estrangement from one’s biological kin as the cost for sworn-brotherhood may be said to be a corollary of what happened to the older generation in this cihua: To solidify their sworn-brotherhood, Zhang Fei and Guan Yu kill each other’s kinsfolk, with Guan Suo and his mother as the only survivors.53 This reminds one of what happens in chapter 34 of Shuihu zhuan, where Song Jiang has Qin Ming’s family slaughtered by Qin’s superior in order to force Qin Ming to join the band. In the cihua, Guan Suo has twelve brothers, all originally bandits in the Taihang Mountains: Both the number twelve and the locale of the Taihang Mountains are reminiscent of the episode in Xuanhe yishi, where the twelve escorts of the huashi gang (convoy of exotic flowers and rockery) become sworn brothers and subsequently form a band of outlaws in the Taihang Mountains.54 One of the best-known episodes in Shuihu zhuan is Lin Chong buying a precious broadsword. Walking on the street, Lin Chong meets the seller, who is complaining that nobody in the city is able to recognize the exceptional quality of the weapon “handed down from his ancestors.” After a little bargaining, Lin Chong buys the broadsword (chapter 7). Only a few chapters later, Yang Zhi, pressed by financial plight, is on the street selling his precious broadsword “handed down from his ancestors.” An idle bully pesters Yang Zhi about the weapon and provokes Yang into killing him (chapter 12). A similar scene occurs in the Liang shi pinghua section of Xinbian Wudai shi pinghua, in which Liu Wenzheng comes upon a man on the street selling a broadsword. When Liu offers to buy it, the seller refuses to sell it to him, claiming that the weapon is to be sold only to a hero. The insult infuriates Liu, who kills the man with the broadsword.55 The same story recurs in the Zhou shi pinghua section of the same work, simply with Liu’s name replaced by that of Guo Wei, who likewise kills an insolent man who is selling his sword on the street.56 A much longer stretch based on a thematic pattern quite similar to that in the Yang Zhi story is in Sui Tang yanyi, where Qin Qiong, trapped in a situation similar to that of Yang Zhi, has to pawn his maces and sell his horse.57 In chapter 5 of Shuihu zhuan, we see a most hilarious scene. Staying for the night at a manor, Lu Zhishen learns that Grandpa Liu, against his own will, is marrying his daughter to Zhou Tong, a bandit chieftain. Proclaiming that he is able to “dissuade” Zhou Tong from the marriage, the big, fat monk masquerades as the bride, lying naked in the bridal bed. After withstanding many caresses and endearments from Zhou Tong, Lu jumps out of the bed and gives the groom a sound beating. The same basic sequence appears also in chapter 18 of Xiyou ji, where Sun Wukong, staying for the night at a manor, volunteers to subdue Grandpa Gao’s monstrous son-in-law and dissolve the undesired marriage. Turning himself into a simulacrum of the girl, the hairy

96

Chapter 3

monkey awaits the husband in bed before he turns back into his original shape and produces his golden cudgel from the ear. The vanquished monster turns out to be Zhu Bajie, who is to become Sun Wukong’s fellow pilgrim.58 In Zhu Youdun’s play Heixuanfeng zhang yi shucai, the action is also very similar. Police Chief Zhao covets the beauty of Li Piegu’s daughter and forces the marriage with the threat of exorbitant taxation. Hearing of the injustice, Li Kui applies heavy powder on his dark face, throws a veil over his unshapely head, and lets himself be carried in a sedan chair to the police chief ’s home. Needless to say, once unveiled, the pretended bride gives the groom both a scare and a beating.59 All three incidents, equally comic, involve a transsexual masquerade by a grotesque-looking male, beating of the tyrannical groom/ husband in the nuptial chamber, and the subsequent dissolution of the marital bond. Dissolution of a forced marriage is actually one of the most popular thematic patterns in early vernacular literature, inseminating a variety of narrative and dramatic actions, although not always as comic as those just noted. In addition to the incident in chapter 5, two more episodes in Shuihu zhuan, obviously akin to each other, are based on the same pattern: Lu Zhishen’s killing of a priest and a monk who drink with an abducted woman in a monastery (chapter 6) and Wu Song’s killing of a priest who embraces a kidnapped girl in a cemetery temple (chapter 31). Both cases involve a forced sexual relationship rather than a formal marriage. In “Bao Longtu Chenzhou tiaomi ji,” the son of the county magistrate Qin takes a village girl from her father by force, and Judge Bao, pretending to be a traveling scholar, gets the girl reunited with her father by sending the magistrate a warning letter.60 More drastic versions of the pattern are to be seen in two other Judge Bao cihua—“Bao Longtu duan Cao guojiu gong’an zhuan” and “Bao Longtu duan Zhao huangqin Sun Wenyi gong’an zhuan”—where it takes nothing less than capital punishment to annul the forced marriages.61 There is another kind of wedlock formation in early vernacular fiction, equally stereotyped. In chapter 49 of Shuihu zhuan, the woman warrior Ten Feet of Steel (Yizhang Qing) overpowers her adversary Stumpy Tiger Wang, who later becomes her husband. This episode reminds one of Xue Dingshan’s fight with Fan Lihua in Xue Dingshan zhengxi and Yang Zongbao’s bout with Mu Guiying in Yang jia jiang. In all three cases, the pattern is the same: The male general is tamed by the prowess of the female warrior, but the man’s military humiliation is subsequently compensated for by the woman’s sexual submission. This pattern appears in a modified form in the cihua “Zhang Wengui zhuan,” where the young scholar Zhang Wengui, on his way to the capital for the civil service examinations, is kidnapped by a bandit chieftain and

The Narrative Pattern

97

taken to his lair, only to be coerced by the robber’s beautiful daughter into marrying her.62 Another recurring scene in early Chinese vernacular literature is the one of military contest. On Liangshan, Yang Zhi does not stand out as one of exceptional military prowess. After joining the band he seems somehow to have his stature diminished. He does not rank among the Five Tigers of the Main Cavalry and has to settle for the lesser status as one of the Eight Tigers of the Light Cavalry. But it is Yang Zhi who demonstrates his martial skills most impressively when he is a government officer. In a military tourney, Yang Zhi is to contend in archery with Zhou Jin, another officer. Yang lets Zhou Jin first shoot three arrows at him before shooting back. He successfully dodges all three shots and, taking his turn, shoots Zhou Jin in the shoulder with the first arrow (chapter 13). A similar sequence can be found in “Xue Rengui zheng Liao shilue,” where Xue Rengui lets Yuan Hu, a foreign general, shoot three arrows at him before he takes his turn and shoots Yuan Hu to death with the first arrow.63 Later in that military tourney in Shuihu zhuan, Yang Zhi fights Suo Chao, a much more skilled warrior than Zhou Jin. They put on a spectacular show and battle to a draw, much to the delight of Commander Liang and the admiration of the soldiers and local residents, who cheer without end. Again, in Xue Rengui zheng Liao shilue there is a similar scene. Qin Huaiyu and Yuchi Baolin, each the son of a revered but aged general, fight brilliantly in an even battle, dazzling the eyes of Emperor Taizong and the military and civilian spectators.64 The Liangshan bandit heroes’ superior martial skills are often the decisive factor for their victories, as most of their major campaigns are won on horseback. A conspicuous exception is their expedition against the Liao, during which the battles are most often not so much between warriors as between the military counselors. After Song Jiang captures the city of Bazhou from the Liao forces with the strategy of feigned surrender, the war becomes a contest in tricky deployment of troops. Song Jiang has his troops arrayed into a NineUnit Eight-Diagram Formation ( Jiugong bagua zhen), which is immediately recognized by the Liao general, Wuyan Yanshou. Wuyan spreads out his army and challenges Song Jiang to identify his deployment, which Zhu Wu recognizes as a Great Monad and Three-Power Formation (Taiyi sancai zhen). After that, each side deploys its army in a variety of formations, one after another, trying to bewilder the enemy. Each formation is recognized, until Wuyan Yanshou is prodded to enter Song Jiang’s Nine-Unit Eight-Diagram Formation and is captured. However, the older Wuyan, the commander-in-chief of the Liao army, sets up a Zodiac Deployment (Taiyi huntianxiang zhen), which completely perplexes Song Jiang and all his counselors (chapters 87–88). This kind

98

Chapter 3

of warfare may seem bizarre, but it is exactly the way of fighting in Qiguo Chunqiu pinghua.65 Sun Bin and Yue Yi, commanders of the Qi and the Yan armies respectively, contest with different types of military formations before the war escalates into one between two even greater masters of deployment, Sun Bin’s teacher Gui Guzi and Yue Yi’s teacher Huang Boyang. One of the deployments set by Yue Yi is called Ninth-Heaven Mystic Queen Formation ( Jiutian Xuannü zhen), which almost wins the war for the Yan forces.66 Ninth-Heaven Mystic Queen, of course, is an important figure in Shuihu zhuan.67 When Song Jiang is driven to the end of his wits by the enemy’s Zodiac Deployment, it is the goddess who appears in Song Jiang’s dream to teach him a strategy to attack it. In the dream, she also reminds Song Jiang of the three volumes of the Heavenly Book (Tianshu) that she bestowed on him (chapter 88). To make the parallel complete, in Qiguo Chunqiu pinghua, Gui Guzi cannot rescue Sun Bin out of Huang Boyang’s Soul-Deluding Deployment ( Mihun zhen) until he regains his three-volume Nether Book (Yinshu).68 This “warfare of deployments” occurs briefly twice in Sanguo yanyi. In chapter 100, Sima Yi’s Cosmic Unity Formation (Hunyuan yiqi zhen) is recognized and dismissed by Zhuge Liang as too rudimentary, and he in turn sets an Eight-Diagram Deployment (Bagua zhen) and traps Sima Yi’s forces in. Later, in chapter 113, the Shu general Jiang Wei and his Jin opponent Deng Ai contend with each other with different variations of Zhuge Liang’s formation, and Jiang Wei, who was Zhuge Liang’s disciple, emerges the victor of the match. Andrew Plaks, calling our attention to the textual parallels between Shuihu zhuan and three other masterworks of the Ming vernacular fiction, properly terms these relations “a web of intertextual cross-reference.”69 That “web” indeed covers not only the early full-length vernacular narratives but also much of early vernacular literature at large. Patrick Hanan observes in his article on Pingyao zhuan: “If we look closely at the Shuihu in the light of the Ping yao zhuan’s patchwork composition, it begins to appear as a mass of variations on familiar themes.”70 Putting Shuihu zhuan back in the “web” of relations among early vernacular works, we can only feel the critical insight in such observations by Plaks and Hanan. What caused the “variations on familiar themes” in different works in early vernacular literature was not any writer’s deliberate transgression of a literary convention as Jin Shengtan asserted; it would be absurd to suggest that the writer of Shuihu zhuan deliberately duplicated a stretch from Ping yao zhuan, a cihua, a zaju, or a pinghua. To be sure, facilitated by print industry, textual derivation was certainly more likely in late vernacular literature, when the text became the dominant mode of existence and the most accessible way of transmission for a literary work.

The Narrative Pattern

99

Yet in early vernacular fiction, where works typically did not appear in print until after their oral antecedents had been circulating for a long period of time, what was responsible for the parallels in thematic patterns was more likely a shared base for story making, a “common storehouse of convention” that “must have belonged initially to oral, not written, literature,” as Hanan puts it.71 Jin Shengtan’s explanation of the recurrent patterns in Shuihu zhuan in terms of the author’s writing skills should not be taken at face value, although his dwelling on different writing techniques certainly serves his own critical agenda well. On the other hand, Jin Shengtan was sagacious in seeing the recurrences not as a symptom of lack of creativity, as he fully recognized the elasticity and resilience of such patterns. The more or less fixed sequences and the linkage scenes supply the narrator with readymade compositional materials. They are easily adaptable precisely because they are frameworks whose details are easily adjustable. The narrator of Shuihu zhuan’s oral antecedents could flesh them out with variant details to accommodate different contextual situations and different audience expectations. The “variations of familiar themes” facilitate story making but never simply “mechanize” it. The result in the narrative is that the stereotypical nature of the patterns is for the most part balanced by the exuberant diversity of detail. Sequences recur but they are seldom mechanically redundant; the same situations crop up again and again but they seldom seem monotonous. What we have seen in Shuihu zhuan itself and in its relations with other works in early vernacular literature is what Vladimir Propp called the “twofold quality” of folkloric—and thus oral—story making: “It is amazingly multiform, picturesque and colorful, and, to no less a degree, remarkably uniform and recurrent.”72 It also reminds us of what Eric Havelock calls the “echo principle” in the Homeric poems. By “echo,” Havelock refers to the recurrent narrative patterns that facilitate both the singer’s recapitulation of the narrative details and the audience’s anticipation of the narration: Echo is something that the ear of singer and audience is trained to wait for. Its mnemonic usefulness encourages the presence of anticipation. . . . Oral mythos is continually stretched forward in this way as it is told in order to assist recall in the reciter’s mind of how the mythos is to proceed, what the plot is to be. Echo, however, is modified. It is not a duplicate, for a duplicate would say nothing more than had already been said, the tale would degenerate into mindless repetition. The echo must accompany a fresh statement of fresh action, but this cannot be excessively moved or inventive.73

100

Chapter 3

As we have seen, the story making in Shuihu zhuan must have operated on such an “echo principle”: Scenes, incidents, and even long narrative stretches constantly echo each other within the work itself, and many of them also echo their counterparts in other works in early vernacular literature. The poiesis of much of the narrative, it can thus be said, was based on a dynamic dialectic between the etic and the emic, between duplication and invention, and between the uniform and the multiform.

4 From Voice to Text The Orality-Writing Dynamic In the previous chapter, the narrative discourse of Shuihu zhuan is discussed in terms of the oral mode of composition and story making. The discussion, I hope, helps elucidate the fact that much of the narrative discourse indeed took shape in an oral milieu, with many elements characteristic of oral literature intact or discernible in its present textual form. Of course, the voice of the storyteller is gone forever, and it is only in the form of the printed text that the narrative exists today. The current chapter addresses the issue of the textualization of the work. My argument here is that the fanben editions of Shuihu zhuan that appeared in the early sixteenth century were the product of a long and cumulative process of textualization based on a dynamic interplay between writing and orality. I will engage briefly the existing theories on the textual evolution of the narrative, but the focus of the chapter will be on a stylisticlinguistic study of the fanben text. Taking advantage of the strong affinities between Shuihu zhuan and the vernacular stories, both from the “early period” and the “middle period” by Patrick Hanan’s division, I will use those stories as a context for the study of some stylistic and linguistic features in the fanben Shuihu zhuan. The purpose, as we will see, is to show that the fanben text of the novel, just like many other works in early vernacular literature that contain fullfledged written vernacular prose, had traversed a long course of evolution and maturation in which both witing and orality played indispensable roles.

The Two Putative Compilers: Shi Nai’an and Luo Guanzhong The specific manner in which the Shuihu cycles were put into writing and further compiled into a book remains unknown and will probably never be known

101

102

Chapter 4

to us, but in the past four centuries the compilation has been attributed to either or both of two names: Shi Nai’an and Luo Guanzhong. Among Ming scholars, Lang Ying (1487–?) and Gao Ru believed that the book came out as a joint effort, composed by Shi and edited by Luo (Shi Nai’an di ben, Luo Guanzhong bian ci);1 Tian Rucheng (1503–?) and Wang Qi (1565–1614) claimed the sole compilership for Luo; and Hu Yinglin (1551–1602) attributed the work to Shi only.2 Later on, when Jin Shengtan truncated Shuihu zhuan into a seventy-chapter version, he dismissed the latter fifty chapters as Luo Guanzhong’s wretched sequel to the first seventy chapters that he claimed to be Shi Nai’an’s superior work. Unfortunately, little biographical data about either Shi Nai’an or Luo Guanzhong is available. Luo Guanzhong was believed to be a playwright and fiction writer living in the late Yuan or early Ming period, to whom several other narratives, including Sanguo yanyi and San Sui ping yao zhuan, have also been attributed.3 Accounts about Luo Guanzhong in different historical sources are casual and often contradictory. In his Lu gui bu xubian, Jia Zhongming (1343–?) notes that Luo Guanzhong was from Taiyuan and was once his “good friend despite a big gap in age” (wangnian jiao), suggesting Luo might be at least twenty years his senior.4 Tian Rucheng in Xihu youlan zhiyu calls Luo Guanzhong a native of Qiantang (present-day Hangzhou) and, out of line with most other sources, puts his lifetime in the Southern Song period.5 In his Xu Wenxian tongkao, Wang Qi agrees with Tian Rucheng on Luo Guanzhong’s nativity but records his name as Luo Guan instead.6 Hu Yinglin claims that Luo was Shi’s disciple, but he does not substantiate that assertion with any evidence.7 In Jiang Daqi’s preface to Sanguo zhi tongsu yanyi, Luo Guanzhong is said to be from Dongyuan.8 Despite all such contradictions, the historicity of the playwright Luo Guanzhong, based on Jia Zhongming’s account of his personal relationship with him, still seems to some extent credible, but it remains questionable to identify him with the person who was supposedly involved in the compilation of several vernacular narratives. Indeed, some modern scholars have suggested that there might be two different persons with the same name, one being the dramatist living in the closing decades of the Yuan, and the other the fiction writer who compiled Shuihu zhuan during the Ming.9 Should that be the case, our knowledge about Luo Guanzhong, the putative compiler of Shuihu zhuan, would be further diminished, because we would have to split the already insufficient biological data for Luo Guanzhong into two pigeonholes, one for the dramatist and the other for the fiction writer. About Shi Nai’an we know even less. All that Lang Ying, Gao Ru, and Hu Yinglin could say about Shi was that he was a native of Hangzhou, and even on that single point of apparent consensus one cannot be certain, since each

From Voice to Text

103

of them might simply be echoing or citing others without committing himself on the issue.10 Later scholars, such as Zhou Lianggong (1612–1672) of the early Qing period, claim that Shi Nai’an lived during the end of the Yuan, but according to Yan Dunyi that assertion must have been based merely on an inference from Hu Yingling’s unsupported assumption about the masterdisciple relationship between Shi and Luo.11 Because of the scarcity of reliable information of Shi Nai’an’s life, Lu Xun proposed that the name of Shi Nai’an might have been merely an invention by the composers of the fanben recensions of the narrative.12 That, however, did not put the issue to rest. In 1946 and 1947, Shen bao, a newspaper based in Shanghai, published several articles on the discoveries of what was believed to be Shi Nai’an’s memorial tablet in an ancestral temple of the Shi clan in Baiju of Xinghua County, Jiangsu Province; the alleged grave of Shi Nai’an in nearby Dafeng County; and a version of Shi Nai’an’s tombstone inscription written by a Wang Daosheng of the Ming period that was included in the genealogy of the Shi clan of Xinghua.13 In the early 1950s, following the appearance of an article in Wenyi bao that brought up the case again,14 the Association of Literature and Art of Northern Jiangsu and the Ministry of Culture of the Chinese central government respectively dispatched research teams to Xinghua. While the report of the former team tacitly supported the claim that Shi Nai’an was from Xinghua, the report of the latter team was somehow not publicized until Nie Gannu, the chief scholar on the team, declared that all the fuss about Shi Nai’an’s nativity in Xinghua was nothing but sheer nonsense: “Not even a faintest vestige of Shi Nai’an was to be found there.”15 Since 1962, there have been new discoveries in Xinghua and Dafeng. The most significant is the epitaph on an unearthed tombstone for a Shi Tingzuo, which indicates that Shi Tingzuo was the fourth-generation offspring of a Shi Yanduan. This Shi Yanduan, according to the epitaph, moved to Zhejiang during the chaotic years of war at the end of the Yuan before he returned to Xinghua when peace was restored, an account that could be brought in line with the assertion by Ming scholars that Shi Nai’an was a native of Hangzhou. This, along with the collateral evidence of the Shi clan genealogy (Shishi jiapu bu) that indicates Nai’an was another name for Yanduan, enabled a number of scholars to conclude that Shi Yanduan was indeed Shi Nai’an, the compiler of Shuihu zhuan.16 Others, however, have expressed serious reservations on the methodology in synthesizing the various sources of evidence. While the authenticity of the tombstone epitaph as an unearthed historical relic is beyond doubt, they have insisted that there is no way to rule out the possibility that the handwritten copy of the genealogy might be, at least in part, a forgery. Since neither side has been able to convince the other, the attempt to clarify the issue of Shi

104

Chapter 4

Nai’an’s life and nativity is at an impasse.17 More recently, as if to emulate Wu Mei’s identification made in the early twentieth century of Shi Nai’an with the Yuan dramatist Shi Hui,18 some scholars have even suggested the possibility that Shi Nai’an was the same person as Nai An, a writer in late Southern Song or early Yuan who compiled Jing-Kang baishi, a popular history of the ending decades of the Northern Song.19 That identification would put Shi Nai’an’s time about a century back in history. With the situation so confusing, the once enigmatic figure of Shi Nai’an is hardly any less enigmatic to us today. The heated debate on Shi Nai’an and Luo Guanzhong will, it seems to me, never lead to a conclusive end. Even if the historicity of Shi and/or Luo can finally be indisputably established, which I seriously doubt, we will still not be able to determine conclusively to what extent they were responsible for the compilation of Shuihu zhuan or even whether they were involved in the compilation at all. It would be a vain effort, therefore, to base our study of the textualization of Shuihu zhuan on any person’s biographical data, as is often the case with many other narratives with oral and popular origins. As John Foley admonishes us, “the usual notion of context as a set of verifiable authorial facts enabling effective criticism proves largely impertinent to oral and oral-derived works of literature.”20

Early Known Editions of Shuihu zhuan While a prototype Shi-Luo edition of Shuihu zhuan at the end of the Yuan or beginning of the Ming is largely unprovable, what are generally believed to be the earliest known editions did not come into being until considerably later. In 1975, a fragment of an early edition entitled Jingben Zhong yi zhuan was discovered at the Shanghai Municipal Library that is accepted by most experts as an authentic print of the Zhengde (1506–1521)-Jiajing (1522–1566) period and possibly the earliest among all extant editions.21 But it was another edition, also of the Jiajing period, that exerted a great influence on later editions. This was the so-called Wuding edition, which was either sponsored or simply owned by the prominent cultural and political figure of the time, Guo Xun (1475–1542).22 Yet while the historical existence of the Wuding edition is generally accepted, no exemplar of it is now extant. What are generally called “Jiajing Fragments” ( Jiajing canben), which consist of eight chapters and one additional page, were once believed to be part of a copy of the Wuding edition, but that theory has been challenged more recently.23 In the following Wanli period (1573–1620), there appeared several editions, including the one with the title Zhong yi Shuihu zhuan prefaced by the pseudonymous writer Tiandu Waichen, who claimed that the edition was a reprint of the Wuding text.24 After carefully considering the reliability of the dating of Zhongyi Shuihu zhuan, however, Andrew Plaks

From Voice to Text

105

suggests that it was the Rongyutang edition entitled Li Zhuowu xiansheng piping Zhong yi Shuihu zhuan, with a preface dated 1610, that should be considered as “representative of the chain of filiation leading down” from the Wuding text.25 Therefore, the textual evolution of Shuihu zhuan of which we have tangible evidence started relatively late, since no textual exemplar can be safely dated earlier than the Jiajing period. This, however, does not mean that there had been no earlier versions of the novel, either in manuscript or in print. That the Guo Xun edition was based on an earlier prototype was suggested by Tiandu Waichen, who states in his preface that Guo “reprinted the book” (chong ke qi shu) while “cutting off the preambles” (xue qu zhiyu).26 Yuan Wuya (fl. 1614) also suggests in his “Zhong yi Shuihu quanzhuan fafan” that the Wuding edition made some major revisions of earlier texts, including the rearrangement of the episode regarding Song Jiang and Yan Poxi, the deletion of the sections on Wang Qing and Tian Hu, and the addition of a part on the expedition against the state of Liao.27 The details about the textual rearrangement in the Wuding edition are not necessarily reliable, since Yuan Wuya might be advertising his own edition as a true inheritor of the Wuding text. But the “fafan” at least indicates the feasibility of the theory that the Wuding edition had been preceded by earlier texts at Yuan Wuya’s time. Apart from the fragments from Jingben Zhong yi zhuan, the Jiajing-Wanli editions supposedly based on the Wuding model were not only the earliest known editions but also the earliest exemplars of the so-called fanben or full recension.28 Since their appearance, the fanben texts have remained relatively stable. Even Jin Shengtan, with his truncation of the narrative and composition of a new finale, did not go much further than making some minor textual changes in the first seventy chapters.29 On the other hand, the accounts by Tiandu Waichen and Yuan Wuya cited above suggest a radical departure in the Wuding edition from its textual precursors. Thus we see a contrast between a possible state of changes in the textual evolution before the Jiajing editions— although no texts from pre-Jiajing times are extant—and the relative textual stability in the fanben tradition after the Jiajing editions, which had clearly established a textual authority. The inauguration of the fanben recension, marked by the Jiajing texts, may therefore be considered as the culmination of the textualizing process. That the Wuding edition must have been antedated by earlier texts seems indisputable, even if we do not trust the remarks by Lang Ying, Gao Ru, Tian Rucheng, and Hu Yinglin about the Shi and/or Luo compilership. In addition to the accounts by Tiandu Waichen and Yuan Wuya about Guo Xun’s editing over an earlier version, there is textual evidence in the fanben text itself that

106

Chapter 4

points to the involvement of some shuhui (literally, book guilds; associations of men of letters in popular culture) members in the composition of the narrative, a point I will return to later. As a linguistic marker, the character “mei” is used as the suffix to plural personal pronouns in the fragments of Jingben Zhong yi zhuan, while in the Rongyutang edition, the suffix appears as “men” in all but four cases.30 Since the Rongyutang edition was presumably based on the Wuding edition, this could mean either that Jingben came into existence before the Wuding edition or it had textual kinship to a pre-Wuding version. Actually, Guo Xun himself might have disclaimed the originality of the edition of which he was putatively the sponsor. Guo composed in 1531 (the tenth year of the Jiajing reign) a narrative called Guochao Yinglie zhuan, honoring his father Guo Ying. According to Shen Guoyuan, this narrative was written in imitation of Sanguo zhi sushuo and Shuihu zhuan.31 This would suggest that in the early years of the Jiajing reign, a certain version of Shuihu zhuan was already in circulation. Indeed, Lang Ying hypothesized that even Luo Guanzhong’s compilation could have been preceded by earlier scripts or texts ( jiu bi you ben), as he endeavored to explain why he claimed that Shuihu zhuan had been compiled (bian), rather than composed by Luo Guanzhong.32

What Could the Nonextant Pre-fanben Texts Be Like? This brings us to the issue of the relationship of the fanben editions to their possible precursors. Probably we can never know when the earliest effort was made to write the narrative down. Since we are talking about texts that are not extant, there is no firsthand evidence that would enable us to say anything with certainty, but the differences between the fanben editions and the editions in the jianben recension might provide us with some clues. While most of the jianben editions actually contain more narrative events than their fanben counterparts, the language in the jianben editions is rough and uneven and the descriptions brief and sketchy, with a lot of elliptical sentences and omissions of logical links. Lu Xun points out that the language in the Rongyutang edition is so much more vivid and detailed than that in the jianben editions that it is “virtually a new version.” Comparing the passage about the snowy night that Lin Chong spends at the army fodder depot in the 115-chapter edition—which he considers a typical jianben edition—with its counterpart in the Rongyutang text, Lu Xun finds the latter to be more than twice the length of the former.33 In relation to the difference between the two textual traditions, there is also the controversial problem of priority. Which tradition started earlier— the fanben or the jianben? Although most scholars agree that the Jiajing edition must have been preceded by earlier texts,34 they have by no means reached a consensus on the issue of whether the general line of textual evolution of Shuihu

From Voice to Text

107

zhuan went from the fanben to the jianben or the other way around. Some, including Lu Xun, Zheng Zhenduo in his early writings, He Xin, Nie Gannu, and Liu Ts’un-yan, believed that the jianben versions represented an earlier stage of the textual evolution and therefore must be closer to the earliest prototype. Others, represented by Hu Shi, Sun Kaidi, and Richard Irwin, argued that the known jianben printings, which were dated either contemporary with or later than the Jiajing-Wanli fanben editions,35 must be booksellers’ abridged versions of the fanben texts.36 The two sides in the debate, it seems to me, are largely at cross-purposes. The seemingly opposite views are not necessarily irreconcilable, as Andrew Plaks so convincingly points out: This seesaw debate is necessarily inconclusive, since it relies on comparing those extant texts that have by chance come down to us to determine historical priority, when we have every reason to suspect that none of these are in fact original exemplars of their respective systems, and may not even be particularly representative. If, on the other hand, we find that we are dealing with two parallel strains, which only after a certain point enter into a relationship of mutual influence, then there is no contradiction in viewing the jianben line in general as evidence of an earlier stage of development, while still recognizing that the particular jianben examples we have in hand may be the result of a later process of abridgment of the sixteenthcentury fanben texts. Or, to look at it from another direction, we can accept that the fanben must have had earlier prototypes, without falling into the trap of assuming that any of the jianben as we have them must provide models of this “original” form.37

We may well maintain the possibility of some kind of relationship between the jianben editions and the early precursors prior to any fanben text, while not losing sight of the evidence in existing jianben versions that points to adaptations from the fanben tradition. In other words, the extant jianben texts may well have two different sources to draw from: the fanben versions and some pre-Jiajing prototype(s).38 The question now becomes: What could those nonextant early texts be like? With none of the pre-Jiajing texts extant, the possibility of any textual study is completely ruled out, but at least we may consider this against the background of the general development of the written vernacular, especially that of vernacular prose. As discussed in chapter 1, an overwhelming majority of the vernacular works, especially those with mature and expansive prose, ap-

108

Chapter 4

pear only in the Ming—in most cases late Ming—editions, even if they had been associated with an oral tradition of earlier times. This fact points to a course of gradual maturation of written vernacular prose. Since vernacular prose as a literary language did not grow full-fledged—both quantitatively in terms of the number of the works in it and qualitatively in terms of the degree of vernacularity—until perhaps the middle of the Ming, its development could have been roughly concurrent with the textualization of Shuihu zhuan. The corollary is obvious. If the Wuding edition was preceded by a series of earlier texts, the prose in the earliest prototypes must have been much less vernacular and much less amplified, possibly mixed with simple wenyan. This certainly agrees with the general course of development for written vernacular prose mapped out in chapter 1. To avoid possible confusion with the jianben editions that came into existence in or after the Wanli period, it would be better for us not to call those early prototypes “jianben” texts, but it seems certain that they must have been much shorter versions with less expanded vernacular prose. C. T. Hsia has insightfully suggested the possibility of a great disparity in language style between the fanben tradition and its precursors. He gives credit to the Wuding edition for its “distinctive contribution” of “a decidedly colloquial style to replace what . . . must have been a style more literary and economical.”39 The contrast between these two language styles—one being more literary and more sketchy and the other distinctly more colloquial and amplified—certainly reminds us of the difference between the dearth of vernacular prose in Yuan-edition zaju and the early nanxi texts and the significant expansion and augmentation of vernacular prose in their late Ming versions. Given the gradual process of vernacularization of prose from the Yuan to the middle of the Ming, it seems a reasonable hypothesis that the earliest Shuihu texts could be very sketchy outlines, initially serving as aids for oral performance. Most likely they did not appear in print but were manuscripts circulating among the storytellers themselves. While these sketches, at a certain point of their development, were to become the bedrock for the later jianben tradition,40 they continued to be expanded and amplified, becoming increasingly detailed in the narrative action and increasingly vernacular in the language style. Elements might be drawn, at first tentatively, from actual oral delivery to reform and enrich the sketchy written versions that might have previously existed, and the new written version in turn could be read by people in the storytelling circle, or used as a guide for another round of oral telling, which again would lead to a still higher level of amplification. Those conjectural transitional texts could probably not be read by the general reading public for at least two reasons. First, they could be sketchy and elliptical and loaded with jargon and homophonic “errors.” Therefore, the texts might not be eas-

From Voice to Text

109

ily intelligible to an outsider. Secondly, public performers usually wanted to keep secret their ways of transmitting the tradition, for what was at stake was nothing less than their livelihood itself. The textualization thus did not have to be subsequent to the synthesis of the Shuihu stories from different popular genres into a unified oral narrative in prose. Rather, they were more likely two synchronous processes, interdependent and mutually facilitating and reinforcing. The gradual and cumulative textualization brought the written version farther and farther away from the original sketchy outlines and finally arrived at a much expanded written narrative, which, after editing and compilation, was ready to appear in print for the reading public. It is impossible for us to chronologize accurately the course of progress in the textualization, but Zhu Youdun’s two Shuihu plays published in 1433 may serve as a milepost. In the original versions of those two plays, the plenum of the bandit heroes is referred to as the “Group of the Thirty-Six,” aligning with Xuanhe yishi rather than Shuihu zhuan.41 It seems that Zhu Youdun at that time was unaware of a key number that was to crop up frequently in Shuihu zhuan, the number of 108, including both the thirty-six major chieftains and the seventy-two minor ones. This may mean, as George Hayden and Y. W. Ma have pointed out respectively in their illuminating articles, that no Shuihu narrative with a full account of all 108 bandits had appeared in print by 1433.42 Although no exemplar of the early textual precursors is extant, what is described in “Xiaochuang ziji,” written by Wu Congxian in the early seventeenth century, could provide us with a missing link.43 Wu claimed that he had read a version of Shuihu zhuan that appeared in a language “unsmoothed and unrefined, peculiar and irregular” (Qi ci zhazha buya, guaigui bujing ).44 Based on Wu’s quotations from the book, the language was basically simplified wenyan, and its narrative details did not completely tally with those in either the fanben or the jianben editions. There is no way for us to know exactly how sketchy the narrative was or how limited its scope, but judging from Wu Congxian’s description, it may have belonged to a very early stage in the transition from the oral to the written. Technically, of course, it was possible for some postJiajing writer to rewrite the narrative in the classical language, but that was extremely unlikely given the general trend in the narrative literature shifting from wenyan to baihua.

The Reciprocal Oral-Literary Transmission Therefore, in the process that ultimately led to the fanben editions, two movements could have been involved that—although running in opposite directions— were mutually dependent. A written version at a certain point of the evolu-

110

Chapter 4

tionary process was the destination of a textualizing movement as well as the starting point for another oralizing movement. It was both prior and posterior to oral delivery, and for that reason it could be both compositional and notational, helping to shape and also being shaped by the oral telling. After the Jiajing fanben exemplars, the oral storytelling of the Liangshan cycles no longer had any role to play in the textual evolution of Shuihu zhuan; but before their advent the transmission of the narrative could be characterized as a process of “oral-literary” reciprocation, for each of the foregoing scripts or texts could mark a link in a gradual process of interaction between writing and orality. Steadily and constantly, the storytellers’ voices fed the growth of the narrative text, resulting ultimately in developed and amplified vernacular prose. A recent study has confirmed that the oral-literary reciprocation may have been the way of textualization shared in many genres of popular performance in premodern China, which was known in the northern storytellers’ jargon as zuannong (compilation/adaptation/reorganization). “The performers’ zuannong could serve as a basis for literary writers’ composition of huaben xiaoshuo; such huaben xiaoshuo in turn were to be recycled in storytellers’ oral performance; and then the newer round of oral performance would lead to the composition of another written version of the huaben xiaoshuo. This constant cyclic movement between the performers and writers constituted the ‘chain of literary creation’ ” (chuangzuo lian).45 What happened to some of the characters’ names in the textual evolution from Xuanhe yishi to the fanben version of Shuihu zhuan may serve as an illustration of the reciprocal oral-literary transmission. Table 4 shows the changes in some of the names: Each person’s name appears in the two works in obviously different orthographical forms, but it remains pronounced identically or similarly. Hearing a name uttered by the storyteller, the scribe may have lost the clue to the “right” characters that were used for it. The reason is simple. A Chinese personal name or even a nickname—unless it is in itself a fixed and familiar semantic unit, as in the case of such an abused name as Fugui, which literally means “wealth and rank”—does not provide a clear-cut semantic message that would immediately and unequivocally dictate the proper use of characters. Accordingly, until the written form of the name became well established, what the scribe could do with each syllable was note down the sound as he heard it, adopting a character possibly different from the one previously used. Even in other cases where a semantic context could be more easily delineated, the scribe would not always be meticulous in selecting the “right” written character. A prefatory piece attributed to Huai Lin, a monk friend of Li Zhi, claims that the latter considered the numerous “misused” characters (ezi jiduo) in the narrative the very hallmark of a text of a popular origin. In the

From Voice to Text

111

Table 4 Comparison of some characters’ names in Xuanhe yishi and the Rongyutang Shuihu zhuan Xuanhe yishi

Rongyutang Shuihu zhuan

_______________________________________________________________________ Li Jinyi Wang Xiong Zhang Qing Mu Heng Huyan Chuo Du Qian Zhang Cen

ıiq ˝z iC pÓ IµÔ ˘d i¬

Lu Junyi Yang Xiong Zhang Qing Mu Hong Huyan Zhuo Du Qian Zhang Heng

cTq ®z iM p∞ Iµ` ˘E iÓ

Rongyutang edition that contains Huailin’s preface, many of the mixed-up usages such as dai a for dai N and de o for de ∫ are left unchanged.46 It was in the Yuan Wuya edition, which appeared a few years later, that such “errors” were corrected. In his “Zhong yi Shuihu quanzhuan fafan,” Yuan Wuya complained that in his edition an onerous job had to be undertaken to correct the “miswritten” characters from the previous versions: In regard to the correction of miswritten characters, a lot was done in the old editions, yet remaining errors are still numerous. In the lines of the introductory verse, for instance, there are four mistakes. A comparison between the present edition and the old one will tell all the others. The uses of nai @ for nai ` and zao º for zao Í may well have merely been slips of the pen, but the falsifying of dai π into dai a, sha Ÿ into sha ˛, and shuan C into shuan ¨, and the confusion between chong R and chong ƒ and between jing w and jing ∫ lead the meaning astray. Instances like these are too numerous to mention here, and they have all been corrected in the present edition.47

Except for the case of zao º and zao Í, which might be a phonetic mix-up as well as an orthographical one, characters in each of all other pairs do not bear any graphical resemblance. But while the characters look different, they sound identical, for each pair is an exact homophone. Such homophonic substitutes would suggest that the scribe did not care so much about orthography as about phonetics. Most likely the original scribe who was responsible for such errors was noting down what he heard; and since what he had put down in writing was to be recycled into another round of oral delivery, it was the representation of the sound that would really matter. Functionally, therefore, homophones

112

Chapter 4

were perfectly interchangeable, and it was practical for the scribe to substitute a character with a homophone that was either more commonly used or had fewer strokes. Such homophones would not “lead the meaning astray,” for the “wrong” word would sound exactly right to the audience. Indeed, they did not become errors until the text-reader communication model for the narrative finally broke away from the oral milieu. The fanben recension marks the establishment of such a model of “pure” reading, where a word’s visual and aural attributes become two mutually conditioning factors. In his study of the bianwen texts, Victor Mair proposes that the texts may “represent various points of development on a continuum ranging from oral to written,” which means that the transition from oral performance to written vernacular text—at the incipient stage of vernacular literature—could only be a gradual process. Similar to what we have seen in the textualizing process of Shuihu zhuan, the bianwen texts are also loaded with homophonic “errors,” which could be a symptom of the “oral-literary” transmission during the evolution from voice to text: “The frequent occurrence of homophonic error is an indication (although by no means proof ) that the individuals who wrote down the transformation texts were more strongly influenced by oral renditions of stories than by written ones. Countless examples could be cited to indicate that the sound rather than the shape of a given character was usually uppermost in the mind of the scribe.”48 As both translator and researcher of the Yuan drama, Stephen West has also noticed the extensive homophonic substitutes in the texts of zaju. The written forms for some lexical items are unstable in the texts, because “they reflect a language that hinges on oral-aural keys—that of spoken literature.” Consequently, a word like zhiwei (just because) is represented by a variety of combinations of characters: u∞ (zhiwei), l∞ (ziwei), h∞ (zewei), ≠∞ (zhiwei), and Ó∞ (zhiwei), all having similar or identical pronunciations.49 Another such example is the texts in Ming Chenghua shuochang cihua congkan. In “Shi-lang fuma zhuan,”50 for instance, such homophonic “errors” occur virtually every few lines, using gang Ë for gang ı and gang ˚, guojiu Íh for guojiu ͧ, tongbao P] for tongbao PM, zhufu ¨I for zhufu ÒJ, zimei lf for zimei nf, shoupa ‚» for shoupa ‚¨ dandan ¶¶ for dandan ÊÊ, and so on. The once rampant homophonic “errors” in the Shuihu texts could mean two things for us. First, since those “errors” were most prone to occur when the scribe’s major concern was the rendition of the sound, they may have been inherited from earlier texts that had arisen immediately from an oral tradition. Second, as those many erroneous characters were not corrected until just before the fanben recension—and some of them even survived in the Rongyutang edition—it may suggest a long process of textual evolution in proximity

From Voice to Text

113

to an oral milieu, which explains the sustained tolerance of such homophonic substitutes. The fanben editions of the narrative, therefore, were the result of such a process of constant dynamic between orality and writing. Precisely like the “textual reprocessing” of the oral tradition about Jesus that Walter Ong considers in his study of the Gospels, this is a process in which “Oral materials are textualized, the textual materials then freely circulated orally, with or without some textual control conjoined to oral control, and then are reprocessed from orality into text again.”51 This orality-writing dynamic gradually reduced the disparities between writing, which had for centuries been alienated from the dayto-day spoken words, and the speech on the storyteller’s tongue. A “more literary and economical” language style of the earlier stage of the textual evolution, a style that was obviously closer to classical Chinese, was replaced in the later stage by a style that is “decidedly colloquial.” Out of this oralitywriting interaction emerged the mature vernacular prose, which was to make possible a brand-new narrative mode for later fiction writers. Indeed, textualization through such reciprocal “oral-literary” transmissions was not unique to Shuihu zhuan but was an inevitable stage in the course of gradual maturation of vernacular prose in general. It explains why no zaju, xiwen, pinghua, or huaben texts with amplified and mature vernacular prose have been found to date concurrently with the heydays of the oral genres with which the texts may have been associated. Take zaju for example. As noted earlier, the Yuan-edition zaju texts contain little or no vernacular prose. In comparison with the Ming-edition zaju plays, the Yuan textual precursors can only be considered as a kind of jianben, as Gu Xuejie suggests.52 Most of these Ming dramatic texts—both the printed editions like Zang Maoxun’s Yuanqu xuan and the manuscripts such as those included in the Maiwangguan Library and later inherited by the Yeshiyuan Library—were based on the so-called inner-palace scripts (neifuben), which were the texts used in rehearsal by the actors and actresses hired by the imperial court.53 This means that the textual evolution of the zaju plays was not based on any writers’ rewriting in their closed studios. Rather, it must have proceeded in close proximity to the tradition of stage performance itself. A script at a certain point of the textual evolution could be an effort both to register the lines onstage and to regulate future performances, thus marking a certain point in the continuum of transition from the oral to the written. The vernacular prose in xiwen underwent a similar process of amplification, as shown earlier in the discussion of the plays in the Yongle dadian and those in late Ming editions. But the best example from the genre to illustrate “oral-literary” transmissions is provided by a more obscure play, Liu Xibi Jinchai ji.54 A manuscript of the play was excavated from a Ming tomb in 1975,

114

Chapter 4

and on the last page, two lines of characters were written that indicate the manuscript was completed between the sixth and seventh years of the Xuande reign (1431 and 1432).55 This manuscript was found to have a cover that was made of several sheets of paper pasted together. As Chen Liming informs us, the sheets of paper used for the cover had actually been part of an earlier manuscript of the play that the Xuande manuscript had evidently replaced. According to Chen Liming, these fragments of the earlier script were considerably more sketchy and stunted in both the action and the vernacular prose than their counterparts in the later version.56 Obviously, the Xuande manuscript represents a major revision and expansion over its precursor. The newer version, probably incorporating elements from actual stage performances, was still meant to be a stage script, as suggested by the numerous homophonic substitutes and the different marks in red ink as signs for stage directions. The people who were engaged in the revision were men of letters in close association with the actors and actresses, as the last page of the manuscript carries the line “Zaishengsi liyuan zhili,” which means, literally, “This has been established by the Troupe of Zaishengsi.”57 We have adequate evidence for the gradual amplification of vernacular prose in the texts from different genres of early vernacular literature. It is significant that almost all works in early vernacular literature either were associated with an oral tradition or belonged to a performative genre. We can now say with some assurance that the prose in early vernacular literature was written and then repeatedly rewritten in proximity to some living oral models. Although in most cases we can only speculate on the specific modes of the orality-writing interaction, it should be accepted as a fact that written vernacular prose had to traverse a course of gradual maturation and amplification based on its contact with orality before it became a full-fledged literary language. Since the gradual augmentation and expansion of vernacular prose was by no means unique to any one work or any one genre but a common phenomenon in early vernacular literature at large, we have to take that as a corollary of the general linguistic climate and the society’s slowly developing awareness of the new possibilities in creative writing. The vernacularization process could only advance at such a pace as the general cultural conditions would allow, and by those cultural conditions the textualization of Shuihu zhuan, just like the textual evolution of so many other works in early vernacular literature, had to be governed.

The Kinship of Shuihu zhuan with Early Vernacular Stories The “oral-literary” transmissions were responsible for the extensive parallels in thematic patterns between Shuihu zhuan and many other texts in early ver-

From Voice to Text

115

nacular literature, such as those discussed in chapter 3. These parallels may be better explained in terms of a shared pool of story-making material, which the works in early vernacular literature both contributed to and drew from. In chapter 3, I deliberately refrained from a closer examination of the early vernacular stories, where thematic parallels to Shuihu zhuan are particularly opulent. In addition to the “exile-imprisonment sequence” found in Shuihu zhuan and Ping yao zhuan, as well as the story “Zaojiaolin dawang jiaxing,” which was mentioned earlier, Patrick Hanan has noted several other parallels. The scene in both chapters 17 and 27 of Shuihu zhuan, where the hero (Yang Zhi and Wu Song respectively) barely escapes becoming chopped into filling for buns, and the scene in chapter 16 where the robbers carry away the loot while the guards are drugged and paralyzed, are both found in the story “Song Sigong da’nao ‘Jinhun’ Zhang.” Two scenes in the novel, the hero’s (Yan Qing) winning of a wrestling tournament on Mount Tai (chapter 74) and the hero’s (Lu Junyi) departure from home to avoid a prophesied disaster (chapter 61), both occur also in the story “Yang Wen ‘Lanluhu’ zhuan.” Additionally, just like Shuihu zhuan, both “Yang Wen ‘Lanluhu’ zhuan” and “Wan Xiuniang choubao Shantinger” are about bandits and are both set in the area of Shandong.58 Several other parallels are also worth our attention. In “Ren xiaozi liexing wei shen,” the husband’s bloody killing of his adulterous wife and her lover in order to vindicate his wronged father strongly resembles the narrative sequence of “killing an adulterous sister-in-law,” of which we have alreadly seen parallels in several Yuan zaju as well. The scene of the hero’s ferocious massacre of the woman and her family in the story is almost a recurrence of Wu Song’s wild killing in the Duck and Drake Bower (chapter 31); and the ending of the story, where the hero, escorted by a group of sympathetic neighbors, gives himself up to the magistrate, reminds one of the similar scene of Wu Song’s surrender (end of chapter 26 and beginning of chapter 27). The scene in which the hero, chased by an enemy, ends up seeking lodging at the enemy’s own home occurs in Shuihu zhuan (chapter 37) as well as in “Cui Yanei baiyao zhaoyao.” Another scene, in which a chief detective complains to his subordinates about their lack of resources in handling a difficult case, is found both in the novel (chapter 17) and in “Kan pixue danzheng Erlangshen.” In the last case, there are even obvious verbal parallels. In both Shuihu zhuan and “Kan pixue,” the police chief sullenly reminds his inept aides that they are in the same room where they have received their pay: [Shuihu zhuan] In ordinary times you were all in this room to get your money to spend. (Nimen xianchang shi dou zai zhe fangli zhuanqian shiyong.) (1: 519)

116

Chapter 4

[“Kan pixue”] You people all used to be in this room to get your money to spend. (Nimen zhongren dou zai zhe fangli zhuan guo qian lai shi de.)59

Also, in both occurrences of the scene the chief’s annoyance is said to grow to a new height after his conversation with his assistants: [Shuihu zhuan] [He] previously had only three-tenths of vexation . . . and now it has added five-tenths more of vexation. ( Dangchu zhiyou sanfen fannao . . . you tianle wufen fannao.) (1: 519) [“Kan pixue”] [He] previously had only five-tenths of vexation . . . and now it has added ten-tenths more of vexation. ( Xianqian zhiyou wufen fannao . . . gen tianshang shifen fannao.)60

Such verbal correspondences to Shuihu zhuan are by no means rare in the stories. In “Ren xiaozi liexing wei shen,” we find expressions, especially in the characters’ dialogues, that also appear in Shuihu zhuan. One such example is “Shuo qilai, zhuang niniang de huangzi” (If I speak out it will be to your disgrace), an expression used by the adulterous wife when she falsely accuses her father-in-law, in front of her husband, of having sexually assaulted her.61 The expression, in slightly different forms, occurs in the Shuihu episodes of Wu Song and Shi Xiu, both of which parallel the story “Ren xiaozi” thematically as well. As Wu Song moves out of his brother’s house, he refuses to tell his brother the true reason by saying, “Shuo qilai, zhuang nide huangzi” (chapter 24; 2: 743). In the Shi Xiu episode, Pan Qiaoyun uses the expression “zhuang nide wangzi” while deceitfully complaining to her husband, Yang Xiong, about Shi Xiu’s “advances” (chapter 45; 3: 1505). It is interesting to note that the word “huangzi/wangzi” takes three different written forms in three places, which suggests that the highly colloquial expression was not yet set orthographically. Again, such homophonic variations may confirm that these parallels are more likely to have taken place amid the orality-writing dynamics, rather than by way of textual derivation removed from the oral model. The following is another example in which the formulaic address of the narrator in Shuihu zhuan to the reader/audience foretelling a character’s impending disaster finds a number of verbal parallels among the early stories: [Shuihu zhuan] If it were I, the teller of this story, living in those days, I would have grabbed him around the waist and pulled him back by force! (Chapter 32; 2: 1038)

From Voice to Text

117

[“San xianshen Bao Longtu duanyuan”] If it were I, the teller of this story, living in those days, I would have grabbed him around the waist and pulled him back by force!62 [“Shiwu guan xiyan cheng qiaohuo”] If it were I, the teller of this story, living in those days, I would have grabbed him around the waist and pulled him back by force, and he probably wouldn’t have had that disaster!63 [“Ren xiaozi liexing wei shen”] If only someone at that time would grab him and not let him go!64

It is important to note that such verbal parallels are not among those “storyteller’s clichés” found virtually everywhere in Ming-Qing vernacular fiction, both full-length narratives and short stories. Instead, they are much less frequently seen, and their occurrences may well be limited only to the early stories and Shuihu zhuan. Both the thematic parallels and verbal correspondences suggest a strong kinship of Shuihu zhuan with the early short stories, a kinship that distinguishes the novel from the other three masterworks in Ming fulllength vernacular fiction. The evolution of Shuihu zhuan, as discussed earlier, had to be a process of prosification of material from different dramatic and chantefablic sources and of synthesis of shorter pieces into a narrative of formidable length. This means that many parts of the narrative at a certain stage of their evolution may have not only paralleled some of the early stories temporally but also had a similar mode of existence. As we recall, the earliest known—although nonextant—Shuihu narratives, the stories of Yang Zhi, Sun Li, Wu Song, and Lu Zhishen, belonged to different subgenres of the storytelling form of xiaoshuo, as did the source material in some of the early stories.65 Indeed, we have to agree with Patrick Hanan that those early stories “come from the same realm of fiction that ultimately produced the Shuihu zhuan.” For that reason, “these early heroic stories are the best context in which to see the rise of the Shuihu zhuan cycle, even better than the pinghua.”66 Of those stories that may have belonged to “the same realm of fiction that ultimately produced the Shuihu zhuan,” only a relatively small number are extant. Most of them, like those within the Shuihu cycle itself, were lost, either before or after they became textualized. In Zuiweng tanlu, sixteen titles, including “Shitou Sun Li,” are identified as gong’an stories; eleven, including “Qingmian Shou,” are listed as podao stories; and another eleven, including “Hua Heshang” and “Wu Xingzhe,” are classified as ganbang stories.67 These numbers indicate that there were dozens of stories belonging to the same realm

118

Chapter 4

of fiction as the earliest Shuihu stories. And we should keep in mind that even those titles in Zuiweng tanlu are listed merely as examples; the actual numbers of the tales from the Shuihu cycle and those in each subgenre of xiaoshuo at the time could have been much larger. This is to say that the relationship between Shuihu zhuan and the early vernacular stories may have been even more extensive than our textual study of the extant stories would suggest.

Stylistic Features in Shuihu zhuan Examined in the Context of the Vernacular Stories From a linguistic and stylistic perspective, we can even see the evolution of Shuihu zhuan in a larger context, one not limited to those stories that share the same thematic bedrock with part of the Shuihu tradition but including all early stories collectively. To be sure, the language in the stories is by no means consistent. It hovers in a wide spectrum between simplified wenyan, as in the descriptive set passages in parallel prose, and most colloquial expressions, often exemplified in the direct speech of the characters. Yet, generally speaking, the plain narrative portions of the stories from the same historical period—those having similar communication contexts where neither the narrator nor the audience has individual stylistic features—largely share the same style, which Patrick Hanan terms “vernacular fiction style.”68 While language employed in the stories may vary and so result in different stylistic effects, the markers of the vernacular fiction style, a set of typical context-bound expressions such as ones introducing lines in verse and ones that lead into a character’s inner speech, are found to be generally much less susceptible to change. Through meticulous analyses and sorting of the markers of the vernacular fiction style, Patrick Hanan has been able to divide the 149 stories into three groups: the early group, dating from a period before the Yuan to around 1450; the middle group, from around 1400 to around 1575; and the late group, from around 1550 to the 1620s. Among the criteria Hanan uses in separating the three groups from each other are three sets of markers of the vernacular fiction style, which are mutually exclusive except in the cases of inordinate numerical preponderance. To establish such stylistic markers for each period, Hanan first sets a “pilot group.” For the middle period, the pilot group is established by internal evidence, such as time setting, place names, and institutional terms that unequivocally point to a Ming dating. For the early period, the pilot group is constituted of stories that conform closely to the stylistic patterns of the pinghua texts and are distinct from those in the middle group. Their “early” dating is also suggested by the fact that they represent the subject matter of oral tales listed in Zuiweng tanlu. From each pilot group, markers are selected from the stylistic features of the “vernacular fiction style,” which are in

From Voice to Text

119

turn tested and reassessed before they are applied as dating criteria for other stories.69 If Shuihu zhuan evolved in proximity to the short stories, much more so than any of the other early full-length vernacular narratives, the linguistic and stylistic features of the short stories may present a very useful perspective for our discussion of the textualization of Shuihu zhuan. From the earliest Shuihu stories in the Southern Song as listed in Zuiweng tanlu to the fanben editions of Shuihu zhuan in the Jiajing period, there was a period of two or three centuries, straddling the early and middle periods for the vernacular stories by Hanan’s division. Although no texts of Shuihu narratives between Xuanhe yishi and the fanben editions of Shuihu zhuan are extant, the close kinship of the Shuihu cycles with the vernacular stories carries tremendous implications for the nature of the narrative discourse in Shuihu zhuan. If the narrative, even though published considerably later, had reached its present form in the late Yuan or the early Ming,70 we should expect the fanben texts to feature exclusively markers of the vernacular fiction style shared by the stories from the early period. If the novel as we have it today was composed at a rather late date— say, in the late fifteenth or early sixteenth century—the text should contain style markers only in common with the stories from the middle period. In that case, the absence of the style markers of the early period would indicate that the narrative was written de novo, inheriting from the early Shuihu sources only their subject matter but not their discourse. However, if the evolution of Shuihu zhuan was a prolonged and cumulative process of prosification and synthesis of material from different oral and performative genres, as I have been arguing, the making of the narrative discourse may have overlapped with both the early and middle periods of the vernacular stories. In that case the plain narrative in the work should show a mixture of markers of vernacular fiction style of both the early and middle periods, rather than a preponderance on either side throughout the entire work. Of course, it would be unreasonable to expect each of Hanan’s criteria to be absolutely infallible. While we see the kindred relationship of Shuihu zhuan with the vernacular stories, we also have to remember that this relationship may not have been exclusive in the narrative’s long course of evolution. We cannot take for granted that the criteria for the stories can all be safely transplanted to the study of the novel. Yet since the purpose here is not exact dating but a rough delimitation of a long period of textualization, these stylistic markers should be generally useful without having to be specifically accurate. It is especially so because the text of the novel provides a large and therefore reliable statistic base. Listed below are some of the stylistic markers for the vernacular stories

120

Chapter 4

that also occur in the plain narrative in the Rongyutang edition of Shuihu zhuan. The markers of the early period are represented by capital letters and the markers of the middle period by small letters: A. changnuodao C. quedai E. buduoshi G. da yi kan shi I. daoshi a. zhengzhi ø» c. jiang ji e. jiubian71 g. dadao

B. jican kejin, yezhu xiaoxing D. kanshi F. shuo you wei liao °SºF H. danjian J. daode b. zheng yao d. zhijiao Ω– f. X pai shifen h. youfenjiao ≥¿–

Table 5 shows the incidence of these stylistic markers in each chapter of the Rongyutang edition, with the frequency of each marker’s occurences indicated by a number:72 A number of things can be said of Table 5. As we can see, most of the stylistic markers in the stories occur in Shuihu zhuan as well, further testifying to a kindred relationship. The stylistic features of both early and middle periods, which are mutually exclusive in the early period stories and the middle period ones, coexist in almost every chapter of the novel. The only exceptions are chapters 31, 66, 76, and 100, where no stylistic features of the middle period are found, and chapter 89, where no markers of the early period are seen. In most chapters, the occurrences of the markers of the early period outnumber those of the markers of the middle period, largely due to the frequent uses of two of the early features, kanshi and danjian. Apart from these two, however, the occurrences of both groups of stylistic features are more or less balanced throughout the entire novel. This fact alone, to be sure, does not definitively prove anything, but it is consistent with the hypothesis that the fanben text of the narrative may have resulted from a long process of textualization, stretching from the early period to the middle period for the vernacular stories. Some of the stylistic features, such as buduoshi, kanshi, danjian, and dadao, appear invariably in “correct” characters, as the proper form of writing is immediately and unequivocally dictated by the meaning of the word. Others, however, appear alternately in the “correct” form of writing and in homophonic variants: quedai is sometimes written as qiadai;73 zhengzhi ø» as zhengzhi øΩ; zhijiao Ω– as zhijiao u–; and youfenjiao ≥¿– as youfenjiao ≥¿Ê.74 The one feature that is by far the most mutable in writing is shuo you wei liao °SºF. It appears in several variants, of which the most frequent forms are the exact

Table 5 Incidence of vernacular story stylistic markers in each chapter of the Rongyutang Shuihu zhuan A B C D E F G H I J a b c d e f g h ________________________________________________________________________________ 1 1 3 1 6 1 2 1 2 1A1 1 3 2 1 1 6 1 1 3 1 1 3 1F4 4 1 1 1 1 4 1 3 6 1 1 1 5 1 2 2 1 5 1 1 6 6 1 7 1d1 1 7 2 1 2 1 2 1 1 8 4 1 1 9 1 6 1 1 2 1 10 1 7 2 1 11 3 1 1 3 1 1 1 1 12 2 1 3 1 1 13 1 9 1 14 1 1 1 1 2 1 1 15 2 6 1 16 2 3 1 17 1B1 1 3 1 1 1 18 2 1 1 2 1 1 19 3 1 2F1 2 1 20 1 1 1 21 5 2 2 22 1 1 2 1 23 2 3 1 1 1 24 1 2 2 1 1 1 1 25 1 1 1 1 26 1 2 1 1 1 1 1 27 1 1 1 1 1 28 1 3 1 29 3 1 1 1 1 30 2 2 1 1 1 31 2 5 1 2 4 32 1 2 1 1 7 1 1d1 1 33 1 3 2 1 34 1 8 3 1 2 35 1C1 7 3 3 1 36 1 3 1 1 1 1 1 1 5 1a1 1 1 1 37 1C1 3 38 7 5 1 1 39 6 2 1 1 3 1 40 3 1 1 1 41 1 2F2 2 1 1 42 1 1 10 2 4 1 continued

Table 5 (continued) A B C D E F G H I J a b c d e f g h ________________________________________________________________________________ 1 2 1 1 1 1 43 2C 6 44 1 1 5 6 1 2 1 45 2 1F1 2 2 46 1B2 2 1 2 1 1 1 47 6 1 2F1 1 F3 48 3 2 1F3 4 1 49 1 2F3 1 1 50 3 1 1F1 1 1f 1 1 51 2 2 1 1 1 52 1 2 1 1F2 4 1 53 3 5 2 1 54 1 1 1F3 1 1 1 55 1 2 1 56 3 1 1 2 1 57 2 1 1F3 1 1 58 1 1F1 2 1 59 1 1 3 1 1 1 1 60 2 1 1 1 1 61 1 6 6 1 62 5 2 4 1 1 63 1 3F1 1 1 64 1 1 3 1 1 65 4 1 1F1 1 1 66 4 3F5 1 67 1 2 1 1 1 68 2 1F1 1 1 1 2 69 1 70 3 2F1 1 71 2 1 1 1 1 72 4 1 1 2 73 1 8 1 1 1 2 1 74 5 1 2F1 1 1 1 75 1 1 76 1 26 77 3 1F6 6 1 1 78 2 1 1d1 79 2 80 1 2 3 1 1 1 1 81 1 3 1 1 2 1h1 82 1 1 1 1 83 2 2 84 2 3 2 1 1 1 1 85 1F2 2 1 4 1 86 87

5 1

2 1

1 2

1 1

1

2

From Voice to Text

123

Table 5 (continued) A B C D E F G H I J a b c d e f g h ________________________________________________________________________________ 88 2 1 1 89 1 90 1 2 1 2 1 91 1 2 2 1 92 1 1 1 1 93 2 1 1 1 1 1 94 6 1 1 95 2 1 96 4 1 97 3 4 1 1 2 1 1 98 2 2F1 99 1 2 2 1d1 1 1 1 100 1 ________________________________________________________________________________ B1: A variant of B written as jican keyin, xiaoxing yezhu. B2: A variant of B written as jican keyin, yesu xiaoxing. C1: A homophonic substitute for C written as qiadai. F1: A variant of F written as shuo yan wei liao. F2: A variant of F written as shuo yan wei jue. F3: A homophonic substitute for F written as shuo you wei liao °—ºF. F4: A variant of F written as dao you wei liao. F5: A variant of F written as yan you wei liao. F6: A variant of F written as shuo you wei jue. a1: A homophonic substitute for a written as zhengzhi øΩ. d1: A homophonic substitute for d written as zhijiao u–. f1: A variant of f written as . . . pai shihou . h1: A homophonic substitute for h written as youfenjiao ≥¿Ê.

homophonic shuo you wei liao °—ºF75 and the close homophonic shuo yan wei liao. I have discussed earlier the homophonic “errors” as a possible “symptom” of a narrative text emerging from an oral tradition where phonetics outweighs orthography, but the fact that these stylistic features vary into homophonic substitutes is particularly significant. Since both stylistic features of the early period and those of the middle period appear in homophonic variants in writing, it may suggest a continuous process of textualization lasting from the early period to the middle period. Admittedly, again, the homophonic variation alone does not amount to indisputable evidence, but it is one more piece in the puzzle. The ending of most of the chapters in Shuihu zhuan, as we all know, is highly formulaic. In most cases the ending prognosticates, in the form of couplets, the action in the following chapter. Since the ending does not tell the story but is only a standard device for chapter division, it is less narrative than narratorial. We are in no position to decide, in clear-cut terms, whether each chapter coincides with an actual session of storytelling or the compilers were responsible

124

Chapter 4

for the chapter division. However, since oral storytelling is dependent on and adaptable to the changing ambiance, it would hardly be imaginable that the storytellers should have divided the Shuihu narrative into any fixed number of sessions. Although it cannot be proven, the chapter division in Shuihu zhuan was more likely done by the compilers. This, of course, does not preclude the possibility that some of the chapters could have coincided with the division of storytelling sessions traditionally or habitually used by the raconteurs. In any case, the chapter division was necessarily a late development, after the process of prosification and synthesis of the material from different sources was fundamentally completed. The language in the chapter ending, therefore, may be particularly indicative of an approximate dating for the completion of the text. The table above provides interesting information in that regard. Out of the one hundred chapters, sixty-five have the phrase “youfenjiao,” thirty-three have “zhijiao,” and twenty-five have both “youfenjiao” and “zhijiao.” All these occurrences, with no exception, are found in the chapter-ending formulas. That is to say, seventy-three chapters carry an ending that features “youfenjiao” and/or “zhijiao.” Such a systematic and extensive use of these two middle period stylistic features in the chapter-ending paragraphs makes it unlikely that the fanben text was completed in the late Yuan or early Ming. To be sure, Hanan’s middle period, from circa 1400 to circa 1575, is delimited rather generously, having a fifty-year overlap with the early period; but in general terms we may consider the heavy use of the middle period stylistic features in the chapter endings corroborative of the suggestions made by George A. Hayden and Y. W. Ma that, when Zhu Youdun published his two Shuihu plays in 1433, the Shuihu tradition could not have reached its fullest textual form with 108 bandit heroes in it. Both factors seem to point to a relatively late date—possibly around the middle of the fifteenth century or even slightly later—for the completion of the text that later appeared in the fanben editions. As one more testimony to Shuihu zhuan’s unique relationship to the vernacular stories, neither “youfenjiao” nor “zhijiao” appears in any chapter endings in Sanguo yanyi and Xiyou ji. The chapter ending in Sanguo yanyi is strictly uniform, consisting invariably of the phrase “zhengshi” followed by a couplet, which in turn leads to the formula: “weizhi . . . , qieting xiahui fenjie.” In Xiyou ji, most chapter endings follow the basic pattern in Sanguo yanyi, but with some variations: “zhengshi” often varies into “zhe zhengshi” or “Yi! zhengshi”; and “bijing” is more often used in place of “weizhi.” The same pattern is also followed in Jin ping mei cihua, but in the ending of chapter 6 we do see an instance of the phrase “youfenjiao.” Even this lone instance could be because of the novel’s textual relations with Shuihu zhuan, as that chapter is about the Ximen QingPan Jinlian adultery.

From Voice to Text

125

Some Linguistic Features Shared by Shuihu zhuan and the Vernacular Stories Shuihu zhuan, as discussed earlier, is an amalgamation of shorter sagas—the Lu Zhisheng saga, the Lin Chong saga, the Wu Song saga, the saga of the robbery of the birthday gifts (shengchen gang ), and so on. Although belonging to the same story complex of Liangshan bandits, these sagas had existed in different oral/performative genres and been largely autonomous and independent from each other before they were synthesized into a long narrative in oral prose. The fact that early period stylistic features are found in ninety-nine of the one hundred chapters in the Rongyutang edition seems to suggest that the textualization process for all those sagas had started at some point in the early period. But the early period, from before the Yuan until around 1450, spans almost two centuries. There had to be a long and gradual process for those sagas to become integrated into the longer narrative, some earlier and some later. If the textualization was not subsequent to but concurrent with the process of synthesis, the textualizing process for different sections of Shuihu zhuan could have started at different moments in the early period. This means that, even though the different sections of the narrative all reached their fullest textual form at a certain point in the middle period, the textualizing process of some sections could have started earlier than others. If this was true, we may expect to find meaningful differences in the language in different sections of Shuihu zhuan. We can continue to experiment with stylistic features, which, considered within the context relevant to their use, may be more reliable than those features that are less context-bound.76 Unfortunately, however, stylistic features have to be from the plain narrative only, and dialogue, where the stylistic context is more difficult to establish, has to be excluded. Yet it is precisely dialogues that are most reflective of the diachronic changes in language, while the plain narrative is free of individual traits and for that reason more resistant to linguistic mutability. This is the reason that, when Patrick Hanan discusses the language properties in his “Trial Group” for determining stories of the early period, he finds it necessary to supplement the study of stylistic features with one of the “linguistic features,” especially modal particles, not only from the plain narrative but also from dialogues. The use of one such particle, “xiu,” for example, is found to be virtually confined to Shuihu zhuan and the stories belonging to the trial group for the early period.77 The result, therefore, “confirms the notion of the trial group as distinguishable by linguistic means.”78 Hu Zhu’an’s article on modal particles in early vernacular texts, which Hanan cites, again evidences Shuihu zhuan’s unique relationship with the vernacular stories. Ten particles, according to Hu Zhu’an, are found frequently

126

Chapter 4

in both Shuihu zhuan and the short stories, and the two sentence-final particles, “zege” and “xiu,” occur only occasionally in other early vernacular texts.79 Patrick Hanan has noted that, apart from Shuihu zhuan and the early stories, the sentence-final particle “xiu” occurs only once in Wudai shi pinghua and once in the twenty-chapter version of Ping yao zhuan.80 I have seen four occurrences of “xiu” in the nanxi texts from the Yongle dadian, including three in Zhang Xie zhuang yuan and one in Huanmen zidi. In the Yuan-edition zaju the use of “xiu” is largely confined to the exclamation “Xiu! Xiu! Xiu,” which occurs in several plays. It, however, does not occur in any of Zhu Youdun’s plays collected in Shemotashi qucong. Nor does it appear in the 1432 manuscript of the nanxi Liu Xibi Jinchai ji, a text contemporaneous to Zhu Youdun’s plays. These findings suggest that by the first half of the fifteenth century, “xiu” was largely out of use, although it does not indisputably prove the case. The occurrence of “xiu” in Shuihu zhuan supports the argument that, although the complete fanben version of Shuihu zhuan may not have come into existence by Zhu Youdun’s time, the textualization process must have started much earlier. Indeed, “xiu” can be safely counted as a linguistic marker for the early period, but its occurrences in Shuihu zhuan are not frequent enough for it to serve as a reliable criterion.81 Fortunately, there are numerous other linguistic features that Shuihu zhuan and the vernacular stories have in common. In addition to the sentence-final particles “xiu” and “zege,” they also include the verb “zuodi,” adverbs such as “zhenge,” “zhengxiege/xianxiege,” and “duande,” modal particles “po’nai” and “wuzi,” and pronouns such as “wuna” and “wudi.” 82 Most of the features occur at roughly comparable frequencies in the stories from both the early and middle periods and therefore will not help differentiate one period from the other. Only “zege” and “wuzi” stand out as most useful for our purpose here: They both have a significant presence in Shuihu zhuan, and each appears in the early stories and the middle period stories at meaningfully different frequencies. If we can establish “zege” and “wuzi” as two linguistic markers for a certain period, their frequency in a section of Shuihu zhuan may suggest the relative earliness or lateness of that part of the text. I will present a survey of the instances of these two features in vernacular stories from both the early and middle periods. My survey does not include those four stories published by Xiong Longfeng (fl. 1590) in the late sixteenth century. Two of the stories, “Feng Boyu fengyue xiangsi” (which has a duplicate in Qingping shantang huaben) and “Kong Shufang shuangyu shanzhui zhuan,” appear in wenyan and will not serve our purpose in any way. The story “Zhang Sheng cailuandeng zhuan” is for the most part a duplicate of “Zhang Shunmei dengxiao de linü” in Yushi ming yan. The only vernacular story that is excluded is therefore “Su Changgong Zhang Tailiu zhuan,” which should

From Voice to Text

127

not significantly affect the results of the survey. Among the other stories, some are not listed here simply because they do not provide relevant linguistic data. These are the stories in which the language is more simplified wenyan than baihua or in which the role of prose is overshadowed by that of verse. Most of such pieces are what Patrick Hanan calls “virtuoso stories.” The following four stories categorized by Hanan as from the early period are therefore excluded: “Qiantang meng,” “Fengyue ruixian ting,” “Qian Sheren tishi Yanzi Lou,” and “Suxiang Ting Zhang Hao yu Yingying.” For the same reason, the following twelve stories classified by Hanan as from the middle period are not listed: “Lanqiao ji,” “Fengyue xiangsi,” “Zhang Zifang mudao ji,” “Zhao Bosheng chasi yu Renzong,” “Yang Jiaoai sizhan Jing Ke,” “Si sheng jiao Fan Zhang ji shu,” “Han Li Guang shihao Fei Jiangjun,” “Kuiguan Yao Bian diao Zhuge,” “Zhachuan Xiao Chen bian Bawang,” “Mei xing zhengchun,” “Feicui xuan,” and “Yan Pingzhong er tao sha san shi.” In the cases in which a story in Qingping shantang huaben has a duplicate in the San yan with only minor adaptations, only the story in Qingping is listed in order to protect the integrity of the statistics. For instance, while “Wenjing yuanyang hui” from Qingping is included, “Jiang Shuzhen Wenjing yuanyang hui” from Jingshi tong yan is left out. Tables 6 and 7 show respectively the frequencies of “zege” and “wuzi” in the stories from the early and middle periods by Patrick Hanan’s division. As we can see from the tables, only two out of the twenty-nine stories (6.9 percent) from the early period do not have either “zege” or “wuzi,” while among the sixteen stories from the middle period, as many as eleven (69 percent) do not show either feature. Also, as indicated by the numbers of the average occurrences of the two features in each group, the chance for them to appear in the stories from the early period is almost ten times as strong as in the stories from the middle period. In the Yuan-edition zaju texts, I have seen only one instance of “zege” in Dongchuang shifa, while “zan,” possibly an earlier written form registering the same expression as “zege,”83 appears frequently in many of the plays. “Zan” may have evolved from the still earlier form “zhe” as a result of the phonetic change of the expression. In the Jiajing edition of Dong Jieyuan’s Xixiang ji zhugong jiao, both “zan” and “zege” are found, but it is hard to tell whether the coexistence of the two forms was from the original text dating from the Jin period or it is evidence of the tampering of later editors. “Wuzi” is seen in the texts of the Yuan-edition zaju, and it often appears as the variant “shangwuzi” |a¤, which in turn frequently adopts homophonic substitutes such as “shangwuzi” |al, “shangguzi” |jl, and “shangguzi” |©l, as Kòsaka has noted.84 In Shuihu zhuan, “wuzi” sometimes varies into the close homophonic substitute “wushi.” The fluid and erratic written forms for each of the two features in early ver-

Table 6 Incidence of zege and wuzi in early period stories Abbreviation a

Title of the story

zege

wuzi

QP2 QP3 QP8 QP11 QP13 QP15 QP16 GJ15 GJ24 GJ33 GJ36 TY6 TY8 TY13 TY14 TY16 TY19 TY20 TY28 TY30 TY36 TY37 TY39 HY12 HY13 HY14 HY21 HY31 HY33

“Jiantie heshang” “Xihu santa ji” “Luoyang san guai ji” “Yinzhi jishan” “Wujie Chanshi si Honglian ji” “Yang Wen ‘Lanluhu’ zhuan” “Huadengjiao Liannü cheng fo ji” “Shi Hongzhao long hu junchen hui” “Yang Siwen Yanshan feng guren” “Zhang Gulao zhong gua qu Wennü” “Song Sigong da’nao ‘Jinhun’ Zhang” “Yu Zhongju tishi yu shanghuang”* “Cui Daizhao shengsi yuanjia” “San xianshen Bao Longtu duanyuan” “Yi ku gui lai daoren chu guai” “Xiao furen jinqian zeng nianshao” “Cui Ya’nei baiyao zhaoyao” “Ji Yafan jinman chanhuo” “Bai Niangzi yong zhen Leifengta” “Jinming Chi Wu Qing feng Aiai” “Zaojiaolin dawang jia xing” “Wan Xiuniang chou bao Shantinger” “Fu Lu Shou san xing dushi” “Foyin Shi si tiao Qinniang” “Kan pixue dan zheng Erlangshen” “Nao Fanlou duoqing Zhou Shengxian” “Lü Dongbin feijian zhan Huang Long” “Zheng Jieshi ligong shenbi gong” “Shiwu guan xiyan cheng qiaohuo”

1 0 0 3 3 7 3 5 4 3 11 1 4 8 5 3 5 4 15 2 2 5 3 1 2 7 1 2 1

1 0 0 1 0 0 0 2 0 0 6 2 0 3 4 2 0 3 2 0 0 1 0 0 0 1 0 1 0

Total

_______________________________________________________________________

Average

3.83

2 0 0 4 3 7 3 7 4 3 17 3 4 11 9 5 5 7 17 2 2 6 3 1 2 8 1 3 1

1

4.83

_______________________________________________________________________

*The prologue of this story is excluded because it duplicates the virtuoso story “Fengyue ruixian ting.” a. Abbreviations used are as follows: QP

Qingping shantang huaben

GJ

Gujin xiaoshuo (also published under the title Yushi ming yan)

TY

Jingshi tong yan

HY

Xingshi heng yan

From Voice to Text

129

Table 7 Incidence of zege and wuzi in middle period stories Abbreviation a

Title of the story

QP1 QP4 QP7 QP12 QP14 QP17 QP18 QP19 QP20 QP27 GJ3 GJ23 GJ26 GJ29 GJ38 TY7

“Liu Qiqing shijiu Wanjianglou ji” “Hetong wenzi ji” “Kuaizui Li Cuilian” “Chen Xunjian Meiling shi qi ji” “Wenjing yuanyang hui” “Cao Boming cuo kan zang ji” “Cuo ren shi” “Dong Yong yu xian ji” “Jiezhier ji” “Li Yuan Wujiang jiu zhushe” “Xinqiaoshi Han Wu mai chunqing” “Zhang Shunmei dengxiao de linü”* “Shen Xiaoguan yi niao hai qi ming” “Yueming Heshang du Liu Cui” “Ren Xiaozi liexing wei shen” “Chen Kechang Duanyang xianhua”

zege

wuzi

Total

0 0 0 2 0 0 0 0 0 0 0 1 0 2 2 0

0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0

0 0 0 2 0 0 0 0 0 0 1 1 0 2 2 0

0.44

0.06

0.5

_______________________________________________________________________

Average

_______________________________________________________________________

*The prologue is excluded as a virtuoso story. a. Abbreviations used are as follows: QP

Qingping shantang huaben

GJ

Gujin xiaoshuo (also published under the title Yushi ming yan)

TY

Jingshi tong yan

HY

Xingshi heng yan

nacular literature suggest that these expressions, which had hitherto existed orally, were only starting to be tentatively recorded in written characters. In the twenty-chapter version of San Sui ping yao zhuan, “zege” is written as either “zege” h” or “zege” hñ in forty-three occurrences, almost evenly attributed throughout the entire length of the work. Both “zege” and “wuzi” make very few appearances, if any at all, in Sanguo yanyi and Xiyou ji. “Zege” does appear in Jin ping mei cihua, but the occurrences are largely confined to the portion of the narrative taken over from Shuihu zhuan. As a matter of fact, of the fifteen instances of “zege” in that part of Shuihu zhuan, eight are carried over in the corresponding portion in Jin ping mei cihua. In the other seven cases, it seems to have been deliberately dropped out of the dialogue, which has been otherwise taken almost verbatim from Shuihu zhuan. For instance, “kan niangzi zuo shenghuo zege” and “Ganniang raoshu zege” in the Rongyutang edition of Shuihu zhuan (2: 771, 2: 786) become “kan niangzi zuo shenghuo” and “Ganni-

130

Chapter 4

ang raoshu” in Jin ping mei cihua (1: 77; 1: 88). The feature “wuzi” does not appear in Jin ping mei cihua, not even in the section taken from Shuihu zhuan. For another comparison, “zege” occurs eleven times in the nanxi texts from the Yongle dadian, including six times in Xiao Sun tu and five times in Zhang Xie zhuang yuan, while “wuzi” does not appear in any of the three plays. In Zhu Youdun’s twenty-two plays collected in Shemotashi qucong, “zege” appears only once, in Yanhua meng, and “wuzi” does not occur anywhere. What we see in these vernacular works is consistent with the data from my survey of the vernacular stories. It seems that the use of “zege” and “wuzi” were indeed on the decline from the Yuan to the Ming and tapered off in the late Ming period. Ling Mengchu’s Er pai, however, is an anomaly, where “zege,” along with the plural suffix “mei,” is regularly used. One possibility is that Ling deliberately used such anachronic linguistic features as part of his strategy to simulate oral storytelling of an earlier period. Since the vernacular stories and Shuihu zhuan may have shared a common storehouse of story-making material and similar linguistic environment, we may examine the occurrences of these two features in different sections of Shuihu zhuan and look for any significant difference in the frequencies of their appearances. Table 8 divides the Rongyutang edition into twenty-three sections, each designated by an English letter. In most cases, each section coincides with what we have called a “short saga,” and in some other cases, a section contains a series of actions that take place in a particular period of the rebellion, although these actions may not be coherent enough to belong to one self-sufficient “short saga.” In still a few other cases, a section may be as short as one or two chapters, serving as a transition between two other longer sections. Since a section can begin or end at different places in a chapter, I attach a subscript letter to the chapter number to indicate that place more accurately; for example, 1–b, 2–m, and 3–e would mean, respectively, the beginning of chapter 1, the middle of chapter 2, and the end of chapter 3. Table 9 shows the use of the two linguistic features, “zege” and “wuzi,” in each of the sections, with the numbers indicating the times of occurrences. For the “length of section” I use the chapter as a unit of measurement, as the chapters are largely comparable with each other in length. Since a section can begin or end at different places in a chapter, the length of a section is not always indicated by a whole number. In some cases a decimal figure has to be adopted to represent a more accurate estimate. Table 9 shows very uneven results. Among all the sections, G and J feature the most frequent uses of “zege” and “wuzi,” with an average of 3.4 and 5.2 times per chapter respectively. These figures may not be surprising, as it has been noted earlier that killing an adulterous woman in order to avenge a wronged or murdered brother or sworn brother is a recurrent thematic pat-

Table 8 Division of sections in Shuihu zhuan Range

Central figure(s) or event(s)

_______________________________________________________________________ A B C D E F G H I J K L M N O

1b – 2b 2b – 3m 3m – 7m 7m – 12m 12m – 20m 20m – 23b 23b – 32m 32m – 42e 42e – 44b 44b – 46m 46m – 50e 51b – 52b 52b – 54e 54e – 58m 58m – 60e

P Q R

60e – 67m 67m – 68e 69b – 71e

S T

71e – 74e 74e – 82e

U V W

83b – 90m 90m – 99e 99e – 100e

Martial Hong’s release of the demons Wang Jin and Shi Jin Lu Da (Lu Zhishen) Lin Chong Yang Zhi and the robbery of the birthday gift Song Jiang (I) Wu Song Song Jiang (II) Li Kui’s visit to his mother and killing of four tigers Shi Xiu and Yang Xiong Three attacks on the Zhu Family Village Lei Heng and Zhu Tong Rescue of Chai Jin from the prison in Gaotang Prefecture Huyan Zhuo Sack of Huazhou Prefecture; Mangdangshan; death of Chao Gai Lu Junyi and the sack of Daming Prefecture Sack of Zengtoushi Sack of Dongping Prefecture and Dongchang Prefecture; final ranking of 108 bandit heroes Excursions of Li Kui and Yan Qing Battles and negotiations between Liangshan and the court; the amnesty for the rebels Expedition against the Liao Expedition against Fang La Death of Song Jiang

_______________________________________________________________________

Table 9 Incidence of zege and wuzi in the Rongyutang Shuihu zhuan Section Length

zege

wuzi

Total

OPC

Section

Length

zege

wuzi

1 0 3 0 11 5 32 25 0 13 5 0

1 0 0.75 0 1.38 1.92 3.4 2.43 0 5.2 0.9 0

M N O P Q R S T U V W Total

3 3.5 2.5 6.5 1.5 2.8 3.2 8 7.5 9.3 1.2 100

0 0 0 8 0 1 2 5 8 0 0 86

2 2 0 1 2 0 4 2 1 2 0 49

Total OPC*

_______________________________________________________________________ A B C D E F G H I J K L

1 1.5 4 5 8 2.6 9.4 10.3 1.2 2.5 4.5 1

0 0 0 0 4 3 24 17 0 9 5 0

1 0 3 0 7 2 8 8 0 4 0 0

2 2 0 9 2 1 6 7 9 2 0 135

0.67 0.57 0 1.38 1.33 0.36 1.88 0.88 1.2 0.22 0 1.35

_______________________________________________________________________

*Occurrences per chapter

132

Chapter 4

tern in several Yuan zaju plays.85 Among the known titles of Yuan zaju, there are Bing Yang Xiong and Zhedanr Wu Song da hu, both by “Hongzi” Li Er,86 and of course a story entitled “Wu Xingzhe” was circulating as early as the Southern Song period, as we are informed in Zuiweng tanlu. It might well be the case that these two sections are among those that were incorporated into the long Shuihu narrative and began to get textualized early. The two linguistic features also have a significant presence in the two Song Jiang sections, F and H. Section S, in which the two features appear 1.88 times per chapter, seems to have started its textual existence rather early as well. This is quite consistent with the fact that Li Kui and Yan Qing are the two most prominent figures in early Shuihu plays, even though most of the plays are not extant. As also noted earlier, the second half of chapter 73 in Shuihu zhuan parallels the plot in Kang Jinzhi’s extant play, Heixuanfeng fujing. There may have been possible correspondences between other episodes in section S and some nonextant early Shuihu plays, such as Gao Wenxiu’s Heixuanfeng qiao jiaoxue and Yang Xianzhi’s Heixuanfeng qiao duan’an. With the texts lost, we have no access to the contents of the plays, but the phrases in the titles—“qiao jiaoxue” (to pretend teaching) and “qiao duan’an” (to pretend judging) might suggest some similarities between the dramatic plots and the action in chapter 74, where Li Kui makes havoc of a yamen and a school.87 Neither “zege” nor “wuzi” occur in section D, the story of Lin Chong, which runs as long as five chapters. This may suggest that—the hero’s pivotal role in the Liangshan rebellion notwithstanding—the Lin Chong story might have joined the narrative relatively late. Indeed, among the major Shuihu figures, Lin Chong’s status is unique, as he is the only one who features as the protagonist in a long section of the novel but whose name is not listed in Gong Shengyu’s Song Jiang sanshiliu zan. One has also to be reminded that Lin Chong’s name is absent from any of the known titles of the early Shuihu zaju.88 The case for Lu Zhishen is quite different. We know from Zuiweng tanlu that a ganbang tale in the Southern Song period carried the title of Hua Heshang, which was possibly a story of Lu Zhishen, who is nicknamed Hua Heshang. In the Yuan zaju Heixuanfeng fujing, Lu Zhishen appears as an important figure, having to clear himself of a misdeed he has been wrongly accused of by Li Kui. Then of course there are the anonymous late Yuan play Huanghua yu and Zhu Youdun’s Baozi heshang zi huansu, both of which feature Lu Zhishen as the leading character. It seems that the story of Lu Zhishen might have been part of the Shuihu cycle ever since the earliest days of the tradition. Yet, our two linguistic criteria occur in section C at a rather low frequency, merely 0.75 times per chapter. This may suggest that the Lu Zhishen saga that found its way into Shuihu zhuan could be from a later tradition, dif-

From Voice to Text

133

ferent from the older one that had found its expressions in the ganbang story and the zaju plays. Indeed, the celibate Lu Zhishen in Shuihu zhuan is radically different from the monk in Zhu Youdun’s play, where he is visited in his monastery by his mother, wife, and son, who take turns to try to persuade him to return to Liangshan. This, however, does not mean the textualization of the Lu Zhishen section in Shuihu zhuan started necessarily later than Zhu Youdun’s play, which dates from 1433. But it may mean that the play could have been based on an older and different tradition that had existed in some texts from earlier times. In Zhu Youdun’s time, the Lu Zhishen story as we have it in Shuihu zhuan may have already become part of the long narrative and the textualization may have been underway, as the existence of the early period stylistic features in that part of Shuihu zhuan would suggest. Yet since the frequency of “zege” and “wuzi” is relatively low in the Lu Zhishen section of Shuihu zhuan, the textualization of the Lu Zhishen saga might not have started much earlier than Zhu Youdun’s time. In some shorter sections, the two linguistic features either make no appearance at all or occur only at very low frequencies. This phenomenon cannot be explained in each of the sections, but in most cases it seems to be quite consistent with what is known about the evolution of the narrative. Section I, which relates Li Kui’s loss of his mother and his killing of four tigers, makes no use of either “zege” or “wuzi.” Sandwiched between sections H and J, both featuring high frequencies of the features, the language in section I looks distinctly different. The result of Li Kui’s excursion from Liangshan is to bring back two new participants in the rebellion, Zhu Fu and Li Yun, both of whom are to become minor chieftains. The possible lateness of this section, suggested by the absence of the two linguistic criteria, may be corroborated by the fact that the plot in the section has no bearing on the gathering of the Thirty-Six, which might be the total number for the chieftains in the earlier stage of the Shuihu tradition. The same may be said of section O, where the two linguistic features do not show up either. A larger part of section O is about the rebels on Mangdangshan and their joining the Liangshan bandits. Again, all of them— Fan Rui, Xiang Chong, and Li Gun—are to become minor chieftains on Liangshan. In section L, another in which the two features make no appearance, we do have two people—Lei Heng and Zhu Tong—who are to become members of the group of the Thirty-Six, but it is still possible that this section could have been a late arrangement, as it clearly serves as a transitional link between two longer narrative sequences, sections K and M.89 Section R, which tells of the addition of Zhang Qing and Dong Ping to the rebels and the ranking of the bandit heroes, also features a low frequency for the two linguistic criteria of 0.36 times per chapter. As members of the

134

Chapter 4

Thirty-Six, Zhang Qing and Dong Ping may have been part of the Shuihu cycle ever since the earliest days of the tradition. Indeed, both names are found in Xuanhe yishi and Song Jiang sanshiliu zan, although in the former text Zhang Qing’s name appears as the homophonic Zhang Qing iC, which becomes in Shuihu zhuan the written form for the name of the minor chieftain Zhang Qing, with the nickname Caiyuanzi. As in the case of Lu Zhishen, however, Zhang Qing and Dong Ping in Shuihu zhuan, where both are district military commanders (Bingma dujian) before joining the rebellion, are very different from those two characters of the same names in Xuanhe yishi.90 It may not be impossible that the stories of Zhang Qing and Dong Ping in Shuihu zhuan, like that of Lu Zhishen, were textualized at a later stage of the tradition. As for the episode of ranking the bandit chieftains, there is a more obvious reason for the possible lateness of the text. Since there had been only thirty-six chieftains in the early stage of the tradition and the number of 108 was a relatively late development, it would be impossible for the ranking of all the chieftains, including the thirty-six major ones and seventy-two minor ones, to take place before the assemblage of all the 108 was accomplished.

Frequency Ratios of Some Early Linguistic Markers to Their Later Replacements Yet while these two linguistic criteria, “zege” and “wuzi,” are certainly helpful in our effort to demonstrate the cumulative process of textualization, such evidence may still look tenuous if not supported by further linguistic data from the text. This is especially because we cannot rule out the possibility that the frequency of a linguistic feature in a particular section, especially a short one, is a mere fluke. A high frequency of the criteria may be a relatively more reliable indication of the earliness of the section, as in most cases a later vernacular narrative would not show a strong penchant for a colloquial usage prevalent in an earlier period but fading away in its own. A low frequency of the criteria, however, should not be accepted unquestioningly as evidence for the lateness of the text, as there is a chance that the storyteller or the scribe, out of personal propensity or some other reason, did not use a current expression as frequently as most of his or her contemporaries. There is another factor that could also affect the reliability of the criteria. As “zege” appears invariably in dialogues, a section that is relatively more “dialogue-light” than others would tend to feature a lower frequency of its use, which may misleadingly suggest the relative lateness of that section.91 The results of the survey above, therefore, should be tested. Both “zege” and “wuzi” have served our purpose well, but they were heavily used in the early period and then faded away, not replaced in either case by a later equiv-

From Voice to Text

135

alent. If, during the textualizing process of Shuihu zhuan, a linguistic feature from an earlier time was being gradually replaced by its successor of a later period, the ratio between the occurrences of the pair in a certain section compared with that in other sections could suggest the relative earliness or lateness of that part of the text. Obviously, the occurrences of the features should be frequent, and if possible, one hopes that more than one of such pairs can be found. The reason is simple: If the survey covers the text at a considerably high degree of density, the chance for the result to be a mere fluke would be significantly reduced. One of such pairs consists of the adverbs “bian” and “jiu.” Both adverbs indicate temporal or logical immediacy of an action and therefore, in that meaning, they are interchangeable.92 Both are used extensively in Shuihu zhuan. “Bian” was the dominant feature in early vernacular literature, while “jiu” came into use later, rarely making its appearance until perhaps the late Yuan.93 The transition from “bian” to “jiu” must have been a very slow and protracted process, as “bian” has not, even to this day, been completely replaced by “jiu,” although “jiu” is now the regular form and much more frequently used in modern Chinese. Since the two features coexist virtually in all works of vernacular literature dated from the late Yuan and after, the ratio between their frequencies may approximately suggest the date of the text. By comparing the ratios between “bian” and “jiu” in vernacular works from different periods, Mei Tsu-lin has been able to verify the dating of a number of dramatic texts and has even proposed the following ratios of “jiu” to “bian” as dating criteria: If the ratio is larger than 1:6 (i.e., 1:5, 1:4, etc.), the text should not be earlier than 1300; if the ratio is larger than 1:2, the text should not be earlier than 1350; and if the ratio is or is larger than 1:1, the text should not be earlier than 1400.94 We are not concerned here about the accuracy of Mei’s criteria, but his effort clearly illustrates the correlation between that ratio in a vernacular work and its relative earliness or lateness. Admittedly, the use of these features could be affected by other factors, such as the work’s stylistic orientation and the writer’s personal linguistic propensity. But since we are dealing not with different works but different parts of the same work, where most contextual conditions may have been highly consistent, the frequency ratio becomes an even more reliable indicator of relative and approximate dating of the text. One does not find in Shuihu zhuan another pair of linguistic features that occur at frequencies comparable to those of “bian” and “jiu.” Notice in the text, however, that the action of carrying or holding is expressed by two different groups of verbs. In the first group, “jiang” is the regular form but “ba” is also used.95 In modern Chinese, both are used only as prepositions in what Wang

136

Chapter 4

Li has termed “dispositive sentences” (chuzhi shi).96 But before the Tang period, both words were used only as verbs, and even after they came to be used as prepositions, they continued to appear as verbs for a long time.97 In addition to their function as verbs, both “jiang” and “ba” also appear in Shuihu zhuan as “dispositive” prepositions,98 which may suggest that the textualization of the narrative was concurrent with the transition of the two from verbs to prepositions. The other group consists of words that are still used as verbs for the action of carrying and holding in modern Chinese: “qu,” “na,” “chi,” “dai,” and “peng.” There is not sufficient textual data to prove that any of these verbs came into use later than the verbs “jiang” and “ba.”99 It seems justified, however, to consider the second group the newer one in relation to the first. As “jiang” and “ba” were in a transitional process from pure verbs to pure prepositions, their chance to appear as verbs was gradually and continuously diminishing, until their job as verbs was completely taken over by the members of the second group. As in the case of “bian” and “jiu,” the ratio of the frequency of the first group to that of the second group may lead to a rough estimate of the date of the vernacular text. Generally speaking, the later a text is, the fewer uses of the first group, and hence the smaller the ratio. Thus we have two pairs, “bian” and “jiu,” and the “jiang” group and the “qu” group, which we will simply call “jiang” and “qu,” using the two words as representatives of their respective groups. Before we examine the ratios of the older features to their later equivalents in different sections of Shuihu zhuan, let us have a look at their occurrences in some other works that appeared in a span of approximately two hundred years, from the beginning of the fifteenth century to the early seventeenth (Table 10).100 The purpose is to test whether the ratios have indeed any correlation to the estimated dating of a vernacular text. The numbers in Table 10 do not give us any clues regarding accurate dating of any of the works, but they do delineate a process of gradual shift from “bian” to “jiu” and from “jiang” to “qu.” This suggests that the transition from earlier linguistic features to their later equivalents may indeed provide a useful context for our study of the textualization of Shuihu zhuan. Yet the relevance of such data from these works should not be taken for granted. For one thing, the dating of some of the texts is problematic. Only Zhu Youdun’s plays are unequivocally dated; in all other works in the group, the date for the publication of a particular edition should not be taken for granted as the date for the text. It therefore remains to be proven whether the linguistic discrepancies between one work and another were caused by the diachronic change of the language or by synchronic disparities across regional or dialectal boundaries. Additionally, the generic heterogeneity of the group may lead one to won-

From Voice to Text

137

Table 10 Incidence of bian, jiu, jiang, qu in some vernacular works Work

bian

jiu

b : j*

jiang

qu

j : q*

Three xiwen plays in Yongle dadiana Xuande xieben Jinchai ji (1432) 1498 edition of Xixiang ji zajub Zhu Youdun’s playsc San Sui Ping yao zhuand Jin ping mei cihuae

112 41 54 319 206 75

5 3 8 84 62 122

22.4 13.67 6.75 3.8 3.32 0.61

7 21 28 99 37 0

1 8 3 50 75 43

7 2.63 9.3 1.98 0.49 0

_______________________________________________________________________

_______________________________________________________________________

*b: j = bian : jiu;

j: q = jiang : qu.

a. For the three plays, only the binbai lines in the scripts are surveyed. This is also the case with the other dramatic works included in the chart. The reason is that the language in the arias is often stylized and much more literary, and therefore may not faithfully reflect the current linguistic development of the time. b. There are eight instances of the adverb “jiu” in the 1498 edition of Xixiang ji, while in all thirty Yuanedition zaju plays there is only one occurrence of “jiu” that can possibly be counted as an adverb. This discrepancy alone seems to be convincing evidence that the binbai prose as we see it in Xixiang ji is for the most part not written by the original Yuan playwright, but a result of the dynamic between writing and oral performance in subsequent times. On the other hand, however, the date this edition appeared in print should not mislead us to think that the vernacular prose in the text represents faithfully the living language at the end of the fifteenth century. Rather, it was a product of a long process of cumulative textualization since the early Yuan. c. The survey covers Zhu Youdun’s twenty-two plays collected in Shemotashi qucong. d. The survey is on the facsimile edition of the twenty-chapter version in Guben xiaoshuo jicheng. e. The survey is on chapters 7, 8, 10, 11, 12, and 13. The first six chapters and chapter 9 are skipped because the narrative in those chapters is heavily dependent on Shuihu zhuan.

der to what extent such discrepancies were determined by divergent stylistic exigencies pertaining to different genres. We need, therefore, a better context to observe the shift from “bian” and “jiang” to “jiu” and “qu.” Ideally, we need a group of works from the same genre, so that formal differences can be precluded as a possible factor for the linguistic disparities from one work to another. Even though accurate dating for each work does not have to be a prerequisite, the correlative earliness and lateness of the works should have been basically and reliably determined. Again, thanks to Patrick Hanan’s effort, we do have such a context with the short stories. Tables 11 and 12 show the occurrences of the earlier features “bian” and “jiang” and the later “jiu” and “qu” in the stories from the early and middle periods respectively. As one can see from the tables, the ratio of earlier features to later features in the stories from the early period is not uniformly larger than that in the sto-

Table 11 Incidence of bian/jiang and jiu/qu in early period vernacular stories* Story

Early features

Later features

__________________________

_______________________

bian

jiang

Total

jiu

qu

Total

13 5 17 11 11 30 29 22 12 11 57 43 21 22 11 22 14 52 44 20 22 30 12 13 88 38 7 59 47 783

21 1 3 0 8 10 3 9 2 5 13 3 8 6 1 0 8 5 10 2 3 4 4 0 10 7 5 12 4 166

34 6 20 11 19 40 32 31 14 16 70 46 29 28 12 22 22 57 54 22 25 34 16 13 98 45 12 71 51 949

1 1 2 1 1 3 1 5 4 2 5 3 8 9 1 2 1 4 10 6 3 2 0 2 13 4 4 9 6 113

5 6 1 2 6 5 1 7 7 8 20 7 5 7 2 4 3 3 12 2 3 2 4 2 18 4 16 4 5 171

6 7 3 3 7 8 2 12 11 10 25 10 13 16 3 6 4 7 22 8 6 4 4 4 31 8 20 13 11 284

_______________________________________________________________________ QP2 QP3 QP8 QP11 QP13 QP15 QP16 GJ15 GJ24 GJ33 GJ36 TY6 TY8 TY13 TY14 TY16 TY19 TY20 TY28 TY30 TY36 TY37 TY39 HY12 HY13 HY14 HY21 HY31 HY33 Total

_______________________________________________________________________

*Bian: jiu

783: 113 = 6.93

Jiang: qu 166: 171 = 0.97 Early features: Later features

949: 284 = 3.34

ries from the middle period. Some stories, such as QP3 and GJ23, are very short, and there is no guarantee that the ratio in a particular story is not a fluke. Collectively, however, both groups of stories are adequately large bases for the statistical survey, and the numbers for the occurrences of the features are too big to be merely fortuitous. As shown, both the ratio of “bian” to “jiu” and that of “jiang” to “qu” drop significantly from the first group to the second group,

From Voice to Text

139

Table 12 Incidence of bian/jiang and jiu/qu in middle period vernacular stories* Story

Early features

Later features

__________________________

_______________________

bian

jiang

jiu

0 2 26 9 7 3 27 19 10 2 13 3 25 11 23 9 189

1 7 1 2 0 0 6 5 6 1 4 1 8 5 0 3 50

Total

qu

Total

_______________________________________________________________________ QP1 QP4 QP7 QP12 QP14 QP17 QP18 QP19 QP20 QP27 GJ3 GJ23 GJ26 GJ29 GJ38 TY7 Total

1 9 27 11 7 3 33 24 16 3 17 4 33 16 23 12 239

0 3 5 2 9 0 9 3 8 5 6 2 6 7 5 8 78

2 5 5 2 3 4 9 4 3 5 5 1 7 4 9 7 75

2 8 10 4 12 4 18 7 11 10 11 3 13 11 14 15 153

_______________________________________________________________________

*Bian: jiu

189: 78 = 2.42

Jiang: qu

50: 75 = 0.67

Earlier features: later features

239: 153 = 1.56

and the overall ratio of the earlier features to the later ones declines from 3.34 to 1.56, a 53 percent decrease. The result of the survey is consistent with what has been seen earlier in the texts from other genres. It confirms that “bian” and “jiang” were indeed in a process of being gradually replaced by “jiu” and “qu” from the Yuan period to the sixteenth century, a long process encompassing the textualization of Shuihu zhuan. This linguistic change must have left its mark on Shuihu zhuan as well, especially since the novel bears conspicuous thematic and linguistic affinities with the short stories, as noted earlier. If some sections in the novel have a textual history longer than that of other sections, as the study of the occurrences of “zege” and “wuzi” suggests, we may expect the ratio of these earlier features to the later ones to differ significantly from one section to another. Table 13 presents the number of occurrences of “bian,” “jiu,” “jiang,” and “qu” in each of the sections in Shuihu zhuan and the ratio of the earlier features to the later ones. The ratio of the earlier features to the later ones varies radically from one

140

Chapter 4

Table 13 Incidence of earlier features and later features in each section of the Rongyutang Shuihu zhuan Section

Early features

_______________________

Later features

_____________________

bian

jiang

Total

jiu

qu

12 35 151 127 240 199 488 558 57 145 148 33 05 146 53 226 73 79 103 189 186 239 21

1 8 25 25 37 22 76 40 3 16 13 0 7 3 5 10 4 4 9 5 12 7 0

13 44 176 152 277 141 564 598 60 161 161 33 112 149 58 236 77 83 112 194 198 246 21

3 11 17 31 44 12 66 69 12 23 18 4 17 35 12 47 10 13 14 48 74 66 4

6 17 69 69 103 32 119 107 20 14 30 10 33 22 8 44 4 23 22 43 30 46 4

Total

E : L*

_______________________________________________________________________ A B C D E F G H I J K L M N O P Q R S T U V W

9 28 86 100 147 44 185 176 32 37 48 14 50 57 20 91 14 36 36 91 104 112 8

1.44 1.43 2.05 1.52 1.88 3.20 3.05 3.40 1.88 4.35 3.35 2.36 2.24 2.61 2.90 2.59 5.50 2.31 3.11 2.13 1.90 2.20 2.62

_______________________________________________________________________

*Early features: Later features. section to another, ranging from 1.43 to 5.50. But before we make any comments, we may compare the results here with the ones we reached in our earlier survey on “zege” and “wuzi.” In Table 14, all the sections in Shuihu zhuan are ranked in two ways, in terms of both the ratio of the earlier features to their later replacements and the average occurrences of “zege” and “wuzi” per chapter. Obviously, in each system, a higher ranking supposedly indicates an earlier start of textual existence for that section. The two rankings do not tally with each other completely. Section Q, which boasts the highest ratio of earlier features to their later equivalents, ranks only number 8 on the right column, with 1.33 occurrences of “zege” and “wuzi” per chapter. Section O, which ranks toward the bottom of the right column with no instances of either “zege” or “wuzi,” has the eighth highest “earlier: later”

From Voice to Text

141

Table 14 Comparison of the two rankings of the sections E : L*

____________________

z & w+

______________________

Rank Section Ratio Section OPC‡ ________________________________________________________________________________ 1 Q 5.50 J 5.20 2 J 4.35 G 3.4 3 H 3.40 H 2.43 4 K 3.35 F 1.92 5 F 3.20 S 1.88 6 S 3.11 E 1.38 7 G 3.05 P 1.38 8 O 2.90 Q 1.33 9 W 2.62 U 1.20 10 N 2.61 A 1.00 11 P 2.59 K 0.90 12 L 2.36 T 0.88 13 R 2.31 C 0.75 14 M 2.24 M 0.67 15 V 2.20 N 0.57 16 T 2.13 R 0.36 17 C 2.05 V 0.22 18 U 1.90 W 0 19 E 1.88 L 0 20 I 1.88 I 0 21 D 1.52 D 0 22 A 1.44 O 0 23 B 1.34 B 0 ________________________________________________________________________________

*Early features: Later features. + zege and wuzi.

‡ Occurrences per chapter.

ratio. For sections A, L, and W, we also see a similar discrepancy between the two rankings. This is, however, by no means surprising. All these sections are short ones, consisting only of one or two chapters. Such statistical fields may not always be large enough, especially since “zege” and “wuzi” never appear in a degree of density comparable to that of “bian/jiu” and “jiang/qu.” The only two sections that may be called “aberrants” are sections E (Yang Zhi and the robbery of the birthday gift) and U (expedition against the Liao). Both are sufficiently long, yet each has its place in the two rankings notably apart. We can only speculate that this might have been caused by some unknown events during the textual evolution of these sections; for instance, a late spurt of exten-

142

Chapter 4

sive textual expansion could significantly lower the ratio of the earlier features to the later ones. Generally speaking, however, most of the sections have comparable rankings in the two columns in the table. Of all the seven sections with an “earlier: later” ratio above 3, six are among the top eight on the right column with highest frequencies of “zege” and “wuzi.” Of the sections with an “earlier: later” ratio below 3, sections N, P, R, M, V, T, C, I, D, and B are either ranked identically in the two columns (M, I, D, B) or have their respective places in the two columns reasonably close.

Stylistic and Linguistic Alluvium as a Result of Cumulative Textualization With the results of the two surveys for the most part corroborating each other, some of the tentative conclusions reached earlier are confirmed. It is indeed most likely that the Wu Song story (section G), the Song Jiang story (sections F and H), the story of Shi Xiu and Yang Xiong (section J), and the stories of the excursions of Li Kui and Yan Qing from Liangshan (section S) are among those sections of Shuihu zhuan that had a longer course of textual development. Among other major sections, the story of Lu Zhishen (section C) and the one of Lin Chong (section D) could have started their textual existence considerably later. Some shorter sections, such as the story of Wang Jin and Shi Jin (section B) and the story of Li Kui’s visit to his mother and killing of four tigers (section I), could also be later adjustments of the narrative. One cannot offer such an estimate for each part of the narrative, since a number of sections, as it turns out, feature inconsistent results in the surveys. Again, the linguistic features employed should be considered generally useful but not always specifically accurate. But obviously I am not trying to achieve the impossible feat of an accurate chronology of the textual evolution of Shuihu zhuan. Rather, my purpose has been to demonstrate that the fanben edition of the novel was the ultimate product of a long process of textual accumulation. That purpose, I think, has been reached. The discussion of the general tendency of vernacularization in a period historically overlapping with the evolution of the Shuihu tradition has provided a new perspective to approach the textual formation of the novel. Using the short stories as the main stylistic and linguistic context, this chapter has demonstrated that the text of Shuihu zhuan could have been “growing” in two different ways. The mixture of the “early period” stylistic features with the “middle period” ones in almost all chapters suggests that all parts of the novel went through a long process of textual evolution spanning from the early period to the middle period, based on the continuously reciprocal “oral-literary” transmissions. The orthographical variations of many of those stylistic markers corroborate that proposition. In

From Voice to Text

143

addition to that, there was a process of conglomeration of the Shuihu sagas from various sources into a lengthy narrative, and the convergence was achieved cumulatively and consecutively. The corollary was that different sections of the narrative started the process of textualization at different times, as evidenced by the striking disparity in the frequencies of some linguistic features and in the ratios between some earlier features and their later replacements.

5 The Engine of Narrative Making Audience, Storytellers, and Shuhui xiansheng In the foregoing chapter, the results of the philological analyses of the fanben text demonstrate a continuous deposition of stylistic and linguistic features from different periods. While we remain still ignorant of many things about the evolution of the Shuihu complex, we can now say one thing with a reasonable amount of certainty: The fanben text of Shuihu zhuan, which presents fullfledged vernacular prose, was “written” and repeatedly “rewritten” amid constant contacts with orality over a long time historically. Yet while the results of such analyses are obviously historicist in nature, the approach to the study of the stylistic and linguistic features is strictly formalistic in itself, confined to the domain of textuality and detached from the historical world where the narrative was produced, received, and textualized. If Shuihu zhuan evolved amid the interactions between writing and popular orality, some questions about the historical existence of the narrative are too imperative to be disregarded: What was the intended audience of the novel’s oral antecedents? How did that specific audience orientation affect the formation of the narrative? Who was responsible for the transmission and textualization? And what impact might that role in transmission and textualization have on the ultimate form of the work? Ironically, it will be necessary to resort to the text again, but only to use it in order to eventually go beyond it for a glimpse into the narrative’s historical mode of existence. I will avoid wild speculations about oral storytelling in premodern China, as most of the historical facts are irrevocably lost, although at some points I will not be able to steer completely clear of hypothetical propositions. The purpose of this chapter is to provide a new perspective for the interpretation of the

144

The Engine of Narrative Making

145

novel in terms of the relationships involved in the poiesis and sustentation of the tradition.

The Hermeneutic Relevance of “Audience Response” In his article “Text as Interpretation: Mark and After,” Walter Ong probes the orality of the original kerygma of Jesus from which the Gospel of Mark crystallized. By placing the Gospel into the context of the oral cycle about Jesus and by looking into the dynamics of both the oral cycle and the textualization, Ong is able to arrive at a new reading. He suggests that the tendency to base textual interpretation exclusively on the literate mentality, which he terms “literary exclusionism,” “can vitiate textual hermeneutic itself.”1 Of course, this “literary exclusionism” is not limited to the field of Gospel studies alone, for Ong goes on to state: Today’s vast literature on intertexuality and on interpretive communities has made it quite clear that texts come into being through interaction with other texts and are interpreted in traditions worked out by specific groups engaged with other texts. But intertextual analysis has commonly paid relatively little attention to the interaction between texts and their circumambient orality. The orality of a milieu can deeply affect both the composition of texts and their interpretation. Orality-literacy contrasts and interactions, we must remind ourselves, involve not merely different “channels” for units of “information” but different noetic worlds and different psychodynamics.2

To be sure, we have access to most oral forms only through the texts that represent them, as writing was the only way to preserve oral materials for centuries before the advent of tape-recording; but their mere textual form does not mean that all our interpretive activities must go along the text-reader axis, completely disregarding its pretextual origination. Indeed, these works, once put in textual form, are to be read—but merely to read them is far from enough for a full appreciation. “It is impossible to read ‘dramatic writing’ adequately,” as Raymond Williams exhorts us, “unless we are aware that it is writing for speech in many voices and for action. In its essential composition it is not a text for silent reading.”3 What Williams has said of dramatic writing can also be said of a narrative text that arose from an oral tradition. We cannot read it adequately unless we take into account the nuances and complications in the circumambient actuality of the original oral communication. One of the hallmarks of oral storytelling is the immediate partnership be-

146

Chapter 5

tween the teller and the audience. No teller can afford to lose his or her listeners, without whom the narrative activity would cease. No audience would just listen passively all the time: They would laugh, they would cry, and they would applaud. And if their patience runs out, they would simply turn away. A good storyteller would have to take hints from the audience’s immediate feedback and make on-the-spot adjustments. The story told is inevitably the outcome of such a negotiation between the audience and the storyteller, which goes on coevally with the development of the narrative itself. The teller is, therefore, never the sole creator of the story, and the audience is not only the addressee but, significantly, a coauthor in the story making. Writing, in contrast, has to go around a slackened cycle of consumption. “Books are objects. On a table, on bookshelves, in store windows, they wait for someone to come and deliver them from their materiality, from their immobility.”4 Until that fortuitous moment when a reader picks up the book and thumbs it open, the message from the writer reaches no one at all. The communication is deferred, and the text becomes, as Paul Ricoeur puts it, “in the air,” “outside the world or without a world.”5 The writer is, in Eugene Eoyang’s incisive words, “a solitary figure, remote, aloof, or reclusive, fending off admirers with exquisite Jamesian tact, or with crude Jonesian brusqueness.”6 This alienation of the writer from the reading public does not mean that he makes no effort to accommodate the reader. He surely does, as Wolfgang Iser clearly shows us.7 But once the writing appears in print, it is detached from the writer and becomes what Ong calls “autonomous discourse.”8 Thus, when the female readers in eighteenth-century England found the death of the beautiful and virtuous Clarissa too much for their tender sensibilities and wanted her to be revived, the writer, Samuel Richardson, was no longer in a position to do anything about it, even if he had wished. “To read a book,” again as Paul Ricoeur says, “is to consider its author as already dead and the book as posthumous. Indeed, it is when the author is dead that the relation to his book becomes complete and, in a way, intact.”9 In a sense, modern literary criticism hinges upon the detachability and the inevitable detachedness of the printed text from the author. The response is made to the text, precisely because the author, removed from the cycle of communication, is neither “responsive” nor “responsible.” But the text is not really responsive either. It never reacts to the reader’s responses, not altering a single phrase of its discourse. It is like a toy tumbler: A reader’s interpretation may bend it in this or that direction, but it promptly “tumbles” over to its immutable form in print, waiting for another reading. To cause any change to a discourse in print, in its extreme cases, would call for the most violent form of “reader response”—namely, book burning. This is what happens in Cervantes’

The Engine of Narrative Making

147

fictional world in Don Quixote, where the curate and the barber commit the chivalric romances to the flames. Over two thousand years ago the first Chinese emperor, Qin Shihuang, did the same thing, in the real world and on a much larger scale, relegating thousands of books to ashes. Such dark moments were by no means rare in Chinese history. This is why the iconoclastic Ming philosopher Li Zhi, predicting the unfavorable reception of one of his books among the conservative literati readers, titled it The Book to Be Burnt ( Fen shu). For a written narrative, the reader’s response has always to be a “postdiscourse” activity. That is to say, it confronts a preexisting discourse and can never be early enough to participate in the process of discourse making. The audience’s response to an oral narrative, on the other hand, would be a contradiction in terms, for the making and reception of the narrative are synchronous and the audience never has a complete and concluded narrative to respond to. Because of the immediacy of the interaction between the teller and the audience, the oral delivery of a story is inevitably conditioned by the audience’s response at every moment of the storytelling. The interaction and negotiation between the teller and the audience, in a sense, become the narrative discourse itself. While the reader’s response is something largely external to a textual discourse, the audience’s role in the process of oral storytelling is internal, a constituent of the narrative discourse in addition to the role of the teller himself. Precisely because of this characteristic of oral narrative, the original audience’s impact can sometimes be traceable even after the narrative has long survived its oral stage of evolution. Folklorists have noticed that the meaning of an oral narrative does not reside so much in what the teller tells as in what the listener hears. They are interested therefore in collecting the audience’s responses to an oral narrative they have received. Alan Dundes, for instance, urges folklorists to expand the texts of oral narratives they collect to include the “oral literary criticism”— that is, comments from the listeners.10 Such comments may be of tremendous value for interpretation, but they, like reader’s comments, are “postdiscourse” responses that have no impact on the discourse-making process. Another pitfall is that the folklorist’s own line of questioning in soliciting comments is inevitably dictated by his or her own presuppositions. Consequently, the audience’s responses thus collected may turn out to be largely the folklorist’s own, as Sandra Dolby Stahl points out to us.11 In dealing with oral or oral-derived narrative texts from a past age, what is more relevant is not such comments from the audience, most of which have long been buried in oblivion, but the possible effects of the audience’s instant reaction and even the audience’s physical presence itself on the process of story making. Hermeneutically, the narrative’s pretextual existence, with its inception and reception, is of vital im-

148

Chapter 5

portance and should not be dismissed lightly. At this juncture, Eugene Eoyang’s following comment on the interpretation of the Tang bianwen is perfectly pertinent here: As historians of literature, we cannot overlook the oral origins of literary fictions. And the role of the audience—whether present or posterior—must be recognized once again. It is an understanding of this audience, actual or tacit, that is crucial to a clear understanding of the modes of fiction. Only with such an understanding can one begin to appreciate that a work is as good as its audience allows it to be.12

The Role of the Urban Audience in Chinese Oral Tradition On the literacy rate in traditional China, scholars have come up with radically differing estimates. The famous Qing palace tutor Hu Xu (1655–1736) lamented that during the first half of the eighteenth century all the nation’s schools reached only about 2 to 3 percent of the entire population.13 In one of his early studies, John DeFrancis puts the literacy rate “through most of Chinese history” at a dismal “one or two percent.”14 The same figure was accepted by Richard Solomon as a “rough estimate of the literate segment of the population between 1600 and 1900.”15 In contrast, some other estimates are distinctly more optimistic. Rejecting Solomon’s figure as groundless, Frederick W. Mote has argued that a literacy rate at 10 percent or more is “virtually inescapable” during that period.16 Barbara E. Ward has suggested that the proportion of the Chinese males who possessed some degree of literacy could be as high as 20 percent through most of the past thousand years, but her conjecture is based on the premise that a “fair degree of literacy” would require knowing only about one thousand characters.17 The figure for literacy could be even higher according to Evelyn Sakakida Rawski, who has estimated that in nineteenth-century China, 30 to 40 percent of the men and 2 to 10 percent of the women knew how to read and write. 18 She is referring to the final stage of imperial China, when the literacy rate following the expansion of education could have been considerably higher than in earlier times. Without a unified standard for literacy in the Chinese context, the lack of consensus among scholars is by no means surprising. Whatever the estimate, it should be noted that, before the advent of vernacular literature in print, to be able to read did not necessarily mean to be able to read literature. It seems safe to say that the overwhelming majority of people were not readers of “literature,” which was written, almost by definition, in wenyan. This, however, did not mean that they were excluded completely from the nation’s literary

The Engine of Narrative Making

149

arena. “It would be a mistake,” as Barbara Ward justly points out, “to see the literature of China as the property only of those who could read. On the contrary, much of the material from the histories, legends, novels, and stories was constantly brought before the wider public—even in the villages—through the medium of the performing arts of storytelling and the popular theater.”19 It was these pervasive activities, more than reading and writing, that were responsible for the inexorable cultural dissemination in a country with a vast territory and an enormous population. The preface to Dong-Xi Han tongsu yanyi (A Popular History of the Eastern and Western Han Dynasties), attributed to the noted late Ming scholar Yuan Hongdao (1568–1610), testifies to the effectiveness of the oral transmission of historical information: Nowadays, from the gentry down to rustic villagers and from septuagenarians down to children, when people are talking about events such as Liu Bang’s uprising at Fengpei, Xiang Yu’s suicide by the Wujiang River, Wang Mang’s usurpation of the throne, and the emperor Guangwu’s restoration of the dynasty, they can tell all the details from the beginning to the end including the names and places. They would get together talking strenuously, from morning till evening and from dusk till dawn, almost forgetting about eating and sleeping. But if you show them the Hanshu [Book of the Han] or the Hanshi [History of the Han], they do not understand them. Even if they try, they are not likely to understand them adequately.20

If one takes into consideration both the relatively small number of people who could read wenyan and the amazing degree of popularization of these historical narratives, one has to agree that among the common people cultural information was conveyed mostly through the channels of oral and performative media.21 “In this world,” as Feng Menglong so succinctly puts it, “cultivated minds are few while coarse ears are many” (Tianxia zhi wen xin shao er li er duo).22 If that was the situation during Feng Menglong’s time in the seventeenth century, when literary readership was significantly enlarged by the increasing publications of vernacular literature—including his own San yan and different editions of Shuihu zhuan—it must have been even more so in the fourteenth and fifteenth centuries, when the oral Shuihu complex was evolving. Historically, the ascendance of professional storytelling in China was closely associated with the rise of urban centers, and the evolution of the Shuihu complex coincided with the period of rapid urbanization and commercialization.23 In the Southern Song period, partly due to the massive immigration of northern Chinese fleeing the invading Jurchen forces and partly because of the

150

Chapter 5

economic prosperity in the abundant Jiangnan area, the capital of Lin’an (Hangzhou) quickly developed into a metropolis with possibly one million households.24 The concentration of population led to further commercialization of the urban areas, which enjoyed a level of prosperity unprecedented in Chinese history.25 During the Yuan the city of Hangzhou continued to boom. In Marco Polo’s description of the Yuan China, his most enthusiastic words were reserved for Kinsai (Hangzhou), which he claims to be “without doubt the finest and most splendid city in the world”: [I]t was stated that the city of Kinsai is about 100 miles in circumference, because its streets and watercourses are wide and spacious. Then there are market-places, which because of the multitudes that throng them must be very large and spacious. . . . There are ten principal market-places, not to speak of innumerable local ones. . . . All the ten squares are surrounded by high buildings, and below these are shops in which every sort of craft is practiced and every sort of luxury is on sale, including spices, gems, and pearls. In some shops nothing is sold but spiced rice wine, which is being made all the time, fresh and fresh, and very cheap.26

While Hangzhou was the predominant and unrivaled metropolis in the Southern Song, the Yuan and the following Ming period saw the rise of many other cities, especially in the Jiangnan area. Hangzhou, as Michael Marmé puts it, was reduced “from a blazing sun to the status of another planet.” We may take Suzhou as an example for those rising “planets” that eroded some of Hangzhou’s splendor. In the short span of fifteen years from 1275 to 1290, Suzhou saw a dramatic growth in its population from 329,603 to 466,158.27 About the city, Marco Polo wrote: The people here . . . live by trade and industry, have silk in great quantity and make much silken cloth for their clothing. There are merchants here of great wealth and consequence. The city is so large that it measures about forty miles in circumference. It has so many inhabitants that no one could reckon their number.28

During the early Ming, Suzhou survived Zhu Yuanzhang’s exorbitant taxation as the emperor’s punishment on the city, which had been the base for Zhang Shicheng, Zhu’s main rival. By 1488, Suzhou had grown into such a commercial hub that the Korean official, Ch’oe Pu, exclaimed that at the port of Suzhou, “merchant ships hailing from Hubei and Fujian gather like clouds.”29

The Engine of Narrative Making

151

The Chinese urban milieu, as Jaroslav Prusek has told us, was quite comparable to the environment in Europe during the thirteenth and fourteenth centuries, which gave rise to the vernacular literatures.30 While the emergence of a bourgeois class in European cities prepared the social conditions for such vernacular works as Decameron and Canterbury Tales, the oral Shuihu complex may have evolved under similar social conditions. The inhabitants in the cities were necessarily motley, but the largest and most stable sector of the urban population was constituted of merchants, shop assistants, craftsmen, and their families. Most of them were unable to read literature in the classical language, but, living at the economic and cultural centers of the nation, they wanted and could afford to treat themselves to some forms of entertainment. As the urban communities were much more compact than those in rural areas, people could easily congregate in sizable audiences. Naturally, storytelling and other forms of public entertainment became popular mainly in the cities and towns. According to Xihu Laoren, there were in the city of Hangzhou five major centers for public entertainment, places called wazi or washe. Within each wazi there were several odeums called goulan. The largest of the five wazi, the North-wa, had as many as thirteen goulan in it, with two of them specially allocated for jiangshi performances. In addition there were twenty wazi in the suburbs.31 The record in Meng Liang lu is slightly different, which puts the total number for wazi in the city and the suburbs at seventeen.32 In Zhou Mi’s Wulin jiushi, which was probably composed in the early Yuan, names of twenty-three wazi in and out of the city are listed. Like Xihu Laoren, Zhou Mi also mentions the North-wa as the largest, containing thirteen goulan.33 In his Chuo geng lu, Tao Zongyi mentions Hu Zhongbin as a storyteller during the Zhizheng period (1341–1368) who “told unofficial histories in the goulan of Hangzhou,” and his younger sister could be his fellow raconteur.34 During the Yuan period, the concentration of storytelling and popular drama in the urban centers might also be related to the government’s crackdown on unlicensed public performers in the rural areas. It is recorded in Yuan dianzhang that two farmers in Shulu County were arrested for gathering about one hundred people putting on public performances (zi ban cizhuan). The document proclaims that, except for professional raconteurs (zhengse yueren), nobody was allowed to practice storytelling or dramatic performance.35 Li Dou’s (fl. 1797) Yangzhou huafang lu describes all forms of entertainment in the pleasure district of Hongqiao in eighteenth-century Yangzhou.36 Several famous storytellers’ names are mentioned, including Wu Tianxu specializing in Sanguo, Xu Guangru in East Han cycles, and Wang Deshan in Shuihu.37 Of course the urban storytelling described by Li Dou was considerably later

152

Chapter 5

than the formative period of Shuihu zhuan; but, given the steady continuity of the Chinese tradition of storytelling, it is not irrelevant to our interest here. About the audience enchantment with the storytelling of the shuohuaren (professional raconteurs),38 history has left us numerous accounts. The Song poet Su Dongpo (1037–1101) tells us how parents would put an end to their children’s mischief by sending them to listen to stories. While listening to the tales from the cycles of the Three Kingdoms, they were sad to hear of Liu Bei’s mishap and overjoyed to hear of Cao Cao’s defeat.39 From Feng Menglong we hear a similar anecdote: A child cut his finger in the kitchen but did not groan. As people were surprised, he simply replied: “I just went to hear the story of the Three Kingdoms. When Guan Yu had his bones scraped by the doctor, he was chatting and laughing as if nothing had happened. How should I feel pain at this!”40 While such accounts tell about the spellbinding power of oral stories over those young listeners, one must not be led to think that storytelling was meant to be mere child’s play. Elsewhere, Feng Menglong dwells on the effectiveness of storytelling in enhancing social values among the audience, and it is said to have the power to make “the timid ones courageous, the licentious chaste, the frivolous sincere, and the insensate perspiring.” Significantly, Feng proceeds to compare the didactic persuasiveness of oral storytelling favorably with that of the Confucian classics: “Even the reading of the Xiaojing [Book of Filial Piety] or the Lunyu [ The Analects] can hardly produce an impact so prompt and so profound.”41 Storytelling certainly owed this “prompt and profound” impact on audiences to its immediate association with social life. The shuohuaren’s ways of making and telling a story were instantly ratified, censured, or rejected in interaction with the audience’s tastes and interests. The actual delivery of the story had to be a sort of homeostatic balance between the shuohuaren and the audience, based on the continuous adjustments on the part of the raconteur to accommodate the audience’s expectations. The conversion from the pro-Wei standing in the semiofficial history Sanguo zhi written in the Jin period to the pro-Shu stance in the popular Three-Kingdom complex during the Song and Yuan periods was well in accord with the changed public sentiment that regarded Shu instead of Wei as the only legitimate regime among the three states.42 After many rounds of negotiation between the storytellers and the audiences, what was told increasingly coincided with what the public expected to hear. Many other features in different Chinese oral traditions have been convincingly explained in terms of storytellers’ attempts to accommodate the interest of a particular audience. Eugene Eoyang, for example, has informed us that the Wang Zhaojun story that developed in China’s northwestern frontier

The Engine of Narrative Making

153

area with an ethnically motley population is significantly different from the Wang Zhaojun story that circulated among the Han people. In the former, the chieftain of the northwestern tribes of Xionggnu, to whom the beautiful concubine of the Han emperor was married as a surety of peace, is portrayed with much greater sympathy.43 A similar case has been brought up by Ma Nancun (Deng Tuo) in his “Sanzhong Zhuge Liang” (Three Types of Zhuge Liang). One portion of the Sanguo cycle tells about the war between the forces of the Shu commanded by Zhuge Liang and the southern aborigines with Meng Huo as their chieftain. According to the story, Zhuge Liang seven times captured Meng Huo and seven times set him free, in order to appease the southern tribes. Ma Nancun tells us that in some part of the southern province of Yunnan inhabited by ethnic groups other than the Han, the story has been reshaped: Zhuge Liang and Meng Huo have exchanged their positions, with Meng Huo as the hero, seven times capturing Zhuge Liang and seven times setting him free.44 While we are blessed with numerous historical accounts of the typology of storytelling, very few detailed descriptions of storytelling scenes can be found in premodern China. One such scene occurs in chapter 51 of Shuihu zhuan, featuring a woman raconteur, Bai Xiuying, who comes from the capital Bianliang to perform in the county town Yuncheng to be close to her lover, the new county magistrate. She specializes in zhuban pindiao, which, as evidenced by the title of her story and the formal characteristics of the performance, might simply be another term for zhugongdiao.45 The woman starts the performance by reciting the verses: “When the baby bird chirps the mother bird returns; / And while the old sheep is bony the lamb becomes chubby.” These lines about self-sacrificing parental love are so touching to the audience that her voice, as Jin Shengtan puts it in his commentary, “penetrates deep into everybody’s ears.”46 Lei Heng, in particular, cheers loudly on hearing these lines, because he is a devoted son living with an old mother. After a few rounds of “singing and telling,” the cheers bring the house down. At the height of the audience’s excitement, Bai Xiuying and her father begin to collect fees from the crowd. Another storytelling scene is to be found in chapter 90. When Li Kui and Yan Qing get into the city of Bianliang to watch the lantern show in the Lantern Festival, they go to a crowded goulan where a raconteur is telling a story of Guan Yu, a key figure of the Three-Kingdom cycle. Li Kui is so carried away on hearing of Guan Yu’s heroism that he shouts out his admiration loudly. In chapter 10 of Shuo Yue quanzhuan, a vernacular narrative about the Southern Song general Yue Fei, we see something very similar. Following two strangers, Niu Gao, one of Yue Fei’s officers, enters an odeum in town where pinghua stories are being told. The storyteller pauses to collect fees after fin-

154

Chapter 5

ishing a session on Yang jia jiang, a cycle about the Northern Song general Yang Ye and his sons. One of the two men gives the storyteller two silver ingots, as he claims that the story is actually about his forefathers. After that Niu Gao follows the two men to another odeum, where the storyteller is telling a story of Luo Cheng, a general in the early Tang period. When the storyteller stops at a critical conjuncture of the plot to collect fees, the other of the two men offers four silver ingots, as he claims that that story is about his ancestors.47 Such episodes are, of course, fictional, but since they are found in narratives that evolved from oral story cycles, they may have been storytelling scenes in stories once told by storytellers and therefore might well be a mirror reflection of the real storytelling situation. Indeed, the dynamic relationship between the storyteller and the urban audience as we see it in such fictional scenes is quite consistent with what we have learned of storytelling in modern China. Based on his field observation of the contemporary Kunming pingshu, Mark Bender tells us that listeners would comment directly to the storyteller before and after the performance and even during the intermissions. This firsthand experience enabled Bender to affirm that “the old storytelling maxim was true: a performer must please his audience or risk not having one.”48 Vibeke Børdahl also reports that among the modern storytellers in Yangzhou, the term “peng da wan” (carry away the big bowl) is often used to warn each other that a blunder in their performance could possibly lead to the audience’s drastic response of confiscating the “big bowl” containing the storyteller’s fees.49

The Urban Audience as Collaborator in Story Making for the Shuihu Cycles As mentioned earlier, Song Jiang, the protagonist in Shuihu zhuan, could have been initially based on a historical figure of that name, the leader of a rebellion toward the end of the Northern Song period. According to Nie Gannu, the earliest stage of the oral complex of Song Jiang stories probably did not deviate too far from the historical records, which depicted Song Jiang and his men as bandits roving about in the vast marshland. This was followed by another stage that centered the activities of Song Jiang and his men in the Taihang Mountains.50 A casual comparison between this and what happens in Shuihu zhuan will indicate that, as a result of the many rounds of interaction between the tellers and the audiences, the major arena for Song Jiang and his cohorts had largely been shifted from rural areas to cities and towns. That the early stage of the oral complex had the bandits based in the Taihang Mountains is evidenced by Gong Shengyu’s “Song Jiang sanshiliu zan.” In the panegyrics on Lu Junyi, Yan Qing, Zhang Heng, Dai Zong, and Mu Hong, Gong Shengyu mentions Taihangshan five times, while the name of Liangshanbo

The Engine of Narrative Making

155

never appears.51 Xuanhe yishi may mark the start of the shift of the locale from the Taihang Mountains to Liangshanbo, for the two places, hundreds of miles apart from each other, are mentioned side-by-side in the curious combination “Taihangshan-Liangshanbo” as the base of the rebellion.52 This is corroborated by some vestiges in the fanben text that may have been inherited from an earlier stage of the story cycles. In chapter 16 of the Rongyutang edition, for instance, there is a verse stanza describing the steep Yellow Earth Ridge, which Yang Zhi and his men, commissioned to escort the birthday gifts to the Prime Minister Cai Jing, have to pass on their way to Bianliang. The last two lines of the stanza read: Don’t you say the paths to West Sichuan are dangerous, Be sure to remember these are the Taihang Mountains. (1: 479)

The Yellow Earth Ridge is said to be in Shandong, under the jurisdiction of Jizhou Prefecture, and therefore hundreds of miles away from the Taihang Mountains. As the setting of the story cycles had changed, the verse line from the earlier time became a “textual fossil” and resulted in this geographical discrepancy. These speculations on the urban audiences for the oral Shuihu complex and the storytellers’ efforts to accommodate the taste and interest of such audiences seem quite consistent with what we know from different sources about popular storytelling from the Southern Song to the early Ming—especially the scorn and disdain shown by both the performers and audiences for country people. During the Southern Song, there was a particular dramatic genre known as zaban or zawang, which specialized in teasing country people, who were ignorant of the sophistication in the cities.53 The sanqu “Zhuangjia bushi goulan” (A Country Bumpkin Knows Nothing of the Theater), by the early Yuan writer Du Shanfu (1202–1283), makes fun of a farmer’s foolery during a session of performance in a goulan: He comes shopping in the city, is attracted into the jolly and boisterous goulan, but has to leave at the height of the performance for an urgent call of nature.54 In chapter 51 of Shuihu zhuan, when Lei Heng is humiliated by the woman raconteur Bai Xiuying for not offering any fees, her father ridicules Lei by saying sarcastically to his daughter: “Have you no eyes, daughter? Can’t you tell the difference between a city gentleman and a rube?” (3: 1676). Such deprecation of the country people is also seen in the other storytelling scene in chapter 90. When Li Kui gets overly excited on hearing the Guan Yu story and shouts out loudly, Yan Qing reprimands him by saying, “How can you be so boorish, brother?” (Ni zendi hao cun) (5: 2911). The epithet “cun,” originally

156

Chapter 5

referring to a village or a farm, is here associated with being “ill-mannered” or “unsophisticated.” A similar attitude toward country people is reflected in the storytelling scene in Shuo Yue quanzhuan. A man in the audience complains against his companion for overpaying his fees to the storyteller: “Those two silver ingots were nothing for you, brother. But when people in the city know it, they would take you for a hick clown” ( Zhishuo dage shi xiangxiaren).55 Of course, this is not to suggest that country people were excluded from storytelling and other public performances. There might be some oral genres that specially targeted the rural people as an audience, as suggested by the line that “The listeners of taozhen were almost exclusively villagers” ( Ting taozhen jinshi cunren).56 But it seems safe to maintain that the audiences of storytelling, especially those genres under the rubric of shuohua, were predominantly urban. From the very beginning, Shuihu zhuan exhibits an almost irrepressible fervor in its descriptions of the urban scenes and life of the urban inhabitants. The narrative is scarcely started when it tells the story of Gao Qiu as a scamp on the streets of Bianliang who later succeeds in ingratiating himself into the future emperor’s favor. In the first twelve chapters, the city of Bianliang provides the setting for a large portion of the narrative action. The three major characters in this part of the narrative—Lu Zhishen, Lin Chong, and Yang Zhi—all live in the middle of the hustle and bustle of the capital before they are finally forced to leave the city. Throughout the rest of the work, Bianliang continues to be a major locale for the narrative events. Both Yang Zhi and Wu Song are put on missions to escort gifts to Bianliang (chapters 16 and 24), and Wu Song even rejoices that the task will give him the opportunity to have some sightseeing in the capital that he has long wished to visit (2: 745). In chapter 72, Song Jiang, who has grown up in Shandong and has never before been to Bianliang, even slips into the capital with a few of his men to watch the lantern show. From the perspective of these intruders who stroll along the streets, the narrative gives an exuberant description of the marketplaces, with “many taverns and teahouses, patronized by people richly garbed in silks and satins, each in a distinctive color” (3: 2337). In addition, it devotes two massive stanzas of verse, which occur near each other in the text, to the glory and prosperity of “the Number One Metropolis under heaven” (tianxia diyi guodu) (3: 2343). Besides Bianliang, the Liangshan heroes are also put in the urban milieu of a good number of other cities and towns. Among them are large cities, such as Beijing Damingfu (chapters 13, 61, 62, and 66) and Nanjing Jiankangfu (chapter 74), and medium-sized and small cities like Dongpingfu (chapters 27, 69), Weizhou (chapter 3), Tai’anzhou (chapter 74), Jizhou (chapters 45, 53), Dengzhou (chapter 49), Mengzhou (chapter 29), Gaotangzhou (chapter 52),

The Engine of Narrative Making

157

Cangzhou (chapters 9, 51), Yunchengxian (chapters 18, 20, 51), Yangguxian (chapters 22–26), Qingfengzhen (chapter 33), and others. It is impractical to look into the narrative’s depiction of all these urban centers, but we can take Yangguxian—which serves as the setting for Wu Song’s story of slaying his adulterous sister-in-law—as an example. Yangguxian is only a county seat, probably a small city, but the residents there constitute a typical urban community. While Wu Song serves as a constable in the police force, his brother Wu the Elder is a peddler selling buns on the street. The latter’s next-door neighbor, Old Woman Wang, runs a teahouse. With the old woman as go-between, Wu the Elder’s wife, Pan Jinlian, takes a lover named Ximen Qing, who is the owner of an apothecary (chapter 24). When Wu Song tries to kill the procuress and the adulteress to avenge his murdered brother, he invites his brother’s neighbors to a dinner at which they are to hear the women’s confessions. Among these neighbors are a silversmith, a man who owns a store that sells paper horses for funeral burning, another man from across the street who sells chilled wine, and finally an old man who peddles noodles—all typical small merchants (chapter 26). Another typical urban center in Wu Song’s story is Mengzhou, to which Wu Song is exiled after his bloody revenge for his brother. As a destination for exiles, Mengzhou is supposed to be an out-of-the-way place, but it turns out to be a booming town. In Mengzhou, Wu Song beats the bully Jiang the DoorGod and restores Shi En’s ownership to his tavern at Happy Grove, a market center in the suburbs of the city. Shi En’s account of what has happened to his tavern testifies to the prosperity of Happy Grove: “Outside the East Gate of this town is a marketplace called Happy Grove. Merchants from Hebei and Shandong all go there to do business. There are dozens of large inns, and twenty or thirty gambling houses and money shops. Since I was a skilled fighter and had eighty or ninety tough ruffians in the prison, I opened a tavern there. All the stores, gambling houses, and money shops had to pay me regular tribute. Even singing girls, upon arriving, had to see me first before they could ply their trade. Money kept coming in every day from so many sources. At the end of the month I was usually able to collect two or three hundred ounces of silver.” (Chapter 29; 2: 909–910)

The almost obsessive adherence to the urban setting in the narrative can indeed be best explained in terms of the storytellers’ efforts to accommodate the interests of an audience constituted mainly by urban inhabitants. The audience’s tastes and expectations, through the dynamics of the audience-teller

158

Chapter 5

interaction, may have dictated the criteria for narratability in the process of story making. The urban milieu and commercial activities had a central position in the narrative, not necessarily because they were particularly palatable to the raconteurs (of course, a raconteur might himself be an urban resident as well), but more likely because they were most appealing to the audience. Even in those parts where the action is put in a rural setting, there is a tendency to “urbanize” the milieu. Throughout the whole narrative, we can hardly find any description of peasants at their farm work, which logically should have been the most common scene in the countryside. Instead, rural people in the narrative are often those who live in manors, where the squire, his family, and his vassals spend most of their time playing with weapons, as in the manor of the Shi family (chapter 2), Chai Jin’s manor (chapter 9), Chao Gai’s manor (chapter 14), the manor of the Kong family (chapter 32), Li Ying’s manor, and the manors of the Hu and Zhu families (chapters 47–50). Since there is nothing really rural about these manors, they are not truly distinct from large residences in the cities, such as Lu Junyi’s house in Beijing Damingfu. In other instances, characters that are supposed to be rural people are fully engaged in commercial occupations. At the foot of Mount Wutai (Wutaishan), where Lu Zhishen stays in the monastery as a monk, there is “a market town of six or seven hundred households. . . . There are stores selling meat, vegetables, wine, and noodles” (chapter 4; 1: 137). Next to the building with the sign “Father and Son Inn,” Lu Zhishen sees an ironsmith’s shop, from where the clang of metal can be heard over a distance. Again, only “thirty paces” from the ironsmith’s shop, Lu finds himself at the door of a tavern “with a banner sticking out from the eaves of the house” (chapter 4; 1: 139). Of course no one would assert that such a bustling market town at the foot of the mountain was totally impossible, but the selection of narrative materials is certainly meaningful: The commercial activities that must have constituted only a peripheral sector of rural life were granted a central position in the narrative. With the rural life somehow urbanized, the world in Shuihu zhuan almost becomes a world of taverns. One can almost say that there are only two kinds of people among the bandit heroes—tavern owners and tavern customers— for we can hardly think of a chieftain of the band who is not at least once involved in a tavern scene. Cao Zheng (chapter 17), Zhang Qing and Sun Erniang (chapter 27), Li Li (chapter 36), and Sun Xin and Mistress Gu (Gu Dasao) (chapter 49) are all from rural areas and are all tavern owners. And when Song Jiang needs lookout posts around Liangshan, he sets up four taverns, one on each side of the bandits’ lair (chapter 70). But many times more numerous are those taverns that are not owned by the Liangshan people, distributed all over the rural world. These taverns, even more so than the manors, are the places

The Engine of Narrative Making

159

where a Liangshan hero meets his comrade(s) or where the narrative action takes a turn. Therefore, when Wu Yong wants to talk to the Ruan brothers, they go to drink in a tavern, selecting a room with “a red-lacquered table with red-lacquered benches” (chapter 15; 1: 441), even though the Ruan brothers live in an impoverished fishing village least likely to boast such a modish establishment. While the descriptions of the physical settings of the narrative action are made fundamentally from the point of view of the urban inhabitants, it is perhaps even more so with the depiction of the characters. Starting with Gao Qiu, most of the villains in the narrative are the scum of urban society. Gao Qiu is a powerful and evil official in the emperor’s favor; he is, in many ways, representative of the seamy side of the bureaucracy. Significantly, the narrative does not initially attack him as a bad official but, instead, exposes his disgraceful and despicable origins. If the dignitary is depicted as having been a young scamp who spent his time gallivanting about the city and dissipating his money in gambling dens and brothels, eventually to be banished from the city for three years (chapter 2), a moral superiority could be easily developed for the audience. With them, poetic justice could take an interesting twist: The detestable big shot is reducible to a small potato, not only brought down to our own level but in moral terms even well beneath ourselves. Like Gao Qiu in his early days, many of the bad guys in the narrative are those bullies, rogues, and scoundrels in an urban community who are hated by the masses of residents. Some of them are unrestrained evildoers who are strengthened by their powerful connections. Gao Qiu’s adopted son, for instance, is the special scourge of the streets of Bianliang, whose favorite pastime is “despoiling other men’s wives” (chapter 7; 1: 228). Another high official’s brother-in-law, Yin Tianxi, who “hurts people with his outrageous behavior” in the city of Gaotangzhou, covets the garden of Chai Jin’s uncle and seizes it by force (chapter 52). And Jiang the Door-God, with the backing of Zhang the garrison commandant, forcibly takes over Shi En’s tavern in Happy Grove (chapter 29). Others are simply unscrupulous local ruffians, such as the Butcher Zheng in Weizhou, who cheats a singing girl into becoming his concubine and afterward blackmails her and her father (chapter 3), and the notorious rowdy and troublemaker Niu Er, known throughout the city of Bianliang as the Hairless Tiger, who pesters Yang Zhi into killing him (chapter 12). And of course there is the licentious and debauched Ximen Qing, owner of a medicine store in the town of Yangguxian. Another adulterer in the narrative, Pei Ruhai, is a monk, but, significantly, “he used to run his family’s silk-thread shop before he joined the order” (chapter 45; 3: 1474), and he is therefore clearly another degenerate of the urban merchant class. It was in the interest of the

160

Chapter 5

urban communities to have such a flock of street vermin eliminated. They would acclaim the heroes in Shuihu zhuan, such as Lin Chong, Lu Zhishen, Yang Zhi, Wu Song, Li Kui, Yang Xiong, and Shi Xiu, because they were precisely the “eliminators” that their urban communities needed. One of the qualities of the Liangshan heroes that is given a privileged place in the narrative is their financial generosity in helping the needy. As Ouyang Jian and Xiao Xiangkai have noted, it is those who are generous in offering monetary help, rather than those with superior military prowess, that become the leaders of the Liangshan band.57 Although the quality of generosity is attached to many of the heroes, it is Song Jiang and Chai Jin who are accorded the highest admiration. When Song Jiang makes his first appearance in the narrative, he is described as a benefactor: He loved to make friends in the gallant fraternity. Anyone, high or low, came for his aid, he would provide his guest with food and lodging in the family manor, tirelessly keeping him company, and giving him traveling expenses when he wanted to leave. Indeed he scattered gold about like dust! He never refused a request for money. He was always making things easy for people, solving their difficulties, settling differences, saving lives. He provided the indigent with funds for coffins and medicines, gave charity to the poor, assisted in emergencies, and helped in cases of hardship. (Chapter 18; 1: 536)

And Chai Jin, descendant of an imperial lineage, is similarly depicted as a benevolent patron. The keeper of the tavern at Chai Jin’s village speaks of his master’s generosity with admiration: “Chai Jin always welcomes bold men from different corners of the world, supporting forty or fifty of them in his home. He’s often instructed us, ‘If any prisoner en route to exile comes to the tavern, just tell him to come to my manor. I will help him with money’ ” (chapter 9; 1: 281). The reputation of both men, Song Jiang and Chai jin, as men of generosity who repeatedly help others with money, becomes one of the major reasons that the Liangshan rebels win over the feelings of both the jianghu haohan (bold men in the gallant fraternity) and a good number of officers in the imperial troops. One recalls the recurring scene in the narrative where a hostile warrior or a general sent by the court kneels down and kowtows to Song Jiang on hearing that Song is exactly the one with the nickname “Timely Rain” ( Jishi yu), which means, figuratively, “one who gives money when money is most needed.” The elaboration on financial generosity may seem indulgent or even out of place in a narrative about a group of bandits who kill and pillage, but we will

The Engine of Narrative Making

161

understand it better if we take into account the role of the urban audience whose daily activities were in trade and commerce. Money is something that pervades the entire narrative; in addition to the heroes’ practice of financial generosity, one thinks of the frequently recurrent scenes of monetary transactions, blackmail and looting for money, and bribery in gold or silver. This prominent role of money can be considered a projection of the audience’s real world, where interpersonal relations were measured increasingly in financial terms. In such economic conditions, personal loyalty between friends, or yiqi, began to assume an important financial significance, and it was the wish of the urban people engaged in commercial business to have resourceful financial allies who could offer money when money was needed. Characters like Song Jiang and Chai Jin, bounteous and accommodating, serve as models of men of yiqi in the audience’s finance-oriented urban world, where a tavern owner or a shopkeeper would be most willing to have a Song Jiang or a Chai Jin as a friend or a neighbor. While the Wu Song-type hero would punish the evil and protect the innocent, as mentioned earlier, the Song Jiang-type hero was needed as well in business and financial activities. Both types of heroes were created almost as two guardian gods for the urban communities, with the former as an upholder of justice and the latter as a warrant for financial security.58 Thus both the setting for the narrative events and the molding of the characters in Shuihu zhuan could have been to some extent influenced by the interests of the urban audiences. C. T. Hsia has noticed in Shuihu zhuan a “curious contrast between the professional polish of several well-developed sagas and the recurrence of certain themes and situations that seem to have resulted from the storyteller’s compliance with the public’s inveterate demand for the familiar and stereotyped.”59 Apart from the indiscriminate attribution of what is “well-developed” to “professional polish” and the underlying prejudice against public tastes, Hsia’s description of the dialogical teller-audience relationship as a shaping force on the narrative is precise. In chapter 3, I covered the recurrent scenes and action patterns in Shuihu zhuan in terms of the oral mode of composition and story making. Here is an additional possibility: Those recurrent elements may also have derived from the needs of the audience. The heavy reliance on the urban settings, especially the tavern scenes, and the molding of human figures into certain urban types can indeed be associated with the audience’s “inveterate demand for the familiar and stereotyped.”

Men of Letters amid Storytellers If Shuihu zhuan was the product of the process of interaction and mutual infiltration between the forces of writing and those of popular orality, such a process did not take place as a matter of course. Traditional Chinese society

162

Chapter 5

was strictly stratified: The demarcation between the fully lettered and cultured on one side and the illiterate and semiliterate on the other was drawn almost equally well between the mental workers and the menial workers and between the ruler and the ruled. Mencius states this most clearly: “Those who labor with their minds rule others, and those who labor with physical strength are ruled by others. Those who are ruled sustain others, and those who rule are sustained by others. This is a principle universally recognized.”60 In a society where the privilege of the lettered was “a principle universally recognized,” the interfusion between the oral tradition and the written culture could be enhanced only under particular social circumstances. To a significant extent, those particular social circumstances were brought about by the system of civil service examinations, which contributed to the large-scale social mobility and reshuffling of cultural forces. As the number of examination candidates kept growing steadily, men of letters were constantly squeezed out of the excessively congested channel to officialdom, which facilitated the contact and confluence of the forces of writing with those of popular orality. The institution of the civil service examinations in traditional China was based on the assumption that a solid scholarship in the Confucian classics was essential for any government office. Under this institution, the ultimate goal of learning to read and write was a success in the examinations, which naturally determined the unchallenged dominance of the classics in the educational curricula. This means that, in the educational system with the civil service examinations as virtually the sole purpose, any lettered person was inevitably trained in the classical tradition. To be sure, the civil service examinations were not the only way for selecting officials in Chinese history. During the Han Dynasty (206 b.c.–a.d. 220), the state’s recruitment of government officials was based on a recommendation system, with examination as an auxiliary method for classifying candidates who had already been recommended.61 Theoretically, the recommendation system was supposed to introduce candidates into officialdom on the basis of moral as well as intellectual merits, but—in the absence of any objective standard for judgment—the system was eventually reduced to a practice of favoritism, or “a useful tool by which the powerful clans could perpetuate themselves.”62 During the Sui and the following Tang periods, the need for establishing a more effective way of selecting government officials was increasingly felt. Thus the civil service examination system was instituted, and it was maintained right through the rest of China’s dynastic history down to 1905. The recommendation system, however, was never officially abolished. Instead, it continued to play a considerable but intermittent role.63 There was, in addition, the practice of selling and purchasing degrees and official positions,

The Engine of Narrative Making

163

which was, however, “generally held within bounds” during the Ming period.64 Indisputably, the examination system remained for most of the time the major doorway to officialdom, and indeed the only doorway for common scholars who had neither powerful connections to be “recommended” into an official position nor financial means to purchase one. If that door to success was completely shut during the first half of the Yuan when the Mongol rulers suspended the examinations, it was only slightly ajar in the periods before and after. The examinations had always been competitive on all three levels.65 After the Yuan government resumed the examinations in 1314, only seven jinshi examinations were held during the last sixty years of the dynasty, and each time the successful candidates—Mongols, Turks, and Chinese all included—numbered only between fifty and one hundred.66 At the lower levels of examination, the situation was probably even worse. Zhejiang Province, which had thirty-two counties ( jun) within its jurisdiction during the Yuan, was allocated a small quota of only twenty-eight licentiates. As happened at least once, Hangzhou did not produce a single licentiate out of the local examinations that were held once every three years.67 In the Ming period, when more public schools were established that were well integrated with the examination system,68 the competition in the examinations reached unprecedented intensity. The number of scholars drastically increased; the number of sheng yuan in Wuxi County, for example, went up from 62 in the year of 1424 to 249 in the year of 1572.69 As the openings of official positions remained relatively few, the Ming authorities had to impose stricter quotas on the number of degree conferees in order to control the size of the population of office candidates.70 Consequently, on all three levels, few candidates would be lucky enough to pass the examinations. According to Ju-k’ang T’ien’s estimation of the situation in the Ming and Qing periods, only one out of ten participants was able to pass the preliminary examination, one out of a hundred to pass the most competitive provincial examination, and one out of thirty to survive the examination on the highest level and acquire a jinshi degree. “Hence, the prospect for a commoner to gain the second degree of juren through the provincial examination was 1 to 1,000, the probability for a first degree sheng yuan to be successful in the final imperial selection was 1 to 3,000. Altogether, the normal chance for any beginner to get the highest degree was 1 to 30,000.”71 This, to be sure, does not mean that only one out of thirty thousand scholars could find a way into officialdom, for even a small number of selected holders of the preliminary degree could eventually become eligible for minor official appointments.72 Still, for the large majority of the scholars, many years’ labor on the printed words of the ancient sages did not raise them to glorious heights but only plunged them into bitter disillusionment.

164

Chapter 5

Among those who had failed in their examinations, many would refuse to reconcile themselves to the failure. “This mass of restless and unemployed scholars,” as Ju-k’ang T’ien has noted, “instead of leading their solitary lives, preferred to congregate and form various organizations.”73 They became engaged in various kinds of activities. Some of them took it upon themselves to report on the martyrdom of chaste women, for writing about other people’s greater misfortunes might provide a relief to their own tension and anxiety.74 Many others became active in popular culture. Historical sources such as Meng Liang lu and Wulin jiushi contain lists of the names or nicknames of some of the noted storytellers during the Southern Song period.75 Curiously enough, the nicknames of some storytellers, especially those in the genre of jiangshi, were formed by adding an academic title to the family name, such as Xu the Gongshi, Zhang the Jieyuan, Zhou the Jinshi, Dai the Scholar, and so on.76 They did not necessarily indicate the storytellers’ real status, of course, but in all likelihood they may suggest that those storytellers had formerly been scholars who were forced to give up their hope for an official career. By having themselves addressed as a jieyuan, a gongshi, or even a jinshi, these lettered men who had been frustrated in their examinations might be seeking a psychological compensation for a lost cause, a fantasy of the fulfillment of a dream that had been shattered in reality. Many other men of letters in the circle of storytelling, however, would stay in the background preparing or improving promptbooks or scripts for the raconteurs. From the Southern Song period to the early Ming, there were trade unions—or shuhui—in almost all the major urban centers for such men of letters.77 Members of the shuhui were called shuhui xiansheng ( gentlemen from the shuhui ) or cairen (men of talent). Again, the tone of veneration in such appellations suggests that they might be people well versed in the art of letters. Shuhui, shuhui xiansheng, and cairen were often used to refer to playwrights in the Southern Song, Yuan, and Ming periods. In Zhou Mi’s Wulin jiushi, six names are listed as people who belonged to shuhui.78 Several of the zaju playwrights listed in Lu gui bu are also said to be members of the shuhui. In the Tianyige edition of Lu gui bu, Xiao Dexiang is said in Jia Zhongming’s elegy to have “demonstrated his extraordinary talent in the Wulin Shuhui.”79 In Jia Zhongming’s elegy for Li Shizhong, also included in the Tianyige edition of Lu gui bu, Li Shizhong, Ma Zhiyuan, Hua Lilang, and “Hongzi” Li Er are all said to be fellow members of a Yuanzhen shuihui, who collaborated on the play Huangliang meng.80 In the three xiwen texts included in the Yongle dadian, Xiao Sun tu claims to have been composed by a shuhui in Hangzhou, Zhangxie zhuang yuan is attributed to a Jiu-san shuhui, and Huanmen zidi is said to have been recompiled by some cairen in Hangzhou.81 In the Chenghua edition of

The Engine of Narrative Making

165

Baitu ji, the play starts with the lines: “If you are not a hero you will not present a sword; / If you are not a cairen you will not compose a poem.” After that, the play accredits its composition to a Yongjia shuhui.82 Shuhui could be active in Zhu Youdun’s time as well. In one of his plays, Tao Yuanjing, the girl Taoer is said to have received her stage name Tao Yuanjing from the “old gentlemen of the shuhui in the town.”83 However, shuhui xiansheng and cairen were more likely terms for men of letters engaged in popular genres in general rather than for playwrights exclusively. In the story “Bai Niangzi yong zhen Leifengta,” for instance, the author is referred to as a cairen.84 In “Jiantie heshang,” the story ends with a short song marking the execution of the licentious monk, which is said to have been composed by “a shuhui xiansheng right on the site of execution.”85 Also, at the end of the cihua “Renzong renmu zhuan,” the text attributes its composition to the cairen. Indeed, the lines illustrate a concerted collaboration between writing and orality: “As men of talent (cairen) have composed this excellent cihua/ Noble and virtuous gentlemen must listen to it.”86 In some early vernacular stories, the term “laolang” is also used to refer to those people who were responsible for an earlier version of the narrative.87 It is difficult to ascertain whether laolang was just a different term for shuhui xiansheng or cairen. In any case, since they were credited with the compilation of the stories, they might not be just ordinary storytellers. If they were not shuhui members, they could be those raconteurs who were experienced enough and literate enough to play leading roles in the trade. It has now been widely noted that the contingent of raconteurs from the Southern Song to the Ming period was a mixed one, in which men of letters mingled with less-educated public entertainers. Yan Dunyi, for example, admonishes us that “we must by no means underestimate the storytellers’ ability or distrust their educational level (wenhua chengdu). According to the historical sources, among those storytellers the ones who were versed in the art of letters to a considerable degree of sophistication were not few in number.”88 Hu Shiying is even more unequivocal in pointing out that, mingled with street entertainers, there were scholars who had failed in the examinations: Among the storytellers and authors of huaben of the Song times, some were professional street performers and others were scholars who had failed to acquire a degree. Those storytellers in Lin’an during the Southern Song, such as Xu the Gongshi, Zhang the Jieyuan, Liu the Jinshi, and Dai the Scholar, were probably intellectuals who joined in the circles of storytelling either because of their failure in the state examinations or because of their interest in the artistry. . . .

166

Chapter 5

It was therefore nothing extraordinary that men of letters joined the shuhui to compile scripts or promptbooks for the storytellers or even took part in the performance themselves.89

In his discussion of the formation of Shuihu zhuan, Richard Irwin attributes the combination and compilation of the materials from different sources to “a shuhui which existed in or near Hangzhou at the end of the Yuan period.”90 It is impossible to verify Irwin’s dating and location of that hypothetical shuhui, but the involvement of shuhui members or people of similar types in the evolution and textualization of Shuihu zhuan seems to be beyond any doubt. Some vestiges of their role, as a matter of fact, have survived the long process of the textual evolution. In chapter 46, when the adultery between Pan Qiaoyun and Pei Ruhai ends up with the monk murdered by Shi Xiu, the narrative takes a turn to acknowledge some shuhui xiansheng as being responsible for some verse lines in the text: “Later, the shuhui gentlemen in the city, knowing all about the affair, took up their writing brushes and wrote this song” (3: 1518).91 Similarly, in chapter 94, where the narrator apologizes to the reader for the tangles of the narrative threads, he attributes the problem to the shuhui’s insistence on giving all the details in the narration: “The reader will see that the details of this episode are like scattered grains of sand. When our predecessors, the shuhui, passed them on, they wanted to recount each individual thing, but it’s hard to tell them all at once” (5: 3028–3029). We do not know for certain the names of any of the shuhui xiansheng that took part in the shaping of the narrative discourse of Shuihu zhuan, but for Shi Nai’an and Luo Guanzhong, the two men to whom the compilation of the narrative has most often been attributed, scholarly inquiries have been made on the possibility of their involvement with the shuhui. Richard Irwin has noted that Luo Guanzhong’s sobriquet Huhai sanren (Unattached Roamer of Lakes and Seas) “hints at failure to secure official employment.”92 Irwin points out that this sobriquet may have been a deliberate imitation of Jianghu sanren, a sobriquet for the Tang scholar Lu Guimeng (?-881?), who roamed about the country after he was disappointed by his failure in the jinshi examination.93 As for the more controversial figure of Shi Nai’an, Yan Dunyi has also proposed that Shi might be a shuhui member: “Only if Shi Nai’an was not only a storyteller but also a member of the shuhui, does [the possibility of ] his participation in compiling as well as preserving Shuihu zhuan deserve serious consideration.”94 However, as so little is known about the life of either Shi or Luo, their roles in compiling the narrative remain largely hypothetical. Even if they were indeed involved in the textualization, they could be merely two of the many, especially since the textualization was such a protracted and gradual

The Engine of Narrative Making

167

process that was not completed until much later than the lifetime conjectured for either Shi or Luo, as my philological study of the fanben text has demonstrated. Yet while the identities of these two men cannot be verified, it can still be said that the textualization of the narrative was the ultimate result of the cumulative efforts by perhaps generations of people like shuhui xiansheng, cairen, and laolang, lettered people who joined in the world of popular orality after they tumbled off the “ladder of success”—the state examinations.

Imprints of the Men of Letters on the Shuihu Discourse To be sure, the system of civil service examinations played a pivotal role in China’s sociopolitical history from the Sui-Tang periods up to the beginning of the twentieth century. Despite the many mediocrities it produced, it did give China some of her most brilliant statesmen in history. Yet among other deleterious effects, the system, by sanctifying the classical texts, contributed greatly to the widening gap between the literary language and the living tongue of the common people. As a mechanism promoting those who excelled in the Confucian classics into officialdom, the examinations took the classical language right to the center of politics. One’s literary competence became almost identical with his knowledge of classical texts, which in turn stood for wisdom and intelligence; and the ability to read and write in wenyan potentially meant the power to rule and govern. Through the examination system, therefore, the classical language became politically triumphant, and it was in the interest of those in power to perpetuate the privilege of wenyan and maintain the disparity “between the esoteric literate culture and the exoteric oral one.”95 The sanctification of printed words from past ages led naturally to the canonization of established literary genres and styles. As a result, there was a longstanding practice in the Chinese literary tradition of imitating the ancients. Mao Dun considers this one of the factors that prevented Chinese literature from a fuller development: “The penchant for imitation is part of human nature, but the propensity to imitation among Chinese literati arose merely from a blind faith in the ancients.”96 His observation is tinted by a radicalism typical of many May 4th intellectuals, but his comment on the conservative force in the classical tradition is not off the mark. During the early Ming period, the rigid Eight-Legged Essay ( bagu wen) became the standard format for examination papers.97 It set specific requirements on all the details in the array of sentences and paragraphs, as well as the general layout.98 Essays of this format written by successful candidates in previous examinations became models that assumed an importance almost equal to that of the Confucian classics themselves. In Rulin waishi, when the examination veteran Ma Chunshang urges

168

Chapter 5

Kuang Chaoren, a younger scholar, to take the examinations, he reminds him of the pivotal role of the model essays: “What books are to be studied today if not our selections of bagu compositions?”99 Such increasingly meticulous formal regulations, enforced politically, proved a hindrance to literary originality. But since the success in the examinations now hinged entirely on the skills in this array of words, candidates had to tuck away whatever literary creativity they might have and force themselves onto the Procrustean bed of the bagu essay. As the civil service examinations were virtually the sole purpose for the entire educational system, it would be hard to imagine that those men of letters who were involved in the transmission and textualization of Shuihu zhuan should have voluntarily given up their hope for an official career. Most of them, we can be sure, must have had some traumatic experience in the examinations before they became completely disgusted and disillusioned. Then, if the audience of the oral Shuihu complex played a role in the narrative making, as discussed earlier, it is reasonable for us to expect those men of letters to have had an impact on the formation of the narrative as well. Shuihu zhuan is, of course, a narrative about the rebels, but its focus is clearly not so much on what the bandit heroes do as rebels as on how they become rebels. As the narrative takes much pain to show, none of the bandits wanted to live as an outlaw in the beginning. With a few exceptions such as the Ruan brothers, almost all of them were part of the Establishment before they joined the rebellion. Many of them were either members of the landed gentry, government officials, or officers in the imperial army. Even Li Kui, the character probably closest to being a representative of the class of peasantry, was a turnkey. While they become bandits because of various personal circumstances, the general pattern remains more or less similar: They are unfairly driven out of the social group to which they originally belonged and are, as the Chinese idiom goes, “forced to go up Liangshan Mountain” ( bishang Liangshan). They have hoped to serve the imperial court and win fame for themselves and their families, but their dreams are shattered through personal misfortunes, and consequently their abilities remain unrecognized. Yang Zhi, an officer in the imperial army who later becomes a bandit chieftain, serves as an example. What he puts in an agitated outcry, after his application to rejoin the military service is turned down, can be taken as a summary for the bitter disappointment of the heroes: “It’s just that I didn’t want to sully the family name. I hoped for a chance to distinguish myself with spear and sword in a border post, to win honors for my wife and opportunities for my sons, and reflect glory on my ancestors. I never expected to get such a rebuff!” (chapter 12; 1: 366).

The Engine of Narrative Making

169

Other bandit chieftains feel the same kind of grievance of being affronted and having their talents slighted before they finally join the rebellion. After his wife is insulted by Gao Qiu’s adopted son, Lin Chong feels that he is not receiving the respect that he fully deserves as the most capable arms instructor in the Imperial Guards: “I have all my talents in vain, as I have no intelligent superior to recognize them [Kong you yishen benshi, buyu mingzhu]. I serve under petty men from whom I have to take a lot of crap” (chapter 7; 1: 233). Not content with making a paltry living by fishing, the Ruan brothers also feel that “All [their] talents are in vain” ( Kong you yishen benshi). In their conversation with Wu Yong, they express a burning craving to have their talents recognized: “We’re not less capable than others [Womende benshi you bushi buru bieren]. But who recognizes our worth? . . . We sell this column of hot blood to the man who appreciates its worth!” (chapter 15; 1: 448–449). An allegorical reading of the whole narrative, to be sure, cannot be totally justified. The work, like many other narratives that arose from an oral tradition, is so discursive that it defies any attempt at a consistent allegorization. But what happens to the Liangshan heroes in the narrative clearly presents itself as parallel to what happened to the men of letters involved in the transmission and textualization of the Shuihu complex. Like the characters in the narrative, these literary people were also outcasts from their original social group, the class of scholars, or shi. As Yang Zhi’s and many other Liangshan heroes’ expectations are rebuffed, for those shuihu xiansheng or cairen who had failed in the examinations, the path to personal and familial fame was also blocked. Since the examinations were not considered a fair measurement of one’s talent, these lettered people would not accept themselves as being men of lesser literary abilities. Instead, just as the bandit heroes in the narrative, they might feel that they had been unfairly driven to the cultural periphery and their talents had been wantonly slighted and neglected. If the frequently recurrent term “benshi” apparently refers to a hero’s martial prowess, it is sometimes substituted in the narrative by the word “cai” or “caixue.” When Wu Song conquers Jiang the Door-God with his fighting moves of the Jade-Circle Steps ( yuhuan bu) and Duck and Drake Feet ( yuanyang tui), such martial skills are said to be Wu Song’s “true talent and genuine learning” (zhencai shixue) (chapter 29; 2: 929). Elsewhere, as the Ruan brothers pour out their bitterness over their nugatory life as fishermen in their conversation with Wu Yong, the narrative comments in a verse stanza: Only because the treacherous treated the talented [ youcai ] unjustly, Heaven sent the baleful stars to the human world. (Chapter 15; 1: 449)

170

Chapter 5

The repeated uses of cai and caixue to refer to the warriors’ martial abilities are highly interesting, as both words, especially the latter one, regularly mean literary talents. A similar case is the use of the word “wenxue” (literary learning). In chapter 19, as Lin Chong launches a bloody coup to kill Wang Lun, the narrow-minded chief of the band, and make Chao Gai the new leader on Liangshan, he calls Wang Lun a “village scholar” with “no literary learning in the bosom” ( Xiongzhong you mei wenxue) (chapter 19; 1: 587), as if a good learning in literature were a prerequisite quality for being a robber! While it is obvious in such examples that cai, caixue, and wenxue do not appear in their most common sense, one is led to wonder about the reason for such catachrestic usages. That reason is not hard to discern. By applying cai, caixue, and wenxue to the bandit heroes, the narrative highlights the comparability between those capable but unhappy warriors and the talented but disappointed scholars. Indeed, at many places in the narrative, the description of the heroes’ frustrated aspirations can be better understood only when we take into account the role of those men of letters in the formation of the narrative. In chapter 11, Lin Chong stays at Zhu Gui’s tavern for the night on his way to Liangshan. Emotionally, he broods over his unfortunate situation of “having a home but not being able to return to it, and having a country but not being able to serve” ( You jia nan ben, you guo nan tou). Yet the main reason for his bitterness, as he reveals in the poem he writes on the wall, is the loss of his official career ( gongming ):100 Like a floating twig I was a wretch adrift, Gone is my official career [ gongming] like a leaf blown away, But someday when I have my way, All east of Mount Tai will be under my sway! (1: 341–342)

Similarly, when Yang Zhi’s application for a position in the army is rejected by Gao Qiu, the narrative presents a stanza lamenting the ruin of Yang Zhi’s gongming: Who would know that when the villainous persecute the loyal, All opportunities for an official career [gongming] suddenly evaporate. (Chapter 12; 1: 365)

When Song Jiang drinks alone in Xunyang Pavilion at one point during his exile in Jiangzhou, he is suddenly overwhelmed by a surge of melancholy. Yet despite his reputation as an exemplary filial son, he does not think of his

The Engine of Narrative Making

171

old father as much as he bemoans his own failure to achieve fame and success: “Though over thirty, I haven’t made my name yet or achieved anything” ( Ming you cucheng, gong you bujiu). Like Lin Chong, Song Jiang also writes a poem on the wall, with the hope that someday he will revisit the tavern after he has earned his place in the world: Since childhood I studied classics and history; Now grown up I am astute and intelligent. Like a tiger enduring in the wilderness, I crouch with tooth and claw, intent. Unfortunately I was tattooed on both cheeks— What a humiliation to be exiled to Jiangzhou! If I can take my revenge someday, The mouth of Xunyang River will be tinted red by blood! (Chapter 39; 2: 1255–1256)

As a clerk in a county magistrate’s office, Song Jiang is known to be able to write legibly and is familiar with administrative procedures, but nowhere else in the narrative is he portrayed as a scholar well versed in classics and history. Nor is it ever said that he tried to pursue gongming beyond his clerical position or suffered any defeat in the examinations. He is exiled for his murder of the woman Yan Poxi, and the exile is actually a clemency for an offense punishable by the death penalty. The craving for revenge and self-vindication expressed in the poem is therefore hardly justifiable by the circumstances and somehow out of context with the Song Jiang story. But if we consider Song Jiang, and indeed Yang Zhi, Lin Chong, and many other bandit heroes for that matter, as creations of a long process of transmission and textualization of the story complex in which the frustrated men of letters played a key role, the seemingly misplaced concept of gongming and the great importance attached to it become more understandable. Amid the writing-orality interaction, those shuhui xiansheng, cairen, and laolang may have projected their own wounded egos onto their portrayal of the bandit heroes. Some of the early commentators on Shuihu zhuan did not miss the parallel between the bandit chieftains in the narrative and the forsaken scholars who were involved in the making of the narrative. The comment in one of the prefatory pieces to the Rongyutang edition on the anxiety and resentment of the Liangshan heroes can be applied equally well to those frustrated men of letters: If the less virtuous are dominated by the more virtuous and the less capable dominated by the more capable, it is only reasonable. But

172

Chapter 5

if the more capable are dominated while the less capable dominate, who can be dominated without feeling the humiliation? It is like a strong man being captured while a weak man is the captor. Can any strong man willingly allow himself to be captured without putting up a fight?101

In the commentary at the end of chapter 17 of the Rongyutang edition, Yang Zhi and Lu Zhishen are said to be two worthy generals (shang jiang ) who are “wandering about [becoming] destitute and lonely” because none of the authorities has good eyes to recognize their worth. “If there is in the imperial court someone like Cao Zheng or Zhang Qing, how can they end up like that?” (1: 526). Again, the commentator does not refer explicitly to the men of letters whose talent was neglected by the government, but the analogy is evident. It was possible, then, for those frustrated literary people to vent their emotions through the transmission and textualization of the narrative. This might be an important reason that, among the different oral traditions that later crystallized into full-length narrative texts, the Shuihu complex must have been the most attractive to those disappointed people of letters, as suggested by the fact that it was the Shuihu narrative that traversed the farthest in the course of vernacularization. Those shuhui xiansheng and cairen could see in the narrative a story of themselves, as they were, in a sense, no less rebels than the Liangshan bandits. The heroes’ capitulation to the imperial court, which amounts to a voluntary surrender of the victors to the defeated, is a conjuncture in the narrative that has received a great deal of critical attention.102 The prominence of that event in the narrative might be interpreted in terms of the role of the outcast scholars in the narrative formation, for whom the emperor’s solicitation of the rebels’ military service might be a development particularly congenial. Just as the rebels still hope to win honor for their families and reflect glory on their ancestors after proving their military superiority, the scholars driven to the margins of society might still cherish the wish that someday they would be summoned into officialdom. Also, just as Song Jiang and his men prove convincingly after their surrender that they are a much more effective force for the emperor than the troops under the command of Gao Qiu and Tong Guan, it could be the hope of the frustrated scholars to have a similar moment of triumph and self-vindication to show the superiority of their talent over those in office. Shuihu zhuan, therefore, can be read as a narrative of a group of warriors who, having been deprived of their gongming, strive persistently to regain it by proving their real worth. This comeback story might be extremely mean-

The Engine of Narrative Making

173

ingful to those who were rejected by the examination system. Shi Jin, after his involvement with the robbers led by Zhu Wu, still hopes for an official appointment: “If I can find my teacher and get some sort of job I can distinguish myself in, I’ll be happy the rest of my life” (chapter 3; 1: 83). For Song Jiang, the rebellion is no more than a detour to a glorious official career, as he makes clear in his repeated admonitions to his fellow rebels.103 In a way, the martial contest through which Lin Chong wins the award of silver from Chai Jin (chapter 9) and the military tournament through which Yang Zhi earns his promotion (chapter 13) suggest an idealized process for official appointment in which the selection is based exclusively on an indisputable measurement of merit.104 Even the ranking of the chieftains—an important and recurrent event in the narrative in which each warrior is given a due place in the hierarchy based on worth or, in the case of the final ranking, on preordained destiny105—is not unlike a chimerical version of a published list of successful examination candidates.106 Yet gongming proves too evasive for the heroes’ pursuit, even after they have demonstrated their outstanding merit. As Yan Qing decides to give up his official rank and withdraw into seclusion immediately after Song Jiang’s triumph over Fang La, the narrative comments in these verse lines: People cling desperately to fame and success, But fame and success may not be your life companions. (Chapter 99; 5: 3217)

Following Yan Qing, several others who have earned their official appointments, such as Ruan the Seventh, Chai Jin, and Li Ying, all decide to retire from their offices as well. Song Jiang and Lu Junyi, the two leaders of the band, are both poisoned to death by the villainous ministers. Even though Song Jiang and his men have lived up to their pledge that they will “never betray the emperor” ( bu fu chaoting ), they are in the end “betrayed by the emperor” (chaoting fu wo). The irony is driven home in the dialogue between Lu Junyi and Yan Qing in chapter 99. When Lu chides his former servant for his “fruitless act” (mei jieguo) in giving up his official rank when he is about “to return home in official finery” ( yijin huanxiang ), Yan Qing replies emphatically that his retirement “is fruitful” ( you jieguo), and adds, quite ominously, that what will turn out to be truly “fruitless” is the master’s adherence to fame and rank (Zhikong zhuren ciqu wu jieguo er). In chapter 100, after Song Jiang’s death, the emperor has a dream at the courtesan Li Shishi’s home. In the dream, the sovereign visits the Liangshan Marsh, where the ghosts of Song Jiang and his friends congregate. Now dead,

174

Chapter 5

Song Jiang reaffirms his “unwavering loyalty” and begs the emperor for the fair treatment he was deprived of while alive. Soon after the dream, the emperor, after certifying Song Jiang’s death, confers on him the posthumous title of the Earl of Loyalty, Chivalry, and Efficacy and decrees that a temple be constructed and dedicated to his memory. Both loyalty and talent now seem to be lavishly recognized and profusely awarded; in other words, Song Jiang finally achieves the gongming he has aspired to all his life—but only when the gongming has become as surreal and hallucinatory as his ghost itself.107 Indeed, the pursuit of gongming is like the pursuit of a phantom for Song Jiang and his friends—just as it was for the men of letters who had been casualties in the state examinations.

Shuihu zhuan as a Work of “Venting Indignation” In the history of Chinese literature, a theory on the genesis of literary writing goes under the term “fafen zhushu,” which can be loosely translated as “vent one’s indignation in writing a book.” According to this theory, personal adversities toughen one’s fortitude and stamina, and therefore they give rise to a strong determination to reassert oneself by realizing one’s literary aspirations. The theory originated with the great historian Sima Qian, who enumerates several instances of fafen zhushu in his famous “Bao Ren An Shu” (Letter to Ren An): When King Wen [of Zhou] was locked up in custody he deduced the principles of the Yijing. When Confucius was in dire straits he wrote the Chunqiu. Qu Yuan was exiled and he composed Lisao. Zuo Qiuming lost his sight and he completed his Guoyu. Sun Bin had his kneecaps chopped off and his works on the art of war were in a long array. Lü Buwei was banished to Shu, and his Lü lan was passed to posterity. Han Fei, trapped in the prison of Qin, [brought into being] his “Shuo nan” and “Gu fen.” The three hundred pieces in the Shijing were mostly works of venting indignation by sages and men of virtue. Such people could not release their pent-up emotions, and therefore they gave accounts of the past and looked forward to the future.108

The best-known exemplar of fafen zhushu, however, was Sima Qian himself. Implicated in the case of one of his colleagues, the historian was inflicted the most humiliating penalty of castration. After that, Sima Qian lived in extreme spiritual pain. But paradoxically, the castration motivated him in his history writing, for he wanted to see his gigantic work, the Shiji, “enshrined in famous

The Engine of Narrative Making

175

mountains and passed on to posterity.”109 He succeeded. After being bent low in humiliation, he bounced back to a lofty height, and the Shiji became the classic for later narratives—both historical and fictional—to emulate. The castration that deprived him of any progeny only made him the forefather to many generations of writers. That Shuihu zhuan is a “work of venting indignation” has been noted by many critics. In one of the prefatory pieces to the Rongyutang edition supposedly written by Li Zhi, for instance, the novel is compared to the literary expressions of fafen by the ancient sages: Sages of the ancient times wrote only when driven by indignation. To write without being indignant is like shivering without feeling cold or groaning without being sick. What that would look like! Shuihu zhuan is a work for venting indignation.110

At the end of chapter 100, the Rongyutang commentator makes the observation that the ending of the narrative, where Song Jiang appears in the emperor’s dream, is “pregnant with deep meaning:” It suggests that the whole narrative is an account of a dream. Otherwise, how can it be that robbers should be made officials when alive and commemorated in the temples after death? [ The compilers] were only taking advantage of this story to vent their indignation [ faxie buping]. If the reader takes it literally, he would be a simpleton talking in a daydream [chiren shuomeng ]. (5: 3274)

In one of the prefaces to Yingxiong pu, a combined printing of the simplified versions of both Sanguo yanyi and Shuihu zhuan,111 the Ming writer Yang Minglang also dwells on the theme of “venting indignation:” “Under the chilly mists and the cold moon, or amid the bitter winds and miserable rain, wouldn’t there be heroes and people of outstanding talents joining each other in chanting the stirring strains, to pour out their shared grievances and indignation?” ( gong tu qi laosao buping ). The two narratives, according to Yang Minglang, are good lessons for rulers on the vital importance of proper selection of talents for the government: Those who are sovereigns must read this book; once they have read it, the heroes will be on the sides of the throne. Those who are premiers must read this book; once they have read it, the heroes will be in the imperial court.112

176

Chapter 5

In other words, the grievances and indignation in Shuihu zhuan are about the disrupted and corrupted order in official appointments, where talent and merit were often disregarded. As the Qing writer Wang Tao (1828–1897) puts it in his preface to a later edition of the narrative, “The men of noble character were out of office while villains held the power. . . . People of honor and talent were frustrated and destitute, even with no place to settle themselves. . . . Only when one reads Shuihu zhuan in this sense does one become a good reader of the novel.”113 While most of the Ming-Qing commentators agree that Shuihu zhuan is an expression of resentment and anger, Jin Shengtan, at least at some points, appears to be a dissenter. In his “How to Read the Fifth Book of Genius,” Jin Shengtan observes: When reading a book the first thing to be taken into account is the state of mind of the author when he wrote it. For example, the Shiji was the product of Sima Qian’s bellyful of stored-up resentment. Therefore he poured [his emotions] into the writing. . . . Shuihu zhuan, on the other hand, is a different matter. Its author, Shi Nai’an, had no bellyful of stored-up resentment he needed to let out. Wellfed, warm, and without anything else to do, carefree at heart, he spread out paper and picked up a brush, selected a topic and wrote out his fine thought and polished phrases.114

Such remarks incurred harsh criticism from the modern scholar Hu Shi. As Hu Shi argues, people who were “well-fed, warm, without anything else to do, and carefree at heart” would compose “Eight-Legged” essays, but never a Shuihu zhuan. According to Hu Shi, Jin Shengtan made a misjudgment, because “he unfortunately had no historical sight and failed to understand that Shuihu zhuan had been for several hundred years the arena for both the common people and men of letters to vent their bellyful of resentment.”115 Hu Shi, obviously, failed to perceive the irony in Jin Shengtan’s remarks, which are just “part of a smoke screen designed to deflect criticism of the unorthodox content of the novel,” as David Rolston rightly points out.116 Actually, many of Jin Shengtan’s interlinear notes and chapter comments express his view much more explicitly. In his comment on the prologue (xiezi), for instance, Jin Shengtan exclaims, “I wonder what kind of grievances the author had in his bosom that he wrote like that.”117 In his “chapter comment” on chapter 18 of the Guanhuatang edition, he puts both Shiji and Shuihu zhuan in the category of yuandu zhushu, or “books written out of enmity and spite.” The words of characters such as Lin Chong and the Ruan brothers, according to

The Engine of Narrative Making

177

Jin Shengtan, are “quarrelsome and resentful, going against the principle of decorum.” “However,” he hastens to add, “since even Sima Qian was not immune to the practice of yuandu zhushu, how can it be used as a charge against the fiction writers [baiguan]?”118 Indeed, it has historically been typical of a Chinese man of letters not to declare his anxiety and frustration straightforwardly but to dredge out emotional turbulence through indirect and roundabout expressions. A literary work would best serve that purpose,119 in which he would “snatch someone else’s cup of wine and pour it over his own distress,” as Li Zhi puts it.120 In the case of Shuihu zhuan, however, the situation may have been even more complicated. The Shuihu complex could have an enormous appeal to those frustrated men of letters, probably more so than other oral traditions such as the Xiyou and Sanguo cycles. Yet, while eager to vent their grievances, the disgraced literary people did not simply take over the subject matter from the oral tradition and write a narrative in the classical language in which they had been trained. Instead, they kept their role within the oral tradition itself, transmitting and textualizing the narrative along the entire course of its evolution. Surely their own interest contributed to the making of the narrative, as we have seen, but their interest was never allowed to become predominant and suffocating of other forces involved in the evolution of the narrative. The Shuihu complex, for the most part, remained an oral tradition for the urban residents, as noted in the discussion of the role of the audience. Obviously, those men of letters had a great respect for the oral tradition and the storyteller’s words, as evidenced by the stylistic and linguistic features from different times kept intact in the fanben text. What had been appealing to those literary people was not only the story cycles about a group of disgruntled warriors but the way those stories were told. As the scholars, after their failure in the state examinations, were in a way disinherited of the legacy of classical learning, the covenant with the classical language was finally annulled. For them, to be engaged in transmitting the story cycles through literary means was in itself another way to release their pent-up emotions, no less so than contributing to the making of the narrative plot itself. Shuihu zhuan, therefore, is a work for “venting indignation” not only because it is a story about a group of bandits but also because it appears in a linguistic medium that had been hitherto disdained and slighted. In other words, the story of a rebellion and the rebellious way in which the story was presented became for the disgraced scholars the double edges of their vengeful blade. The content and the form thus perfectly matched each other, each being a part in the concerted chorus of protest. It is such a drastically rebellious work that its adoption of a rebellion as its subject matter turns out only to be too fitting.

178

Chapter 5

The linguistic medium of the narrative, a radical departure from the bureaucratese, was most suitable for telling the stories of the people at the other end of the social spectrum, those who fight against the bureaucracy, and the match between the defiant spirit in the new literary form and the defiant spirit of the rebels in the narrative gives the work its commensurate “objective correlative.”

6 Literary Vernacular and Novelistic Discourse The rise of written vernacular as a new literary language, in China as in the West, was inevitably the result of a long process of the interaction and interpenetration between the forces of written culture and those of orality and of a gradual confluence of literary consciousness with oral sensibilities.1 In the cultural context of early premodern China, the persistent transmission and textualization of the Shuihu story cycles was a major part of the interface between oral and written traditions. With modern knowledge on the nature of oral culture and the relationship between orality and writing, we can now pay tribute to Shuihu zhuan as a great literary work while at the same time recognizing the oral provenance of the bulk of the narrative discourse. This, as we have noted, does not mean to discount the self-conscious artistic efforts by the men of letters involved in the editing and compiling of the work. The biggest paradox, however, is that the literary sensibility of those men of letters found its best expression in the storyteller’s voice. Cast in the particular sociopolitical conditions and the literary climate of the time, this literary sensibility was one for revolt, rebellion, and rebirth. As the men of letters felt the pressing need for political protest as well as new literary possibilities beyond wenyan, popular orality happened to be their most reliable and most powerful ally. The oral mode of existence of Shuihu zhuan’s antecedents is by no means a contradiction to the now generally accepted appraisal of the novel as belonging to high literary culture. The fact that Shuihu zhuan, especially in the fanben recension, drew fascinated critical attention from such prominent literati figures as Li Zhi, Hu Yinglin, Yuan Hongdao, and Jin Shengtan indicates that the work in print, unlike its oral antecedents, no longer belonged to the goulan or wazi. Yet if the written vernacular prose in the novel was a prod-

179

180

Chapter 6

uct of the long-term orality-writing interaction, this question has to be addressed: In what ways does the new language medium contribute to the literariness of the work? As this discussion will show, what distinguishes vernacular fiction from its wenyan precursor is not merely a difference in language style. Rather, the inauguration of vernacular prose brought about a most profound change in the conception of narratability—the very idea of how stories should be told. As the earliest full-length narrative in mature vernacular prose, Shuihu zhuan presents a type of narrative art that is fundamentally different from that in wenyan fiction.

Literary Vernacular and the Interpretational Ramifications of Textualization Different from any utilitarian and practical use of the language, the vernacular prose in Shuihu zhuan serves a function that is primarily aesthetic. It is literary vernacular—literary in both the sense of being written and the sense of pertaining to literature. Textualized largely from storytelling, the literary vernacular in the novel opened up a new territory for reception and interpretation. As already discussed, the story making of the narrative was, to a significant extent, determined by the dynamic interaction between the storytellers and the audiences. Like any other oral tradition, it was bound tightly with a social, cultural, and circumambient contextuality, and the telling of the story was a living event that took place between living people. Yet as the storyteller’s words now appear in print as literary vernacular, what was once a living event becomes a fixed object. Involved in this process was both death and immortalization, like the paradox in the making of a zoological specimen: In order for an animal to be preserved for many years’ research, it has to die in the first place. The death in textualization was thus, as Walter Ong suggests, a death of fecundity.2 The spoken words were buried in the text, but out of the tomb arose literary vernacular, much more versatile than its oral predecessor. As the words’ flesh-and-blood ties with the living environment are now severed, the text becomes ready to fit into any number of reader’s interpretational contexts. Related to this change of context is a drastic enlargement of the communication framework, the triangular model formed by the message and the emitter and the receiver of the message. In an oral storytelling situation, the storyteller is the ultimate voice that delivers the story, and the audience is those who hear it. They are living and interactive contemporaries, each at one end of the communication line. Their interaction determines the communication message, but as living people they are not part of the message itself. In the case of the Shuihu story cycles, the storyteller and the audience obviously did not

Literary Vernacular and Novelistic Discourse

181

live in the same world as Song Jiang and his cohorts. Rather, they were creators and interpreters of the fictional world in the narrative. With the textualization, however, both the storyteller and the original audience become objectified along with the oral narrative itself. They are no longer living people at different poles of the communication line but become part of the message, or the metamessage, for the words in print always suggest a different communication model, the one between the writer/scribe and the reader. The speaker and listener of the living voice now become fictionalized. Like Marlow and his audience on the deck in Joseph Conrad’s Heart of Darkness, they become personae for a “writer,” for the printed words always point to the existence of a Joseph Conrad behind the storytelling scene on the deck—the ultimate producer of the text that is meant for the reader. In Gérard Genette’s terminology, the storyteller, who was an “extradiegetic narrator” in the oral storytelling, now turns “intradiegetic” in the novel. As a result, the raconteur and audience, who have lost to the reader their status as ultimate interpreters, become themselves part of the text to be interpreted. More significant, while in the oral storytelling situation the interpreter and the message formed a kind of continuum, in the written communication model they are separated by the mute marks of print. The oral narrative was the result of the parley between the storyteller and the audience, but once it appears in a book it becomes something to be scrutinized, reflected upon, and played with by the reader. This “separation of the knower from the known,” to put it in Eric Havelock’s words, makes possible the “recognition of the known as an object,” which in its turn enables analysis of the matter.3 Analytical thought requires distance. Through textualization, the rambling narrative of the Shuihu cycles was wrapped up into a spatial unity. With the whole text as an object for close study, a reader is now able to find in it different possibilities of literary interpretation.

Wenyan as Language for Narrative To fully appreciate the role of literary vernacular, we must revisit briefly the history of Chinese narrative literature. The purpose is to examine the ways in which the two different linguistic media—wenyan and baihua—contribute to or even determine some contrasting narratological features in different narrative genres. “As the dominance of historiography disintegrated, fiction rose” (Shitong san er xiaoshuo xing ).4 The rise of vernacular fiction has always been associated with the decline of classical historiography, which was for many centuries the most privileged narrative genre in China. Classical Chinese historiography, whether in the annalistic-biographic form initiated by Sima Qian or in the chronological form following the tradition of Chunqiu (Spring and Au-

182

Chapter 6

tumn Annals), is primarily about “history makers”—namely, kings/emperors, generals, and ministers. They are similar to Northrop Frye’s hero in the “highmimetic” mode of fiction, being “superior in degree to other men” and having “authority, passions, and powers of expression far greater than ours.”5 What is presented in Chinese historiography is in a sense a homogeneous world, not unlike the one Georg Lukacs found in Western classical epics: When the world is internally homogeneous, men do not differ qualitatively from one another; there are of course heroes and villains, pious men and criminals, but even the greatest hero is only a head taller than the mass of his fellows, and the wise man’s dignified words are heard even by the most foolish.6

Even as Chinese historiography is known for its successful portrayal of a vast variety of characters, these characters belong largely to the same social strata, far removed from that of the common people. This sense of distance is further reinforced by the gap in time. Like Western epics, Chinese historiography tells stories of an “epic past” that is “absolute and complete.”7 It is a past that stands there over a distance, purely as a tradition, and nothing can be added to or taken away from it. This valorized distance—a distance that is both sociopolitical and temporal—is well in keeping with the “official air” of the genre. It was no accident that Sima Qian’s history writing had to be based on the imperial archives; and in the subsequent dynastic periods the compilation of historiography was often put under the direct supervision of a prime minister. “The early Chinese histor,” as Sheldon Hsiao-peng Lu puts it, “was less a free inquirer of the past than a state agent.”8 History becomes an authoritative discourse, enshrined in a distant zone organically connected to a past age inhabited by people who were socially superior. It is kept away from the familiar vicinity of the mundane life and stands in untouchable grandeur and loftiness. As the world of historiography is not only far removed from the present but also elevated far above our own world of the commonplace, it would only seem natural that stories of that world should be told in a language different from our own and that people in that world should speak differently from our next-door neighbor. Here wenyan seems a perfect fit as a literary language: As it was not spoken by anyone among the readers, it contributed to the sense of distance, the inapproachability, and the official dignity and authoritativeness of classical historiography. For the other narrative genres in wenyan, such as zhiguai (stories of ghosts and spirits) and chuanqi (tales of marvels) that flourished mainly during the Six Dynasties (317–589) and the Tang periods respectively, historiography was

Literary Vernacular and Novelistic Discourse

183

the unchallenged model. Most of the writings in these genres went under the rubric of historical records, either as derivatives from or as supplements to historiography. To assuage the inveterate Chinese distrust of unbridled fictionality, the narration of miraculous and supernatural events had to be legitimized by reference to a “reliable” source—often a documented history. This alignment of what was unmistakably fictional with historiography was probably the only way out of a dilemma. On the one hand, the fictionality of the tales had to be camouflaged, for the author was supposed to be a transmitter of what had been proved truthful rather than a fabricator of something fabulous and fantastic,9 and therefore a cloak of a certain amount of factuality or at least varisemblance was needed.10 On the other hand, the supernatural beings and miraculous happenings were too outlandish to be put in a contemporary and therefore familiar milieu. For this predicament, an alignment with historiography seemed an apt compromise. If the zhiguai or chuanqi stories could pass for histories, then they should have a “legitimate” claim to being truthful and factual while maintaining an unbridgeable distance from the reader’s immediate world. It may be said that the major narrative genres in wenyan stake their raison d’etre on a sense of distance they conjure for the reader. It is here that the classical language provides its best service. Estranged from the most recent development in the living language, wenyan was largely insusceptible to the contemporary social milieu. Precisely because of its transcendence over an immediate reality that is necessarily material, tangible, and familiar, wenyan helps enhance the respectability of historiography as an official discourse and the sense of “otherworldliness” in the stories of the miracles and ghosts. What is distant may not look distinct. Yet to look distinct was usually not a major aesthetic concern in wenyan literature. As a literary language removed from the “here and now,” wenyan lacked a signifying capacity that could cover all physical details in the vast external reality. In classical poetry, wenyan’s incorporeal referential orientation contributed to poetic suggestiveness, which does not point to the physical and the explicit as much as to the invisible and the ineffable. A similar suggestiveness was often the goal to achieve in wenyan narrative as well. Since the purpose of literature was to identify with the Dao that was intrinsic to the mutable reality, focusing on the physical details was considered not only unnecessary but simply detrimental to the moralistic as well as the aesthetic integrity of narrative art. This accounts for the fact that wenyan narratives often feature a sketchy delineation of events, where the temporal and spatial specifics are missing.11 In this respect, wenyan narrative embodies the same ideal of artistic pursuit as the Chinese ink-and-wash landscape painting, where a mountain could be delineated in just a few strokes and

184

Chapter 6

a lake could be represented simply by a blank space. Since corporeal imitation was not the aesthetic interest here, external likeness was not only unsought but simply dismissed as unworthy. Not aiming at creating a replica of the total actuality but suggesting some outlines of it, wenyan narrative in its relation to the “real” world is not unlike what Robert Scholes and Robert Kellogg call the “illustrative” narrative, which depends more on artistic tradition and convention than on the representational and referential power of language itself.12 In relation to such an artistic interest, there is a particular conception of narratability that regarded the recounting of a past action or a sequence of actions as the objective of narrative without committing itself to a simulation of the linguistic milieu of such actions. As noted earlier, there is no wenyan writing that is completely immune to the invasion from the vernacular. What we call wenyan narratives are often in different mixes with vernacular ingredients. Typically, however, a wenyan narrative tends to be sustained by a linguistic detachment from the narrated world. Since what really matter are actions, characters in the narrative appear primarily not as speakers, but as doers. The narrative tells the story from without, so to speak, and the characters are seldom given a chance to dramatize their actions from within by speaking in their own voices. The narrated world is thus largely a factual world without a fully developed linguistic dimension of its own. In extreme cases it can be a silent world, where the residents do not talk but are only talked about by the outside voice of the narrator. This is of course not to say that characters in a wenyan narrative are mute. Actually, at least for historiographers, recording what people said was considered as important as recording what they did, as is indicated by the famous phrase zuoshi jishi, youshi jiyan ( The first historiographer records events, while the second records speeches).13 The characters do “speak,” to be sure, but they do not speak in their own voices. Rather, they all speak in a more or less identical language—the words of wenyan that the writer puts in their mouths. This does not mean that speech presentations in wenyan do not contribute to a diversified characterization. On the contrary, they often result in an extremely vivid and lifelike portrayal of characters. For a sample of speech presentation as a means of characterization, let us look into a section of “Ying Ning,” one of Pu Songling’s wenyan stories. I deliberately select such a late work as an example of wenyan narrative so as to preclude the possibility to explain the different mode of narrative in terms of any natural change of the language over the course of time. In the story, a young man initiates a dialogue with a girl, his distant cousin with whom he is secretly in love, by showing a flower that the girl has left on the ground earlier:

Literary Vernacular and Novelistic Discourse

185

The young man produced the flower from his sleeve to show the girl. Looking it over, the girl asked, “It has withered. Why did you bother to keep it?” “You are the one who left it on the ground. That’s why I wanted to keep it.” “For what?” “To show my love. Since we met I have been lovesick but just couldn’t get a chance to see your flowery beauty again. Fortunately you are showing pity on me.” “It’s indeed a small thing. How can I be that stingy since we are kinsfolk? When you leave, I will tell the servant to give you a big bundle of the flowers from our garden.” “How silly you are, sister.” “Why do you say I am silly?” “What I love is not that flower, but the person who picked that flower.” “We are relatives, and love goes without saying between us.” “The love I want is not love between relatives, but between husband and wife.” “Any difference?” “Husband and wife sleep together at night.” “But I’m not used to sleeping with someone else.”14

Even if we do not read the rest of the story, from the dialogue we see two human figures remarkably true to life. The young man, obviously more disposed to sexual love, appears to be a glib-tongued seducer. But the girl is perhaps the more impressive character. She is kind, generous, and lively (in the rest of the story she laughs literally all the time), but the most salient characteristic of the sexually innocent maiden is her silliness and puerility. As we look into the way the dialogue contributes to the portrayal of the character, we find that this effect of characterization is achieved through the meaning of her speech, not the way she speaks. While she is so callow as to not follow the young man’s play on the metaphor of the flower and even to miss the explicit overture in the phrase “sleep together,” she expresses all her silliness and naïveté in extremely sophisticated wording. For instance, generously promising that she will not grudge the boy the flowers in the garden, she asks the rhetorical question, “Zhiqi hesuo jinxi?” And to tell the young man that between relatives love is something that can be taken for granted, what she actually says is: “Jiafu zhi qing, ai he dai yan.” Words such as “jinxi” (to grudge) and “jiafu” (relatives) reflect a highly cultured and sophisticated elegance that is quite incon-

186

Chapter 6

gruous with the innocuous silliness of a teenage girl. Basically, both interlocutors’ speeches in the dialogue are on the same stylistic plateau as the words in the narration: The narrator speaks for his characters. Although the dialogue gives us two distinctly different characters, that difference is brought about through semantic means in the speech presentation, not through stylistic diversification. Thus in wenyan narrative there is a unique mode of narration in which both the words of the narrator and the spoken words of the characters are assigned from a largely unified linguistic consciousness. This mode can be very potent semantically and very effective in characterization. However, regarding language in the narrative as belonging only to the narrator but not to the narrated world, this mode does not aim at presenting characters in their respective linguistic consciousness. A narrative written in “pure” wenyan, therefore, would be a field completely dominated by a centripetal linguistic force, which can level all differences between individual voices and dialects of various social groups. It would therefore present an illusory image of verbal-ideological homogeneity of the social reality, forcing the diversity of life into a ready-made linguistic uniform. This kind of “pure” wenyan narrative is of course a mere postulation, serving our analytical purposes only. Even in the story “Ying Ning,” a section of which I cited above as an example of linguistic homogeneity, the girl’s mother speaks a somewhat more worldly language in keeping with her identity as a seasoned old woman. (Both the girl and her mother, as it turns out toward the end of the story, are fox spirits.) What we can say is that, allowing exceptions, the predominance of wenyan tends to limit the role for linguistic stratification and stylistic diversification in the narrative, especially in speech presentation.

Literary Vernacular as Language for Narrative While for many centuries wenyan remained more or less unitary, the living Chinese language never stopped being stratified. While the divergent and heterogeneous dialects pertaining to different social groups were for the most part kept out of the wenyan narrative, they easily found their way into different forms of oral and performative art, especially public storytelling. Popular orality thus became the locus where different types of linguistic consciousness were registered and orchestrated. Chinese storytelling had a very close kinship to popular drama and was emphatically performative. What was particularly appreciated by the audience was the skill of the storyteller in assuming the roles of the characters in the story and emulating their spoken words. “Alongside the epic and lyrical elements,” as Prusek has noted, “an important part of the storytellers’ text was its

Literary Vernacular and Novelistic Discourse

187

dramatic coloring.”15 This is evidenced by the often-quoted lines by the Tang poet Li Shangyin (813–858), which inform us that the Tang storytellers of the Sanguo cycles vividly imitated the stutter of Deng Ai, a general of the Jin.16 To achieve the desired effects in their performance, professional storytellers took great pains in attuning each character’s speech in accordance with his or her personality and social identity. The famous Ming storyteller Liu Jingting (1587–ca. 1670), while learning the art from his mentor Mo Houguang, was told to “distinguish different people’s personalities and investigate different local customs.” Years later, Liu became a master raconteur whose handling of narrative actions was “extremely vivid and detailed,” but what was most remarkable was his ability to dramatize the speech of his characters.17 The Ming scholar Zhang Dai gave a short account of Liu’s depiction of Wu Song in telling the story of Wu Song killing the tiger: “When Wu Song enters the tavern, there is nobody in. [In Liu Jingting’s dramatization] Wu Song shouts roaringly for the attendant, and one can almost hear the reverberation in the empty wine urns in the shop.”18 A similar report is on the performance of Hu Zhong, Wang Shizhen’s (1526–1590) family storyteller. When Hu was rendering the speech of an emperor in his story, he sounded quite royal, calling himself zhen or guaren (I, the sovereign) and addressing other characters as qing (my minister).19 Such historical accounts are corroborated by our knowledge of Chinese storytelling in modern times. In his report on the narrative voices in Yangzhou pinghua, Vibeke Børdahl informs us of the storyteller’s different speaking styles, which enable him or her to “play on a spectrum of different articulations regarding pitch, tempo, breathing, dialect strata of high and low pronunciations, imitation of other dialects, etc.”20 By emulating numerous social dialects in living words, the oral narrative was able to present a simulacrum of the stratified language itself and portray a character’s unique ideology in his or her own discourse. That is precisely what Chinese vernacular fiction inherited from its oral antecedents. With literary vernacular, which itself could mature only through repeated interaction with its oral model, vernacular fiction was able to present not only a character’s acts but also his or her own words. Characters from all walks of life are presented here as speaking persons. The narrative becomes a polyphonic field, a diversity of social speech types, under no hegemony of a unitary linguistic consciousness of a writer/narrator. The difference in presenting a character’s speech in Shuihu zhuan from that in a wenyan narrative was repeatedly noted by the Ming-Qing commentators. In the chapter where Yang Zhi escorts the birthday gifts in the Yuan Wuya edition, there is a marginal commentary on the verbal skirmish between Yang Zhi and the old chief steward: “The army man speaks exactly like an army

188

Chapter 6

man, and the steward speaks exactly like a steward.”21 Likewise, Jin Shengtan celebrates this linguistic stratification and individualization with great enthusiasm: “The Shuihu zhuan does not contain literary language particles like zhi, hu, zhe, and ye. Each individual character is made to speak in his own individual way. This is truly marvelous skill!”22 As repeatedly noted by Jin Shengtan, the characters in the novel are not only portrayed by what they say but also by how they say it. The language of Ruan the Second (Ruan Xiaoer), a rustic fisherman, is often incoherent, but Jin Shengtan acclaims the incoherency as the narrator’s “masterful stroke of depiction” (moshen zhi bi) of a man who is illiterate ( butong wenmo).23 The coroner, He the Ninth (He Jiu-shu), has a peculiar propensity to use idioms incorrectly, which is, according to Jin Shengtan, typical of “a petty man who desperately wants to polish his words but ends up making them coarser.”24 This allegiance to living speech types rather than to conventions in a literary language was largely responsible for the lifelike depiction of the characters. Again, just as Jin Shengtan puts it, “When the Shuihu zhuan describes the personalities of the 108 persons, there are truly 108 different personalities. In other books, the people they describe all look the same, be they as many as a thousand or as few as two.”25 These individual voices in the narrative are not mutually independent or self-sufficient in themselves. Rather, they are engaged with each other dialogically, setting each other off, and conditioning and manipulating each other. As a result, such voices are not only utterances of particular persons but also verbal expressions of mutually interacting types of social consciousness. One of the indicators of the heterogeneous forms of social consciousness is the characters’ self-reference. The characters, especially the Liangshan heroes before they finally join the bandits, refer to themselves in many different ways, in conformity with their awareness of their respective places in the society. Again, Jin Shengtan was too perceptive a critic to miss this: “[Among the 108 persons] some call themselves wo; some an; some sajia; some xiaoke; and some wo laoye.”26 The subtle shades of difference among these self-appellations carry different social, ideological, and geographical connotations. “Wo” is the normative and therefore neutral word for the first person singular, and in the novel it is the most often used term when the speaker does not feel the need to highlight his social position in relation to that of the addressee. Both “an” and “sajia” are dialectal variants of “wo,” with the former mainly used by people from the north and the latter by people from the northwest. “An” often suggests a candid but crude personality often associated with northern rustics, and therefore we frequently hear it from the mouth of such a character as Li Kui. “Sajia” reminds one of the valor and vigor of a guanxi dahan (a big fellow from west of the Pass), and when Yang Zhi encounters Lu Zhishen in a bitter clash, it is the

Literary Vernacular and Novelistic Discourse

189

common accent betrayed by the utterance of “sajia” that helps turn adversaries into comrades (chapter 17; 1: 505–506). “Xiaoke” (I, the humble one) is in the narrative mostly used self-deprecatingly by someone of an inferior social status, while “wo laoye” (I, the master) is often an indicator of the arrogance and impudence of a strong man or a local bully. Thus, when we hear Wu Song call himself “wo laoye” we know the impetuous hero is provocative in trying to pick a fight; and when Song Jiang, the chieftain of the bandits, repeatedly refers to himself as “xiaoke Song Jiang” even to his captives, we feel his purposeful violation of the social codes attached to this appellation, and the excessive modesty that puts him on the border of hypocrisy. Even more heterogeneous are the ways the characters address each other. When Shi Jin meets Lu Zhishen for the first time in a teahouse, he addresses Lu (chapter 3; 1: 86): “Ganwen guanren gaoxing daming?” (May I be so bold as to ask your honorable name, sir?). After his reply, Lu Zhishen asks virtually the same question, but the language is strikingly different: “Ganwen age, ni jiao shenme?” (May I ask your name, brother?). It is easy to see that this brief verbal exchange is made on two different levels of linguistic stratification. Shi Jin, who is from a wealthy family and obviously the better educated of the two, launches the conversation in a polite and formal manner, and his language is more literary than colloquial, as suggested by such a parallelism as “gaoxing daming” (lofty family name and grand given name). In contrast, Lu Zhishen’s words represent a lower stratum of linguistic consciousness. As an illiterate, he lacks the ability to polish his words—and as a forthright character, he would not have bothered to do so anyway. He calls the younger man “age” (my brother), which—unlike Shi Jin’s word “guanren” (sir, gentleman)—is oriented toward a private rather than a societal and hierarchical relationship, and his question “ni jiao shenme” carries none of the deference in Shi Jin’s question. His language is one of spontaneity, almost a “language of innocence” before it becomes loaded with social and public codes. But even such a language is not without its ideological edge. If Shi Jin’s words are appropriate for the occasion and are in full compliance with the social convention of decorum, Lu Zhishen’s crude voice is clearly that of a person who defies any established order, linguistic or social. Let us have one more example from chapter 38, where Li Kui, an uncouth peasant-turned-prison-guard, for the first time meets Song Jiang, who is in the company of Dai Zong, an official and a mutual friend of the other two. Here is the conversation among the three: Li Kui looked Song Jiang over. “Who’s this swarthy fellow?” he asked Dai Zong.

190

Chapter 6

The superintendent smiled to Song Jiang. “You see how rude this guy is, sir. No manners whatsoever.” “I want to ask what’s rude about that?” said Li Kui. “You should have asked, ‘Who is this gentleman?’ Instead, you said ‘Who is this swarthy fellow?’ If that isn’t rude, what else is? Well. Let me tell you who he is. He’s the noble brother you’ve been wanting to join for so long.” “Not that Blacky Song Jiang, the Timely Rain from Shandong?” “How dare you be so insolent!” Dai shouted. “Referring to him like that! Don’t you have any sense of rank? Kowtow to him, quickly. What are you waiting for?” “I will if he’s really Song Jiang. But if he’s just some idler, what the f—— should I bow to him for [Wo que bai shen niao]! You are not just making sport of me, are you, brother?” “I am indeed Blacky Song Jiang of Shandong,” Song Jiang assured him. Li Kui clapped his hands. “My blessed grandfather [Wo na ye]! Why didn’t you say so earlier to give Iron Ox the delight?” He flopped to the ground and kowtowed. (2: 1215–1216)

The outspoken Li Kui does not mince his words. He sees a man of dark complexion and calls him “swarthy” and “blacky,” without any sense of impropriety in doing so. He tells a naked truth but fails to realize that in sociolinguistic conventions, truth, like a Chinese bride, often has to be veiled. He is immediately reprimanded as being “rude” and having no “sense of rank,” but, as the marginal comment in the Yuan Wuya edition points out, it is precisely this rudeness that makes Li Kui “a genuine man” (zhenren), unencumbered with the artificiality and pretensions of social propriety.27 Dai Zong, as an official, is obviously a much more sophisticated speaker who is apt to adjust his locutions according to his “sense of rank.” While his words represent a different social consciousness, one that wrestles with that of Li Kui’s, his own speech itself is internally dialogized. His reprimand of Li Kui in a speech meant for a social inferior sounds imperious or at least condescending, but his question, “Who is this gentleman,” demonstrating to Li Kui the “proper” way to speak, clearly belongs to a different social dialect, something that is meant for the ear of a person of respectable status. There are many places in Shuihu zhuan where the same character’s voice is stratified and dialogized into different social dialects, especially when the person’s address to another character is set off by an inner speech directed to him/herself. One such example is with Dame Wang, the hostess of a teahouse

Literary Vernacular and Novelistic Discourse

191

and the procuress for the lecherous Ximen Qing. When the enamored Ximen, after pacing back and forth for a long time in front of Pan Jinlian’s house, comes to the teashop, the shrewd old woman immediately sees an opportunity for profiteering: “That stud [shuazi ] sets a fast pace,” she said to herself. “I’ll spread a little sugar on the tip of his nose, just out of reach of his tongue. That rascal [nasi ] is good at wheedling favors in the county offices, but this revered Ladyship [laoniang] will make him pay his dues here.” (Chapter 24; 2: 758)

To his face, however, she repeatedly addresses the wealthy man as “your Honorable” (daguanren) and greets him with both respect and intimacy: “Haven’t seen you in days. Have a chair” (2: 759). Then she sets out to joke merrily about the man’s passion. When Ximen is back in the teahouse one more time and overpays her for the tea, the old woman switches back to her inner monologue: “Got him! . . . The stud [shuazi ] is really hooked.” Meanwhile, she becomes even more solicitous and speaks to Ximen with all hospitality: “Your Honorable looks thirsty to this humble old woman [laoshen]. How about a cup of steeped broad-leafed tea?” (2: 761). The duality of the old woman’s language illustrates the ways in which the style of one’s speech is dictated by one’s consciousness of his or her position in the structure of social relations. Her speech to Ximen is regulated by social norms and by her possible gain or loss resulting from the choice of words. But when she speaks to herself—that is, when the voice is hushed and therefore not emitted as a message of social communication—she is temporarily exempted from those social regulations. Thus, we see the difference between these two types of speech: In addition to the shift in the general tone from one of cold and deliberate calculation to one of hospitality and relaxed merrymaking, the deprecating terms “shuazi” (stud) and “nasi” (that rascal) are replaced by the obsequious appellation “daguanren” ( your Honorable), and the insolent “laoniang” (this revered Ladyship) gives way to the selfdenigrating “laoshen” (this humble old woman).

The Change in the Narrative Mode: From Linguistic Apathy to Linguistic Empathy For Shuihu zhuan, therefore, the external world is primarily a linguistic reality— a reality that reverberates with all kinds of voices. A major part of the narrator’s job, as the above discussion shows, is to represent the characters’ spoken words. Yet any representation of the objective world that passes for being mimetic is inevitably shaped by a subjective consciousness. All the characters’

192

Chapter 6

voices in Shuihu zhuan are of course dominated by the voice of the storyteller/narrator himself. While a character’s speech in the narrative may contain transmissions and interpretations of another character’s voice, it is itself relayed by the storyteller on a higher level. A character’s speech, therefore, is inevitably a “voice within a voice,” a speech that is wrapped up in the narrator’s linguistic consciousness. This engagement of the transmitter with the transmitted provides a dialogizing background for the speech presentation throughout the narrative. As demonstrated in Shuihu zhuan, to narrate a story is to reenact a linguistic experience, to reproduce different voices in their original pitches and tones, and then to orchestrate such voices into an artistic construct. What was involved in the shift from classical Chinese to vernacular prose as the language for narrative was a restructuring of the relationship between the linguistic medium and the narrated world. In wenyan narrative, this relationship can be termed “linguistic apathy.” The story is narrated in a linguistic system that is, for the most part, not shared by the world in the story. The writer does not identify with that world linguistically; rather, the narrated world is taken almost as a linguistic vacuum, which the writer is to fill with the “voice” of his own. In Shuihu zhuan, this “linguistic apathy” is replaced by what can be called “linguistic empathy,” where the narrator endeavors to empathize with different linguistic sensibilities in the narrated world and imitates the various speech types of the characters. What I mean by “linguistic empathy” overlaps in some measure with what Jin Shengtan means by “gewu” in one of his prefaces to the Guanhuatang edition of Shuihu zhuan.28 “Gewu” is a term originally employed in Confucian theories of epistemology, where it refers to the effort to recognize the rationalist principles immanent with things in the external world. Jin Shengtan took over the term and somewhat modified it in his narrative theory. In Jin’s critical terminology, “gewu” means, according to Liu Xinzhong, the effort by the narrative writer to try to understand his characters by following the words and actions of real people in life.29 To achieve that goal, one has to study the social milieu that shapes a character’s action patterns and linguistic consciousness. In his explanation of the point, Jin Shengtan borrows the Buddhist term “yinyuan sheng fa,” which means that all phenomena in the world arise from their respective causes and conditions.30 But a more explicit and stronger concept he advances is “qindongxin,” or “transferring one’s mind,” which he considers to be the key factor for lifelike depiction of characters in Shuihu zhuan: It is beyond any doubt that Shi Nai’an was not an adulteress or a thief, but how could it be possible that the adulteress he presents

Literary Vernacular and Novelistic Discourse

193

seems such a lifelike adulteress and the thief he presents such a lifelike thief? Oh! I know it! If one is not an adulteress one does not really understand an adulteress; and if one is not a thief one does not really understand a thief. But if we say Shi Nai’an was neither an adulteress nor a thief, we talk of the Shi Nai’an before he came to write. . . . Once he transferred himself into the imagined positions of an adulteress or a thief, he could not only be an adulteress or a thief but many other characters.31

Jin Shengtan’s “qindongxin” means the effort to conceive a plausible chain of causality from the character’s own perspective, so that the narrative action can fit well with the character’s personality and psychological conditions. Particularly, it suggests a momentary adoption of the character’s linguistic role in order to speak for him or her in a real-life voice, as Jin Shengtan repeatedly praises the narrative for its verisimilitude in rendering a character’s speaking voice in full compatibility with his or her social status. For later writers of vernacular fiction, “linguistic empathy” means, first of all, identification with the linguistic stance of a storyteller on a metadiegetic narrative level. Even when no character speaks in the story or when a character’s speech is narrated instead of quoted, the writer still seems to be transferring himself into someone else’s voice—namely, the voice of a storyteller, real or imagined. The entire narrative is therefore assumed to be the recording of an oral delivery. To various degrees, most of the vernacular narratives following Shuihu zhuan all maintain the existence of a storyteller as some kind of a “modulator” of the narrative discourse, even though the narrative may not have been directly derived from a full-fledged oral tradition. The adoption of such a persona may have served as the writer’s strategy of self-protection in an age when writing in the vernacular could still be considered beneath a literatus’ dignity; but, when virtually all other sectors of the cultural life were still dominated by wenyan, the linguistic empathy with a storyteller could be the only pretext, or “pre-text,” for a writer to temporarily neutralize his “writer’s sensibility” trained in the classical tradition and to write instead in vernacular prose. In this light, the simulated oral context may be not only a deliberately chosen writing craft but also a move necessitated by the cultural conditions. Confronting the deep-rooted precept that yu, or common speech, had no place in literary writing, the writer probably could feel justified in writing in the vernacular prose only by empathizing with an oral model, a fictive storyteller who was supposed to be the original producer of the discourse. Commenting on the writing process of vernacular fiction, Sun Kaidi considers the writer’s linguistic identification with a storyteller a necessary ac-

194

Chapter 6

climatization for writing in vernacular prose. What the writer was supposed to do, according to Sun, was to “let the pen serve as mouth”: To let the pen serve as mouth was even harder than using the mouth itself. . . . Everything that a storyteller expressed with his mouth and gestures now had to be expressed with the pen. It was very difficult and required great effort. . . .While the writer was writing at a desk in his study, he could not consider himself a writer. Instead, he should regard himself a storyteller. His writing in the study was exactly like telling a story at a place of public entertainment, with the audience listening to him in front of the stage.32

“To let the pen serve as mouth” means a voluntary surrender to oral sensibility and the assumption of the linguistic role of a storyteller, and it also means the adoption of the narrative mode of storytelling, in which the narrator constantly shifts his linguistic gears in order to dramatize different speaking voices of the characters. For the later writers of vernacular fiction following the models of Shuihu zhuan and other early novels, therefore, “linguistic empathy” became a twofold process: By empathizing with a storyteller, they put themselves in a position to empathize with various living speech types.

Linguistic Mimesis and Novelistic Discourse By now, the word novel has been generally accepted as the English term to refer to works in Chinese full-length vernacular fiction, especially Shuihu zhuan and other masterworks in the genre. As the label has been attached to vastly different narrative works in the Western tradition itself in the last three centuries, the relative imprecision of the term in the Chinese context has often been forgivingly neglected. But the transplantation of the English term may be more providential than it seems. As a shared “family name,” it highlights some kind of affinity between the two narrative traditions and sets the stage for meaningful comparative studies. In a thought-provoking article, Andrew Plaks argues cogently that, despite the differences in many aspects, Chinese vernacular fiction and the Western novel share enough elements to be “members of the same generic class.”33 One central feature that both narrative forms have in common, as Plaks points out, is “the aesthetic expectation of a ‘realistic’ representation of some phase of human experience.”34 While a realistic impression is sometimes due to the selection of the subject matter, it is, in a more important sense, the product of the means of the representation. For the most part, “the novel’s realism does not reside in the kind of life it presents, but in the way it presents it,” as Ian Watt succinctly notes.35

Literary Vernacular and Novelistic Discourse

195

In an attempt to specify the defining features of the eighteenth-century English novel as a new literary genre, Ian Watt adopts the term formal realism, by which he refers to “a set of narrative procedures which are so commonly found together in the novel, and so rarely in other literary genres.”36 The novel’s distinctive “narrative procedures,” as Watt informs us, include the repudiation of traditional plots and the particularization in character portrayal and in the depiction of the time-space setting. Yet all such efforts at creating a realistic account of the actual experience of individuals involved a change in the language medium, an “adaptation of prose style to give an air of complete authenticity,” because the prevalent prose style of the Augustan period was “much too literary to be the natural voice of Moll Flanders or Pamela Andrews.”37 To Watt, the most successful aspect of Moll Flanders is Daniel Defoe’s prose. It is not well written in the regular sense, but its relative crudity, paradoxically, is precisely where the novel’s “literariness” lies, as “it is remarkably effective in keeping us very close to the consciousness of Moll Flanders as she struggles to make her recollection clear.”38 In other words, Defoe let Moll Flanders speak in her own “natural voice” in order to produce an “authentic” autobiographical memoir of a lowly woman. Samuel Richardson, a much more conscious literary innovator than Defoe, paid an almost meticulous attention to the linguistic sensibility of his characters, especially the feminine linguistic code of his heroines, Pamela and Clarissa. For Richardson, the epistolary form was not merely a method to tell the story, but a window through which to look into the character’s mental process that must be played out in her own words. To achieve in the letters a likeness to a spontaneous transcription of the heroine’s inner speech, Richardson had to break with the traditional decorum of prose even more thoroughly than Defoe did. As a result, his prose, while “neither elegant nor pungent enough to have been used in comedy or satire,” is “redolent of the moral and social milieu” of the world depicted in the novel.39 Speech presentation in the narrative, of course, has a history much older than the novel itself. As we all know, it is what Plato means by the word mimesis, by which he refers to the narrative mode in the Homeric epics where the poet “makes a speech in the person of someone else.”40 Plato favors the mode of diegesis, in which the poet narrates the story himself, over mimesis, and he does so on moral rather than aesthetic grounds. Yet Plato’s mimesis does not really mean imitation of the “natural voice” of the character; it is simply what modern critics mean by “direct speech.” Homer, restrained by the metrical scheme of the epic verse, does not attempt to present Chryses’ “real-life” voice in the beginning scene of the Iliad, which Plato cites as a typical example of the Homeric mimesis. Instead of making us believe “that it is not Homer but an aged priest who is talking,” as Plato claims,41 the “direct speech” is unmis-

196

Chapter 6

takably in Homer’s own epic language, even though it is put in the mouth of the aged priest. Plato’s mimesis is, therefore, an “illusion of mimesis,” for what is represented is not really the character’s speech itself but mostly the semantic meaning of that speech. The situation is similar to what we have seen in Chinese wenyan narrative: The writer’s own language signifies the character’s speech without attempting to imitate it. Yet if there is indeed such a thing as mimesis in a narrative work, the object of the imitation has to be language itself. “Mimesis in words,” as Gérard Genette puts it, “can only be mimesis of words.”42 In an important sense, the novel, in contrast to the narrative genres before it, is based on mimesis of words. It is, first and foremost, a world of speaking persons. As the novel purports to be a realistic representation of human experience, it requires that each character bring his or her unique linguistic being into the novelistic world. This linguistic mimesis is vital, as no human experience is complete without a linguistic dimension to it, because one’s use of language is always a most reliable barometer of his or her socio-ideological existence. Explaining why the novel must present not only a character’s acts but, more importantly, his or her own words, Bakhtin observes: It is impossible to represent an alien ideological world adequately without first permitting it to sound, without having first revealed the special discourse peculiar to it. After all, a really adequate discourse for portraying a world’s unique ideology can only be that world’s own discourse, although not that discourse in itself, but only in conjunction with the discourse of an author.43

The novelist, therefore, does not resist types of linguistic consciousness that are alien to his own. Instead, he actively incorporates a multiplicity of social voices into his work and makes the novelistic discourse a representation of the diversified and stratified linguistic reality, or “heteroglossia,” as Bakhtin calls it. All the social voices are not merely transmitted. Rather, they are artistically represented and orchestrated, which always suggests a dialogic interaction between the imitator (the authorial or narratorial discourse) and the discourse being imitated. The novel thus becomes a narrative genre of “artistically representing the image of a language,” as Bakhtin puts it.44 Indeed, what we call “linguistic empathy” in Chinese vernacular fiction is strikingly similar to the representation of “heteroglossia” in the Western novel. Like its Western counterpart, Chinese vernacular fiction is characterized by a tendency of linguistic decentralization, where an authorial discourse is constantly intersected by other discourses at different linguistic planes. The nov-

Literary Vernacular and Novelistic Discourse

197

elistic world, as we have seen in Shuihu zhuan, is one inhabited by speakers of heterogeneous social dialects. The alignment with living speech types makes vernacular fiction vastly different from wenyan narrative, but wenyan itself is not completely jettisoned. Reduced to the status of a social dialect, wenyan is incorporated into the novelistic discourse, representing a particular type of linguistic consciousness. As an ingredient, it contributes to different mixes with colloquial elements to represent various speech types. In other words, Chinese vernacular fiction is a form of linguistic synthesis. Like the Western novel, it is a genre that artistically presents the sum total of linguistic experience of the social community. It is easy to see that, in imitating the multiplicity of social dialects, vernacular prose has to be the main written medium. If the eighteenth-century English novelists had to break with the norm of traditional prose in order to represent their characters’ “natural voices,” the linguistic preparation for the rise of the novel in China had to be an enormously more strenuous process. It was by no means a coincidence that the emergence of the earliest Chinese novels, especially Shuihu zhuan, also marked the debut of literary vernacular, as full-fledged vernacular prose was not a by-product of the rise of vernacular fiction but a prerequisite for it. Without the arduous process of vernacularization, driven by the constant interaction between writing and orality, there would have been no linguistic conditions for a novel like Shuihu zhuan. While Shuihu zhuan owes its written medium to the textualization of oral prose in storytelling, it also took over from the oral tradition the manner of dramatizing speaking voices, which became a convention in vernacular fiction and turned it into a genre compatible with the Western novel in that aspect. The oral provenance of the early works in vernacular fiction is therefore by no means something extraneous to the new genre or a handicap for it to overcome. Instead, it was a most essential part of the cultural condition that made Chinese vernacular fiction the great literary genre it is. Linguistic mimesis, initiated in storytelling and established as a new narrative mode in Shuihu zhuan, paved the way for the literary realism exemplified by such great novels as Jin ping mei and Rulin waishi, and, ultimately, by Honglou meng. While reviewing the achievements of Chinese vernacular fiction, one must keep in mind that this glorious tradition started with a seemingly modest attempt—the attempt to get voices into print.

Notes

Introduction 1. Patrick Hanan gives this description of the use of the vernacular as the new genre convention in prose fiction: “If the vernacular is restricted to a small number of functions that seem to call for its use, it is also true those functions are confined to the vernacular. For example, the novel is, with the exception of a few virtuoso experiments, always in the vernacular, or at least in some intermediate language between classical and vernacular. On the other hand, the author’s or the editor’s preface, if there is one, is always in the classical, even though it is addressed to the same reader. This can be justified in a number of ways, but the immediate reason is simply genre convention” (The Chinese Vernacular Story, 10). 2. Gao Ru, Baichuan shuzhi, 82. 3. Yuan Hongdao, “Yu Dong Sibai shu,” excerpted in Zhu Yixuan, ed., Ming Qing xiaoshuo ziliao xuanbian, 613. 4. Shen Defu, Wanli yehuo bian, Juan 25. Excerpted in Zhu Yixuan, ed., Ming Qing xiaoshuo ziliao xuanbian, 615–616. 5. The chronological closeness of the early editions of Sanguo and Shuihu might be one of the reasons for the fact that the two works were often mentioned side-by-side in the bibliographical sources and appeared subsequently in pair editions, as Andrew Plaks has noted (The Four Masterworks, 366–367). 6. Many scholars have noted the difference between the language in Sanguo yanyi and Shuihu zhuan. Patrick Hanan, for instance, has made the observation that in the Sanguo yanyi, “the dual narrative methods [of the classical and the vernacular] are brilliantly synthesized,” while “a dozen or so stories” and Shuihu zhuan are “true vernacular fiction, as distinct from the tradition of fictionalized history, which generally uses a more Classical language” (The Chinese Vernacular Story, 8–9). 7. The work, compiled by Hong Pian, originally consisted of sixty stories, as the title of the collection indicates. The stories were lost until Ma Lian rediscovered fifteen of them in 1929, which soon appeared in a facsimile edition in Beijing under the title Qingping shantang huaben, after the name of Hong’s studio. Ma Lian acquired twelve more stories a few years later, five of which are

199

200

Notes to Pages 4–8

incomplete to different degrees. More recent editions of Qingping shantang huaben (Shanghai: Gudian Wenxue Chubanshe, 1957; Shanghai: Shanghai Guji Chubanshe, 1992) include all twentyseven stories and an article by A Ying describing the fragments of two other stories that he had discovered, “Feicui xuan” and “Mei xing zhengchun.” Whether the last two stories were part of the original Liushi jia xiaoshuo remains open to debate. While most of the stories are fundamentally in the vernacular, two of them, “Lanqiao ji” and “Fengyue xiangsi,” are in wenyan. So is the fragment of “Feicui xuan,” according to A Ying’s description. We know of only one vernacular story that was published earlier than Liushi jia xiaoshuo. The 1498 edition of Xixiang ji includes in its prefatory matter a story entitled “Zengxiang Qiantang meng,” the story of a young scholar and the ghost of a beautiful girl. See Xinkan qimiao quanxiang zhushi Xixiang ji (Guben xiqu congkan), 3a-6b. 8. See John DeFrancis, The Chinese Language: Fact and Fantasy, 71–73. 9. Ibid., 89–115. 10. See William Hannas, Asia’s Orthographic Dilemma, 101–124. 11. Ibid., 124. 12. For the percentage of the Chinese characters used today that are semantic-phonetic composites, Insup Taylor gives the estimate of 80–90 percent by citing three Chinese sources. See Insup Taylor, Writing and Literacy in Chinese, Korean and Japanese, 53. 13. Four Masterworks of the Ming Novel, 302–303. 14. The Singer of Tales, 132. In his later writings, especially in his article “Perspectives on Recent Work on Oral Literature,” however, Lord seems to be revising this view and shifting to a more receptive stance on the existence of “transitional” texts. 15. Literacy and Orality: Studies in the Technolog y of Communication, 141. Original italics. 16. I am relying on Einar Sveinsson’s account of the oral as well as written sources of the Njáls Saga in his Njáls Saga: A Literary Masterpiece. According to Sveinsson, the saga was written about 1280, some three hundred years after the events in the saga had actually occurred. While the writing of the saga might have been based on an oral tradition three hundred years old, a lost genealogical document written about 1180 could have shortened the span of oral transmission for some aspects of the saga by one hundred years. The written document in its own turn was based on preceding oral transmission. Additionally, oral tales about the death of Njall found their way into earlier sagas, including the earliest version of Landnamabok, which in turn might have launched another round of oral transmission until the oral versions again became fixed in a written form in Njáls Saga. 17. In his Uneasy Narrator: Chinese Fiction from the Traditional to the Modern, Henry Y. H. Zhao calls the earliest period of the development of Chinese vernacular fiction the “Rewriting period,” lasting from the Southern Song to the late sixteenth century, when Jin ping mei was written (10–11). Although Zhao does not define the “Rewriting period” in terms of the dynamics between orality and writing, it seems to me that it was coeval with the process of vernacularization, in which writers wrote on oral models and toward a full-fledged written vernacular. 18. Translated by J. H. Jackson (Shanghai: Commercial Press, 1937; reprint, Cambridge, MA: C & T, 1976).

Notes to Pages 9–13

201

19. Shijing jinzhu jinyi, Wang Yunwu, ed., and Ma Chiyin, annotator (Taipei: Taiwan Shangwu Yinshuguan, 1970), 405. 20. Jin Shengtan, “Shuihu zhuan xu er,” 7.

Chapter 1: Vernacularization before Shuihu zhuan 1. Bernhard Karlgren, The Chinese Language: An Essay on Its Nature and History, 57. 2. “Introduction” to Gudai Hanyu, Guo Xiliang et al., eds., 2. 3. John DeFrancis, Visible Speech, 264. 4. Lü Shuxiang, “Wenyan and baihua,” 73–75. 5. Xu Shen (58–147), Shuowen jiezi (Sibu beiyao ed.), 9/1/7a. 6. I take “diao” to be in the sense of decorating with colorful paintings. Both James J. Y. Liu (Chinese Theories of Literature, 22) and Vincent Shih ( The Literary Mind, 9) translate “diaose” as “sculptured colors.” 7. Wenxin diaolong zhu, “Yuan Dao,” 1–2. For the translation, cf. Vincent Shih, The Literary Mind and the Carving of Dragons, 8–9 and James J. Y. Liu, Chinese Theories of Literature, 22–23. 8. Cang Jie’s legendary invention of the Chinese script is mentioned in Lüshi Chunqiu. Gao You (fl. 205) states in a note that Cang Jie “imitated the markings of birds in the invention of writing” (Lüshi Chunqiu, in Zhuzi jicheng 6: 203). In Huainanzi, Cang Jie’s invention is said to be marked by a rain of grain from heaven and cries of ghosts in the night. Gao You suggests in a note that the character “gui” ( ghosts) might be a typo for “tu” (rabbits): the rabbits wailed at the prospect that their hair would be plucked to make writing brushes ( Zhuzi jicheng 7: 116–117). In his Lun heng, Wang Chong (27–97) challenges the credibility of raining grain and crying ghosts as ominous phenomena incurred by the creation of writing symbols. He argues that, since the making of the characters was an act in accord with heaven and earth, it should never be the cause for any such strange happenings ( Zhuzi jicheng 7: 52). In Wenxin diaolong, Liu Xie seems to be synthesizing the two traditions about Cang Jie from Lüshi Chunqiu and Huainanzi: “As the markings of birds were clear and illustrative, characters were created. . . . When Cang Jie brought the written symbols into existence, ghosts cried and grain rained from heaven” ( Wenxin diaolong zhu, 2: 623). 9. Kong Yingda (574–648), annotator, Zhouyi zheng yi (Sibu beiyao ed.), “Xi ci,” 8/3a. It is suggested elsewhere in “Xi ci” that the Sixty-Four Hexagrams and the Eight Trigrams had different origins, but both from external nature. The Hexagrams are said to have been based on the auspicious and ominous signs in heaven, while the Trigrams were thought to be imitations of the patterns on the body of the dragon-horse and on the back of the divine tortoise that appeared respectively on the Yellow River and the Luo River. See Zhouyi zheng yi, 7/17b. 10. In his Literate Revolution in Greece, Eric Havelock considers the visual development of the written signs as a hindrance to instantaneous verbal communication: Strictly speaking, written orthography should behave solely as the servant of the spoken tongue, reporting its sounds as accurately and swiftly as possible. It need not and should not have a nature of its own, and the Greek sys-

202

Notes to Pages 13–16 tem when it was invented seemed to have conformed to this specification. But, as we shall see, the artistic fascination of the Chinese with the calligraphy of the ideogram has had its counterpart in the development of scripts and their elaboration in European and Arabic countries. This visual development of the written signs has nothing to do with the purpose of language, namely instantaneous communication between members of a human group. (53)

11. William Hannas, Asia’s Orthographic Dilemma, 106. 12. Wenxin diaolong zhu, 1: 2. The English paraphrase of the term is by James J. Y. Liu, Chinese Theories of Literature, 23. 13. Wenxin diaolong zhu, 1: 2. 14. Ruan Yuan, Yanjingshi ji (Congshu jicheng ed.), 567. Quoted and translated in James J. Y. Liu, Chinese Theories of Literature, 104. 15. Laozi, Daode jing, chapter 1. Wang Bi (226–249), annotator, Laozi zhu, in Zhuzi jicheng, 3: 1. 16. Zhuzi jicheng, 3: 34. 17. Zhuangzi, “Tian Dao.” Zhuangzi jijie, in Zhuzi jicheng, 3: 87. 18. Kong Yingda, annotator, Zhouyi zheng yi, “Xi ci,” 7/18a. 19. Ibid. 20. Wen fu yizhu, 18. 21. Zhuangzi, “Tian Dao.” Zhuangzi jijie, in Zhuzi jicheng, 3: 87. 22. Wen fu yizhu, 42. 23. Wenxin diaolong zhu, 2: 494–495. 24. Kong Yingda, annotator, Mao shi zheng yi (Sibu beiyao ed.), 1: 1/3a. 25. Stephen Owen, Traditional Chinese Poetry and Poetics: Omen of the World, 58. 26. Ibid., 45. 27. Zuo zhuan, “Xianggong ershiwu nian” (Twenty-Fifth Year of Duke Xiang). In Kong Yinda, annotator, Zuo zhuan zhushu, 36/7b. 28. The ten genres listed in Lu Ji’s Wen fu are shi, fu (a descriptive essay form), bei (monumental inscription), lei (elegy), ming (tablet inscription), zhen (epigram), song (panegyric), lun (expository essay), zou (memorial), and shuo (polemic). In Liu Xie’s Wenxin diaolong, the thirty-four genres, which include all ten from Wen fu, are discussed successively in twenty chapters, from chapter 6 to chapter 25. 29. This refers to the two guwen (ancient language) movements in the ninth and sixteenth centuries. The most prominent names involved in the first guwen movement were Han Yu (768–824) and Liu Zongyuan (773–819). Tired of the prevalent pian wen style in prose writing with rigid rules on parallel structure and tonal juxtaposition, Han and Liu advocated a revival of “ancient prose,” a return to the freer and more flexible prose style of the Qin and Han periods. The second guwen movement seven centuries later, although usually considered less “revolutionary,” can almost be called a mirror repetition of the first. At a time when the styles of prose writing exem-

Notes to Pages 17–21

203

plified by the so-called Eight Masters of Tang and Song were advocated as an antidote to the stifling bagu wen (Eight-Legged Essay), Li Mengyang (1473–1530)—with the support of six other members of his school known as the Former Seven Masters (Qian qi zi)—argued for the superiority of the even older models of prose writing from the Qin and Han periods. 30. Collected in Gudai hanyu, 3: 745–751. 31. Since the text of Shishuo xinyu as we have it today is based on a twelfth-century edition, we cannot be entirely sure that those vernacular elements were not added later. 32. For an account of the impact of early Buddhism—especially the translation of Buddhist texts—on the development of the written vernacular, see Victor H. Mair, “Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages.” 33. Dating of the Chan yulu texts is a task of formidable difficulty. From the extant editions it is usually very hard to tell to what extent the original text has been tampered with by their editor(s). With those yulu texts that have more than one edition extant, scholars have indeed noticed massive variations among the different versions. See Du Jiwen and Wei Daoru, Zhongguo Chanzong tongshi, 18ff. Obviously, extensive editorial tampering may compromise the authenticity of the written vernacular in a yulu text as the product of an earlier age. 34. Bi yan lu, No. 12. In Chanzong yulu jiyao, 722. 35. Ibid. 36. The assessment of the influence of the Chan yulu on its Neo-Confucian counterpart is an issue open to debate, their formal similarities notwithstanding. Song Neo-Confucians were no doubt familiar with the form of Chan yulu, but that does not necessarily mean their own form of yulu was derived from the Buddhist model. For one thing, yulu, even though not called this, had been a genre of writing in classic Confucianism centuries before Chan Buddhism ever came into existence, the best-known example being, of course, the Confucian Lunyu (The Analects). But of course it is debatable whether the language in the Lunyu faithfully reflects the spoken language of Confucius’ time. 37. For an essay on the relationship between Neo-Confucian yulu and the traditional form of commentary on the classics and on the implication of the use of colloquial language in Neo-Confucian exegeses, see Daniel K. Gardner, “Modes of Thinking and Modes of Discourse in the Sung: Some Thoughts on the Yü-lu (‘Recorded Conversations’) Texts.” 38. See Wing-tsit Chan, Chu Hsi: New Studies (Honolulu: University of Hawai‘i Press, 1989), 375. 39. Forty such tablet inscriptions are collected in Yuandai baihua bei. More imperial decrees and other bureaucratic writings of the Yuan period are to be found in Da Yuan shengzheng guochao dianzhang, which appear in a vernacular somewhat more accessible to the modern reader. Included in the volumes are also accounts of exemplary judicial cases, which are less encumbered by transcriptions of Mongolian words. 40. Zhu Dexi, a leading Chinese linguist, put the dates for the original editions of both textbooks around the middle of the fourteenth century. See Zhu Dexi’s article, “Laoqida yanjie, Pu tongshi yanjie shuhou.” 41. Wang Jide (?-1623), for example, deplores the dilution of the regional distinction of nanqu

204

Notes to Pages 21–22

(southern drama) by the linguistic influence from its northern counterpart: “The use of words in nanqu and beiqu (northern drama) should not be mixed up. Nowadays, words such as zhe, wu, nin, and zan appear in nanqu. . . . Those are indeed grossly inappropriate forms.” See Wang Jide, Qu lü, 148. 42. Although there has been a more indulgent and broader definition of “bianwen” under which the term is used to refer to virtually any texts of popular literature discovered at Dunhuang, including jiang jing wen (sutra lectures), huaben, and other genres, I am following the more stringent and narrower definition suggested by Victor Mair. About different definitions and the corpus of bianwen, see Victor Mair, T’ang Transformation Texts, 9–32. 43. Victor Mair, “Lay Students and the Making of Written Vernacular Narrative: An Inventory of Tun-huang Manuscript,” 5. 44. This formulaic phrase appears most frequently in “Hanjiang Wang Ling bian,” “Li Ling bianwen,” and “Wang Zhaojun bianwen.” Although “ruowei chenshuo” seems the standard form, it sometimes appears in variants such as “ruowei,” “chu ruowei,” and “chu ruowei chenshuo.” 45. T’ang Transformation Text, 88. Being “several generations removed” from their oral precursors by no means abrogates or discounts the oral origins of many of the bianwen texts. Rather, it is one more example that reinforces the argument I will unfold: Before written vernacular, especially vernacular prose, could be written with ease and confidence, the textualization of an oral discourse of an extended length might have to be a process involving continuous interaction between orality and writing. 46. About the possible influence of Tang storytelling on popular entertainment in later periods, Victor Mair has pointed out the shared feature of “picture-storytelling” in the bian performance and the pinghua (expository tale). Based on that evidence, Mair proposes that “it is possible to say with some assurance . . . that expository tales were the direct descendants in the Yuan period of transformation storytelling. We might even go so far as to say that pinghua was essentially bian with a Sinicized name” (Painting and Performance, 3–4). According to Mair, the recitation from precious scrolls, or baojuan, may also have derived from bian storytelling (Painting and Performance, 8–9). 47. In the text of “Zhao Lingzhi dielianhua guzici,” printed as an appendix to a modern edition of Wang Shifu’s Xixiang ji (Shanghai: Shanghai Guji, 1978), no prose passages are presented. Instead, for each prose passage only the beginning and ending phrases in Yuan Zhen’s tale are given, which testifies to the verbatim duplication. 48. Literally, the formula can be translated as: “Respectfully, I ask my accompanist to resume the previous tune.” Presumably it was used when the storyteller shifted from prose narration to singing. The formula also appears in some bianwen texts and the story “Wenjing yuanyang hui,” collected in Qingping shantang huaben, which could have been related, in one way or another, to some kind of a storytelling script—most likely a guzici. 49. Both “gong” and “diao” are terms used for the names of musical modes in ancient China. Theoretically there were eighty-four modes, including twelve gong and seventy-two diao, although in actual performance only a much smaller number of modes were used. The term “zhugongdiao”

Notes to Pages 22–25

205

literally means a combination of different gong modes and diao modes. The name is derived from the musical organization of the genre. Different from guzici, where all the songs belong to the same musical mode, in zhugongdiao any two succeeding sets of verses are usually set in two different musical modes. The form of zhugongdiao, according to the Southern Song scholar Wang Zhuo in his Bi ji man zhi, was invented in the Northern Song (960–1126) by Kong Sanzhuan, a raconteur from Zezhou (115). For a comprehensive study of the genre, see Wilt L. Idema, “Data on the Chu-kung-tiao.” 50. The extant nanxi text Zhang Xie zhuang yuan begins with a short narrative in a prosimetric form. Although the narrative calls itself a zhugongdiao ( Xiao Sun tu deng sanzhong, 13b), I do not include it in the group of extant exemplars of the genre simply because it is not comparable with the other three texts in length. 51. In Zhong Sicheng’s Lu gui bu, Dong Jieyuan is listed first among the names of “renowned masters of the previous generation” (103). Most likely Jieyuan was not the man’s name. As the title for the first place in the preliminary level of the civil service examinations, it was often used as a generic honorific for scholars, especially for those who did not earn a degree in the examinations. 52. Ming Jiajing ben Dong Jieyuan Xixiang ji, 1, 8–9. 53. In his “Song-Jin-Yuan zhugongdiao kao,” Zheng Zhenduo comments on Xixiang ji zhugongdiao that the “style of its prose, just as in Zhao Delin’s (Zhao Lingzhi) Dielianhua guzici, is completely from guwen” (Zhongguo wenxue yanjiu xinbian, 2: 854). 54. Xinkan qimiao quanxiang zhushi Xixiang ji, 1: 36/a-37/a. For an English translation of this edition, see Stephen H. West and Wilt L. Idema, The Moon and the Zither. 55. Little is known about Wang Shifu’s life. About Xixiang ji’s authorship, there is an abandoned theory on the collaboration of Wang Shifu and Guan Hanqing. That theory must have been active during the sixteenth century, as Wang Shizhen (1526–1590) recorded in his Qu zao (Elegance of Arias) that some of his contemporaries believed that the ending scenes of Xixiang ji had been composed by Guan Hanqing (31). The alleged collaboration suggests that the two playwrights must have been contemporaries. Since Guan Hanqing is believed to have lived in the late Jin and early Yuan (See Wang Guowei, Song-Yuan xiqu shi, 140), Wang Shifu could not have lived much earlier or later. In Zhong Sicheng’s Lu gui bu, both names are listed as “deceased renowned masters and men of talents of the previous generation,” also suggesting that the two were contemporaries (104, 109). Lu gui pu, however, attributes Xixiang ji zaju to Wang Shifu alone, and so does Zhu Quan’s Taihe zheng yin pu in the early Ming. 56. The earliest known but nonextant text of Xixiang ji zaju was included in the Yongle dadian, a vast assemblage of collectanea compiled in the first decade of the fifteenth century. Zhu Youdun (1379–1439), an early Ming dramatist, could have published another edition of the play, as Ling Mengchu (1580–1644) claimed that his edition of Xixiang ji was faithfully based on that of Zhu’s. That claim, however, has been called in question by W. L. Idema in his “Zhu Youdun he Xixiang ji” (Zhang Zugang, trans. In Han Sheng ed., Xixjang ji xinlun, 131–140). In their introduction to The Moon and the Zither, Stephen H. West and Wilt L. Idema call Ling’s assertion

206

Notes to Pages 26–27

“just a sales pitch,” and they believe that Ling might have based his edition on a copy of the 1498 edition instead (15). Apart from the four pages discovered in 1980 at a bookstore in Beijing that might be remnants from a mid-fifteenth-century edition, the complete Hongzhi edition dated 1498 is the earliest textual exemplar of the play. For an account of the editions of Xixiang ji zaju from the Ming up to the modern times, including its translations in foreign languages, see the West and Idema introduction cited above (1–27), and also see Han Sheng, “Xixiang ji gujin banben mulu jiyao.” 57. The scarcity of spoken lines in the Yuan-edition zaju plays can sometimes be an annoyance to scholars. David Hawkes, for instance, expresses his disappointment by saying that “Our scholarly gratification at the discovery of an early text must not blind us to the possibility that it may be a very bad text” (“Reflections on Some Yüan Tsa-chü,” 81). 58. In the modern edition of JiaodingYuankan zaju sanshi zhong (Taipei, 1962), the text of a Yuan-edition zaju takes up only ten to twelve pages on average. Guan Zhang shuang fu Xishu meng, which contains no binbai, is only seven pages long. Wang Shifu’s Xixiang ji consists of five parts. In the 1978 reprint (Shanghai Guji) of the Hongzhi edition, each part occupies twenty to twentysix pages, not counting the space for the notes. Even if we consider each part as a regular play in its own right, its length still more than doubles the regular length of a Yuan-edition zaju. 59. Zheng Zhenduo, “Xixiang ji de benlai mianmu shi zenyangde?” 610. 60. Quoted in Jiang Xingyu, “Panguo Shuoren Xixiang dingben de ‘jicheng’ he ‘chuangzao’,” in his Xixiang ji kaozheng, 121–137, 128–129. Panguo Shuoren, according to Jiang Xingyu, was the pseudonym for Xu Fenpeng (121). 61. About the editions of the Shuihu zaju, see Wang Xiaojia, Shuihu xi kaolun, 21. Also see Fu Xihua, Yuandai zaju quanmu. Y. W. Ma has noted an interesting and significant fact about the dating of the Shuihu plays: All the extant texts of the Shuihu plays attributed to Yuan playwrights actually date much later than the two plays written by Zhu Youdun, both of which appeared in 1433. See Y. W. Ma, “Cong Zhao’an bufen kan Shuihu zhuan de chengshu guocheng,” 148. 62. The names and nicknames of the bandits in Zhu Youdun’s plays are closer to those in Xuanhe yishi than to Shuihu zhuan. Also, as in Xuanhe yishi, the total number for the bandit chieftains in the plays is 36, rather than 108 as in Shuihu zhuan. However, the difference of the content in a Shuihu play from that of Shuihu zhuan, unless combined with other evidence, does not necessarily establish the priority of the play over the narrative. For instance, the plot in Li Kaixian’s (1502–1568) Baojian ji, based on the story of Lin Chong, deviates from Shuihu zhuan substantially. Early in the play, Lin Chong, before he is forced to go to join the Liangshan band, claims that he is the one who has killed Fang La, which runs diametrically contrary to the long tradition from Xuanhe yishi to Shuihu zhuan that Liangshan bandits defeat Fang La after their surrender to the emperor. Li Kaixian, however, most likely wrote the play after Shuihu zhuan came out in print, as evidenced by his account in Ci Xue (286) of several literati scholars’ discussion of Shuihu zhuan. Even a play as late as Ling Mengchu’s Song Gongming nao yuanxiao zaju features a plot radically different from that of Shuihu zhuan: The courtesan Li Shishi’s royal patron, the emperor Huizong (r. 1101–1125), has a rival in the young poet Zhou Bangyan, who is banished out of the capital as

Notes to Pages 27–28

207

a punishment by the jealous emperor. When the sovereign pays another visit to Li Shishi, the courtesan is away from home seeing her unfortunate lover off, leaving the Liangshan chieftain Yan Qing at her home entertaining the emperor. Obviously such a plot must be from some source other than Shuihu zhuan. In fact, it is based on an anecdote contained in the Southern Song scholar Zhang Duanyi’s Gui er ji (Excerpted in Shuihu zhuan ziliao huibian, 31–32). Xuanhe yishi also contains a different episode about the emperor’s jealousy, in which Li Shishi’s other lover, Jia Yi, is almost put to death (66–88). In the case of Zhu Youdun’s Shuihu plays, however, their precedence to Shuihu zhuan in print does have a strong corroboration. Judging by Zhu Youdun’s preface to his Baozi heshang zi huansu, collected in Shuihu ziliao huibian (Ma Tiji, ed., 83), it seems that he had no knowledge whatsoever of Shuihu zhuan as a book; and given the fact that he was one of the leading scholars of his time, it would be unlikely that he should have been unaware of the book if it had indeed come out in print. For discussions of Zhu Youdun’s Shuihu plays and their implication for the dating of the earliest edition of Shuihu zhuan, see George A. Hayden, “A Skeptical Note on the Early History of Shuihu zhuan,” W. L. Idema, “Zhu Youdun’s Dramatic Prefaces and Traditional Fiction,” and Y. W. Ma, “Cong zhao’an bufen kan Shuihu zhuan de chengshu guocheng.” I will return to this issue in chapter 4, where the frequent instances of linguistic features in Shuihu zhuan typical of a period earlier than Zhu Youdun’s time will be presented to support the argument that the textualization of Shuihu zhuan was a long process that had started much earlier than Zhu Youdun’s plays but did not reach its fullest form until later. 63. W. L. Idema, The Dramatic Oeuvre of Chu Yu-tun, 37. 64. See Yu Weimin’s foreword to Song Yuan si da xiwen duben, 1–2. 65. Zhao Jingshen preferred to consider Gao Ming as merely an adapter of the play, on the grounds that the title Cai Bojie Pipa ji is mentioned in Xu Wei’s Nanci xulu as a nanxi of Song-Yuan origin. Zhao, however, noticed that the plot in the older and nonextant version, where Cai Bojie is struck to death by lightning as a punishment for his betrayal of his parents and wife, might have been drastically different from the later version of the play attributed to Gao Ming. See Zhao Jingshen, Yuan Ming nanxi kaolue, 43–44. 66. Wang Guowei suggests that Jingchai ji might be written by the Ming prince Zhu Quan and Sha gou ji by the early Ming scholar Xu Zhongyou, but neither attribution has been universally accepted. The attribution of Sha gou ji is particularly problematic, since a nanxi entitled Sha gou quan fuxu, which may well have been another name for Sha gou ji, is mentioned in the Yongle dadian text of Huanmen zidi cuo lishen ( Xiao Sun tu deng sanzhong, 56b). Wang, however, questions some Ming scholars’ attribution of Bai yue ting to the Yuan playwright Shi Hui. See Wang Guowei, Song Yuan xiqu shi, 212–222. 67. For the different Ming editions of the four xiwen plays, see Song Yuan si da xiwen duben, 5–6, 178, 271–272, and 392. 68. See Liu Xiaopeng, “Yongle dadian san ben xiwen yu wu da nanxi de jiegou bijiao.” 69. Photoreprint editions of both texts are collected in Guben xiqu congkan, entitled respectively Yuanmo Cai Bojie Pipa ji and Li Zhuowu piping Pipa ji.

208

Notes to Pages 28–31

70. In the modern typographical edition of Song Yuan si da xiwen duben (1988), each play runs over one hundred pages. 71. About the language style in chuanqi plays of the late Ming period, Ling Mengchu complains in his Tanqu zazha: “Today’s plays not only compete with each other in the ornamentation of the arias, but also contend with each other in the baroque luxuriance of the spoken lines. Even for a common dialogue, no casual and colloquial words [xianyu] are used. Instead, it has to be rendered in stringent parallelism” (259). 72. According to Meng Yuanlao, a storyteller called Huo Si was particularly good at telling stories of san fen (tripartition), which referred to the period of the Three Kingdoms, and another with the name of Yin Chang specialized in the stories of Wudai shi (History of the Five Dynasties) (Dong jing menghua lu, 30). 73. For Wudai shi pinghua and Xuanhe yishi, a general consensus has been reached that they most likely date from the Yuan period. The date for Lianggong jiu jian remains obscure, but some scholars suggest that it cannot be a Song text. See Zhongguo wenxue shi (Shanghai: Fudan Daxue, 1996) 3: 137. Of the two versions of Da Tang Sanzang qujing shihua (or rather the same version with two different titles, the other one being Da Tang Sanzang fashi qujing ji), Zhang Peiheng puts the date in the Yuan period (“Guanyu xiancun de suowei ‘Song huaben’,” 164), while Patrick Hanan considers them “both probably of the Southern Song” (The Chinese Vernacular Story, 7). 74. Chinese Vernacular Fiction, 85. 75. One instance of a pinghua’s verbatim incorporation from a classical source is in Qin bing liuguo pinghua, in which Li Si’s long memorial to Qin Shihuang, arguing against the emperor’s decree to dismiss officials of nonnative origins, is copied verbatim from Shiji. See Song Yuan pinghua ji, 2: 641–643. 76. The number could be 150 if the prologue to “Song Sigong da’nao ‘Jinhun’ Zhang” is counted as a separate story, as Patrick Hanan points out (The Chinese Short Story, 16). 77. Among the thirty-four “early stories,” Hanan puts fourteen in “Group A,” the “earliest” of the “early,” which were “mostly written in the Yuan.” “Group C” consists of twelve stories, which were “mostly written in the early Ming.” The other eight in “Group B” are more difficult to date, but they are unlikely to be earlier than Group A or later than Group C. See The Chinese Short Story, 152–169. 78. Such accounts are seen in Meng Yuanlao’s Dong jing menghua lu, Naide Weng’s Ducheng jisheng, Xihu Laoren’s Xihu Laoren fansheng lu, Wu Zimu’s Meng liang lu, Zhou Mi’s Wulin jiushi, and Luo Ye’s Zuiweng tanlu. The first two works were written in the Southern Song period, while the other four were perhaps composed in the late Song or, more likely, the early Yuan. 79. Ducheng jisheng, 97. 80. See Patrick Hanan, The Chinese Short Story. Also see Tan Zhengbi, San yan Liang pai ziliao. 81. For the types of subject matter in some of the earliest vernacular stories and their possible kinship to the varieties of subject matter in the oral genre of xiaoshuo, see Patrick Hanan, The Chinese Short Story, 170–174. Even for those early stories whose stuff material is not referred to

Notes to Pages 31–38

209

in either Ducheng jisheng or Zuiweng tanlu, Hanan proposes the hypothesis that they might correspond to a different oral period, later than the one reflected in these two works. 82. Ducheng jisheng must have been written no later than 1235, as it has a preface by Naide Weng dated that year. If a vernacular story reflects the oral storytelling of the period described in Ducheng jisheng and appeared in the textual form at the end of the Yuan, the oral version and the written version of the story would have been separated by at least 130 years. 83. Hu Shi gudian wenxue lunji, 709–710. 84. Yan Dunyi, Shuihu zhuan de yanbian, 151. 85. Zang Maoxun, Yuanqu xuan, 1: 3. 86. Wang Jide, Qu lü, 148. 87. Li Yu, Xianqing ouji, in Ming Qing zaji, 3: 47. 88. See Wang Guowei, Song Yuan xiqu shi, 97–98. 89. Wu Mei, Gu qu zhu tan, 62. 90. Gu Xuejie, Yuan Ming zaju, 29. 91. Most of the zaju plays listed in Lu gui bu, Lu gui bu xubian, and other bibliographical sources are attributed to definitive playwrights. Attributions, of course, do not always indicate authentic authorship, but for early vernacular fiction, even attributions are rare. None of the stories in Qingping shantang huaben, for example, is attributed to any particular name. For nanxi, largely due to its much less esteemed status during the Yuan and Ming periods, many plays also remain anonymous. 92. Qian Xuantong, “Rulin waishi xin xu,” in Rulin waishi huijiao huiping ben, 2: 785–786.

Chapter 2: Told or Written: That Is the Question 1. Chinese Vernacular Fiction: The Formative Period, lxiv. 2. See Sheldon Hsiao-peng Lu, From Historicity to Fictionality: The Chinese Poetics of Narrative. 3. “Hou Meng zhuan,” Song shi, 32: 11114. 4. Song shi, 32: 11141. 5. Excerpted in Shuihu zhuan ziliao huibian, 12–14. 6. The eight types of oral stories in Zuiweng tanlu are as follows: lingguai (spirits and demons), yanfen (rouge and powder), chuanqi (marvels), gong’an (court cases), podao (broadsword), ganbang (staff ), shenxian (immortals), and yaoshu (sorcery). This classification is somewhat different from that in Meng liang lu, which does not contain the last two categories, shenxian and yaoshu (Dong jing menghua lu wai si zhong, 312). In Ducheng jisheng, podao, ganbang, and faji biantai (upstarts and nouveau riches) are given as subdivisions of gong’an. As in Meng liang lu, the list does not include shenxian and yaoshu (Dong jing menghua lu wai si zhong, 98). 7. For the titles of the stories that may have belonged to the Shuihu cycle and of other stories, see Luo Ye, Xinbian Zuiweng tanlu, 4b. 8. Gong Shengyu’s “Song Jiang sanshiliu zan” is included in Zhou Mi’s (1232–1298) Guixin zashi xuji (Congshu jicheng ed., 276–287). It is also collected in Shuihu zhuan ziliao huibian (21–24). In the prefatory note, Gong Shengyu mentions portraits of the Liangshan bandits drawn by Li

210

Notes to Pages 38–40

Song. It was therefore possible that his verses were intended to accompany Li’s portraits. But since Gong was himself a famous painter, it was equally possible that his verses were meant for portraits drawn by himself. 9. Shuihu zhuan ziliao huibian, 21. 10. The dating of the work is difficult because the compiler may have drawn from a variety of sources, as evidenced by the stylistic inconsistency in different parts of the work. On the title page in the Qing edition published by Shiliju congshu (facsimile copy collected in Guben xiaoshuo jicheng), it claims to be a reprint of a Song edition. Obviously, publisher Huang Pilie (1763–1825) believed in Song authorship. In a postscript to that edition, “Chongkan Songben Xuanhe yishi ba” ( Xuanhe yishi [Guben xiaoshuo jicheng], 199ff), the pseudonymous writer Shanhaiju Zhuren supports the dating by citing the text’s avoidance of the character “dun,” which was part of a Song emperor’s name. In the prefatory note to Xuanhe yishi in the modern edition of Song Yuan pinghua ji, the work is also said to have been composed in the Song period (255). The strong patriotic emotion about the conquest of the Northern Song by its belligerent neighbor Jin and the humiliation suffered by the two abducted Song emperors may indeed suggest that part of the work could have been based on Song sources. On the other hand, some other internal evidence points to a later dating. For instance, the first emperor of the Song dynasty, Zhao Kuangyin, and his father Zhao Hong’en are referred to simply by their names, which was contradictory to the general practice of Song writers. The prediction that the Song dynasty would have to move its capital from the north to the south also suggests that the work was probably written after the end of the Song. Everything taken into account, the compiler might be a subject of the Song who survived the dynasty and composed the book in the early Yuan. Richard Gregg Irwin has put the dating of the work around 1300 (The Evolution of a Chinese Novel, 25–26), which is probably as close an estimate as we can get. 11. Xuanhe yishi, 51–61. 12. Literally, the title Xuanhe yishi means “Unrecorded Events during the Xuanhe Reign.” 13. See Hennessey, “Classical Sources and Vernacular Resources in Xuanhe yishi,” Chinese Literature: Essays, Articles, Reviews 6 (1984): 33–52. 14. In sources from the Song, the dating of the quelling of the Fang La rebellion is not consistent. Xu Mengxin’s San chao bei meng huibian (Excerpted in Shuihu zhuan ziliao huibian, 3–5) puts it in the fourth year of the Xuanhe reign, while Yang Zhongliang’s Tong jian changbian jishi benmo (excerpted in Shuihu zhuan ziliao huibian, 11) dates it to the third year. Since the compiler of Xuanhe yishi puts the whole Shuihu section under the heading of “Xuanhe sinian,” obviously he considered that it was in the fourth year that Song Jiang surrendered and joined the campaign against Fang La. 15. Between Gong Shengyu’s “Song Jiang sanshiliu zan” and Xuanhe yishi, it is difficult to decide which was earlier. Both Gong Shengyu and Zhou Mi, who included “Song Jiang sanshiliu zan” in his Guixin zashi xuji, lived long lives and died about twenty years after the demise of the Southern Song. It could, like Xuanhe yishi, have been composed in the early Yuan period. I, however, tend to believe in the priority of “Song Jiang sanshiliu zan” for three reasons: (1) Gong

Notes to Pages 40–42

211

Shengyu states in the prefatory note that he did not see anything written about the band apart from the note in “Hou Meng zhuan.” (2) Such prominent Liangshan figures as Gongsun Sheng and Lin Chong are conspicuously missing from Gong Shengyu’s list. (3) Unlike the case in Xuanhe yishi, there is no internal evidence in “Song Jiang sanshiliu zan” against a possible Song dating. Deng Deyou, however, thinks otherwise. He suggests that Xuanhe yishi may have preceded “Song Jiang sanshiliu zan” on the ground that the names of the bandits in the latter are more consistant with those in Shuihu zhuan. See his article “Tan Xuanhe yishi yu ‘Song Jiang sanshiliu zan’ de shidai xianhou wenti.” 16. Hu Yinglin (1551–1602) observes in his Shaoshi shanfang bicong that the language in the Shuihu segment of Xuanhe yishi is “crude and coarse” and therefore must be from the “popular lore” (lüyan sushuo) (Excerpted in Shuihu zhuan ziliao huibian, 87). 17. Shuihu xiqu ji, 1: 10–11. 18. See George A. Hayden, “A Skeptical Note on the Early History of Shuihu chuan.” 19. The basic line of Song Jiang’s self-introducing monologue runs as follows: Song Jiang was previously a clerk in Yuncheng County. When he was drunk he killed a woman called Yan Poxi. After giving himself up, he was exiled to Jiangzhou but was rescued by Chao Gai, the leader of the Liangshan band. After Chao Gai’s death during the third attack on the Zhu Family Village, Song Jiang was made the leader of the rebels. This summary is somewhat different from what happens in Shuihu zhuan, but it is for the most part consistent in all Shuihu plays attributed to the Yuan authors. In Shuang xiangong and Yan Qing puyu, the monologue includes the additional event of Song Jiang knocking over a candle stand after the murder of Yan Poxi and consequently setting the house on fire. 20. In the play, the daughter of an old tavern owner is abducted by two knaves who impersonate the Liangshan chieftains Song Jiang and Lu Zhishen. Believing his fellow rebels to be the offenders, the righteous Li Kui insists that Song Jiang and Lu Zhishen be identified by the tavern owner and his fellow villagers. As he is finally convinced of his comrades’ innocence, Li Kui goes to see Song Jiang with a bundle of sticks tied to his back and asks for a good flogging. The play ends with Li Kui’s capture of the two true culprits. What happens in the play parallels the second half of chapter 73 of Shuihu zhuan, where Chai Jin, instead of Lu Zhishen, is the alleged accomplice of Song Jiang. 21. Judging by the titles, however, some nonextant Shuihu plays might have contained episodes that are included in Shuihu zhuan. The plots in Hei Xuanfeng qiao duan’an (Black Whirlwind Plays a Judge) and Hei Xuanfeng qiao jiaoxue (Black Whirlwind Plays a Teacher) might parallel the two episodes contained in chapter 74 of Shuihu zhuan. In the first episode, Li Kui wears the robe of a county magistrate in the yamen, ordering the clerks to play the plaintiff and defendant. In the second, he intrudes into a Confucian school and scares the students away. It is also possible that Yan Qing she yan (Yan Qing Shoots Wild Geese) and Zhe danr Wu Song da hu (Wu Song Kills a Tiger with a Broken Pole) correspond respectively to what happens in chapter 110 and chapter 23 of Shuihu zhuan. 22. See his “Shuihu zhuan jiuben kao,” 121–134.

212

Notes to Pages 42–45

23. In the early Shuihu plays attributed to Yuan authors, all the bandit heroes are those who were to become major chieftains in Shuihu zhuan. Yan Shun, who is a minor chieftain in the narrative, appears in the Yuanqu xuan version of Yan Qing puyu, but the name was only an expedient change from Yan Da in an earlier version. 24. Huan laomo, however, is an exception. In the play, Liu Tang and Shi Jin, two would-be Liangshan chieftains, are constables in Dongping Prefecture and join the Liangshan rebellion only at the end of the play. The action in another extant play, Song Gongming pai jiugong bagua zhen, is about the expedition against the Liao and therefore after Song Jiang’s surrender to the emperor, but that play is probably of late Ming authorship. 25. Shuihu zhuan de yanbian, 113. 26. The earliest play based on the episode is probably Sha Xi (The Killing of Yan Poxi), a kunqu attributed to the late Ming playwright Xu Zichang. See Ma Tiji, Shuihu shulu, 464. 27. Some of the playwrights’ inventions were incorporated in Shuihu zhuan. Indeed, it may be one of the reasons for the highly episodic narrative structure in the chapters following the final ranking, as suggested by Y. W. Ma in his “Cong zhao’an bufen kan Shuihu zhuan de chengshu guocheng.” 28. Lu gui bu, 110, 112, 114. 29. Ibid., 172. 30. Ibid., 284. 31. For more about the play Xiyou ji, see Yan Dunyi’s article “Xiyou ji he gudai xiqu de guanxi.” 32. A Yuan version of the play is included in Jiaoding Yuan-kan zaju sanshi zhong, 413–425. 33. “Shuihu zhuan de yanhua,” 110. 34. Sun Kaidi suggests that some of the flaws in the narrative plot such as the basically duplicate campaigns against Zhujia Zhuang and Zengtou Shi are due to its drawing from the northern cihua in addition to its main source of the southern cihua. See his article, “Shuihu zhuan jiuben kao.” Sun’s hypothesis of a Shuihu zhuan cihua is reiterated by Richard Gregg Irwin in his Evolution of a Chinese Novel, 39–43. 35. The text is available in the modern edition of Da Tang Qin Wang cihua (Beijing: Wenxue Guji Kanxingshe, 1955), and a photoreprint of a late Ming edition is included in Guben xiaoshuo jicheng. 36. At the beginning of the story the posited storyteller addresses his audience: “Today please listen to me telling a cihua called ‘The Pearl Shirt.’ ” See Yushi ming yan (Shanghai: Shanghai Guji, 1992), 1. It is perhaps an example of using the word “cihua” synonymously with “huaben.” Qian Xiyan, Feng Menglong’s contemporary, states in his Xi xia: “At the beginning of each cihua there is a segment of qingke [a treat], which serves as desheng lishi touhui [an advertising prologue]. That was a trick of people in the Song times to analogize this with that and to make something out of nothing” (excerpted in Shuihu zhuan ziliao huibian, 150–151). The formal features that he describes here are obviously those of some huaben. Qian, too, might be using the word “cihua” not to refer to that particular type of prosimetric storytelling but as another name for “huaben,” the more general term for stories.

Notes to Pages 45–46

213

37. The Chinese Short Story, 44. 38. It is recorded in Yuan dianzhang that, except for licensed professional entertainers, peasants were not allowed to perform cihua ( banshuo cihua). See Yuan dianzhong, No. 57, 780. 39. For details of the discovery of the texts and about their alleged fifteenth-century owners, the wealthy Xuan family, see Anne E. McLaren, Chinese Popular Culture and Ming Chantefables, 15–31. 40. Ming Chenghua shuochang cihua congkan, which consists of the texts for sixteen cihua and a nanxi, first appeared in a photoreprint edition in 1973, published by the Shanghai Municipal Committee for the Preservation of Cultural Relics and the Shanghai Museum. Two more reprint editions came out in 1979, by Shanghai Wenwu Chubanshe and Taiwan Dingwen Shuju. 41. See Zhao Jingshen, “Tan Ming Chenghua kanben Shuochang cihua.” Zhao notes that the wood-carving styles of the pictures and characters in the four cihua texts on Hua Guan Suo are very similar to those in the Yuan pinghua texts, which suggests that these cihua might be reprints of some Yuan prototypes (20). In his Reading Illustrated Fiction in Late Imperial China, Robert E. Hegel discusses both the affinities and differences between the print style of the pinghua and the shuochang cihua. A significant difference, as Hegel points out, is with the illustrations: While the pinghua contain only illustrations occupying the top portion of a page, in the shuochang cihua there are both such page-top illustrations and the new format of half-folio pictures. Even in those pagetop illustrations in the shuochang cihua, as Hegel informs us, the movement of the characters illustrated usually proceeds from right to left, in the same direction as the reading of the text, “in contrast to the pinghua” (179). Still, the heterogeneity of the print styles among the shuochang cihua texts suggests that some of them may have had earlier textual origins than others. The only cihua that feature the format of page-top illustrations are the Hua Guansuo tales, a fact that Anne E. McLaren has also noted in her Chinese Popular Culture and Ming Chantefables (59). 42. Ye Dejun uses the term “yuequ” for the type of verse in prosimetric art that follows strict tonal schemes in order to fit the well established tunes, such as that in Zhugongdiao, and the term “shizan” for the other type of verse less controlled under musical patterns (Song Yuan Ming jiangchang wenxue, 212–217). Ye’s division between yuequ and shizan is approvingly cited by Anne E. McLaren in Chinese Popular Culture and Ming Chantefables, 79. 43. The verses in all sixteen shuochang cihua are full of what Milman Perry and A. B. Lord would call “formulas and formulaic expressions”—stock material fitting conveniently into the metric pattern to facilitate the process of verse making. It is an interesting and significant phenomenon that all sixteen cihua are uniformly in the same rhyme, which obviously has maximized the applicability and interchangeability of the formulas and formulaic expressions not only in different parts of the same cihua but at a higher level in different cihua as well. No scholarly consensus, however, has been reached on the nature of the texts in shuochang cihua. David T. Roy, while agreeing that they are based on formulaic diction, considers them “to be examples of written-formulaic, rather than oral-formulaic, composition.” See David T. Roy, “The Fifteenth-Century Shuochang cihua as Examples of Written Formulaic Composition,” 114. 44. All textual citations from Shuihu zhuan in this book are from the photoreprint of the

214

Notes to Pages 46–49

Rongyutang edition collected in Guben xiaoshuo jicheng, entitled Li Zhuowu piping Zhong yi Shuihu zhuan (5 vols.). Unless otherwise noted, all English translations are mine. 45. “Shuihu zhuan jiuben kao,” 124–125. 46. Ibid., 133ff; 141. 47. Ibid., 138–142. 48. The Evolution of a Chinese Novel, 42. 49. Ibid., 42ff. 50. By my estimate, “Shilang fuma zhuan,” one of the longer cihua among the sixteen texts, is slightly over twenty thousand characters long; and “Bao Shizhi chushen zhuan,” one of the shorter pieces, is only about half that length. 51. None of the cihua texts has a ruhua or similar narrative accessories that characterize the early vernacular short stories. The limited scope of narrative action in a cihua is suggested by some of the titles, which are modest but most pertinent to the plot: “Bao Longtu Chenzhou tiaomi ji” (Lord Bao Sells Rice in Chenzhou), “Renzong renmu zhuan” (Emperor Renzong Acknowledges His Mother), and “Bao Longtu anduan wai wupen zhuan” (Lord Bao Judges the Case of a Deformed Pot). 52. Ming Chenghua shuochang cihua congkan, 329–332. 53. Ibid., 494–496. 54. Ibid., 358. 55. Whether or not the two homophonic terms “pinghua ˚‹” and “pinghua ≠‹” have been used interchangeably is an issue open to debate. “Pinghua ≠‹” seems to be the one of older origin. In the titles of the five popular histories published during the Zhizhi reign (1321–1323) of the Yuan, “pinghua ≠‹” is the term uniformly used. In those cases the term denotes the text rather than the form of storytelling. But on other occasions it is more difficult to tell. In his Xihu youlan zhiyu, Tian Rucheng (1503–?) gives an account of the oral performance by blind men and women in Hangzhou: “They learn to play the pipa, sing xiaoshuo and pinghua ≠‹ of the old and present times in order to make a living. This is called taozhen. Usually they tell stories from the Song times, about the old customs in the capital of Bianjing” (368). It is not clear whether by “pinghua ≠‹” Tian refers to the text or a genre of oral performance. A similar example is in Shen Defu’s (1578–1642) Wanli yehuo bian, where Guo Xun, Tian Rucheng’s contemporary, is said to have asked the eunuchs in charge of pinghua to put on before the emperor daily performances based on his Yinglie zhuan (excerpted in Ming Qing xiaoshuo ziliao xuanbian, 1: 205). In Zhongguo pingshu (pinghua) yanjiu, Tan Daxian states that “pinghua ≠‹” was just another name for jiangshi during the Yuan period (20). If that was the case, the term could refer both to the text and to the oral performance itself. The other term, “pinghua ˚‹,” according to Chen Ruheng, came into use only in the early fifteenth century (Shuoshu shihua, 98), and its reference to an all-prose and full-length form of storytelling seems less controversial (possibly only because the storytelling with which the term is associated belonged to times relatively more recent). In chapter 90 of the Rongyutang edition of Shuihu zhuan, the storytelling performance is referred to as “pinghua ˚‹” (5: 2910), which might suggest a relatively late date for the completion of the fanben text. In Tan Daxian’s Zhongguo

Notes to Pages 49–51

215

pingshu ( pinghua) yanjiu, the term “pinghua ˚‹” is used to refer to both the text of the story and the form of storytelling after the Ming times, although he prefers another name, “pingshu ˚—,” for the storytelling genres in the north. 56. For an account of storytelling from the Song down to the modern times, see Chen Ruheng, Shuoshu shihua. Also see Tan Daxian, Zhongguo pingshu ( pinghua) yanjiu, 20–31. In the 1950s, there were, according to Zhang Qingping, more than forty registered professional raconteurs in Yangzhou pinghua, of whom thirty-one were in Yangzhou and the rest in Shanghai, Zhengjiang, and Nanjing. Their repertoire included Sanguo, Shuihu, Yue Fei zhuan, Sui Tang, and Xiyou ji. See Zhang Qingping, preface to Wu Song (Nanjing: Jiangsu Renmin Chubanshe, 1978), 1. 57. About the possible length of a jiangshi and any other forms of popular history telling in the Yuan and early Ming times, no record is available. We may get some idea, however, from the fact that the story of Wu Song told by the modern raconteur Wang Shaotang ran as long as seventy-five days. See Zhang Qingping, preface to Wu Song, 2. 58. The edition prefaced by Tiandu Waichen is extant only in its early Qing reprint by Shiquge. Shen Defu may have been the first to identify the pseudonymous writer of the preface as the scholar and bureaucrat Wang Daokun (1525–1593). In his Wanli yehuo bian (excerpted in Shuihu zhuan ziliao huibian, 147), Shen suggests that the Xin’an edition of Zhong yi Shuihu zhuan was a reprint of the original Guo Xun edition, and then he goes on to assert that the writer of the preface to the Xin’an edition, Tiandu Waichen, was a pseudonym of Wang Daokun’s. The identification has been called into question by modern scholars such as Yan Dunyi (Shuihu zhuan de yanbian, 176–177) and Nie Gannu (“Shuihu wulun,” 146–148). 59. Zeng Zuyin et al., eds., Zhongguo lidai xiaoshuo xu ba xuanzhu, 124. 60. In Shaoshi shanfang bicong, Hu deplores that the deletions in the Fujian editions of the narrative had made them much inferior to the editions that he had seen twenty years earlier (excerpted in Shuihu zhuan ziliao huibian, 214). 61. Shaoshi shanfang bicong, excerpted in Shuihu zhuan ziliao huibian, 131. 62. “Bangmang wenxue yu bangxian wenxue,” in Lu Xun lun wenxue yu yishu, 2: 493–494. 63. Xihu youlan zhiyu, 468. 64. The edict was issued in July 1753. Quoted in Wang Liqi, “Shuihu yu nongmin gemin,” 73–74. For a fuller account of the imperial prohibition of literary works, including Shuihu zhuan, by Emperor Qianlong, see Tai-loi Ma, “Novels Prohibited in the Literary Inquisition of Emperor Ch’ien Lung, 1722–1788,” in Critical Essays on Chinese Fiction, 201–212. Also see Okamoto Sae, Shindai kinsho no kenkyû. 65. I am borrowing the theoretical model from M. H. Abrams’ The Mirror and the Lamp: Romantic Theory and the Critical Tradition, which views the literary work in its relations to the author, the audience, and the universe. Abrams’ model is in my judgment still a useful one, although it has largely neglected the dialectic factors in literary composition. In his Chinese Theories of Literature, James J. Y. Liu proposes a revised version of Abrams’ scheme as a theoretical framework for his discussion of the Chinese literary tradition (9–13). Zhang Yuechao also considers Abrams’ model applicable to Chinese literature. He proposes that what theoretically distinguished Chinese liter-

216

Notes to Pages 52–53

ature from Western literature until recently was a difference between the mimetic-orientedness of the West and a dominantly expressive Chinese poetics as embodied in a motto such as “Shi yan zhi” (Poetry speaks of one’s intent). See Zhang Yuechao, “Zhongxi wenlun fangmian jige wenti de chubu bijiao yanjiu.” 66. “Da Deng Shiyang.” Fen shu, Xu Fen shu, 4. 67. For the critical tradition of pingdian commentary, see Part One of How to Read the Chinese Novel (David L. Rolston, ed.), which contains chapters by Rolston on the sources, historical development, and formal aspects of pingdian commentary and another chapter by Andrew H. Plaks on the terminology and central concepts in the tradition (1–123). The book also contains English translations of excerpts from some important pingdian commentaries on Ming-Qing fiction. For a somewhat more distinctly narratological approach to fiction pingdian, see David L. Rolston’s more recent book, Traditional Chinese Fiction and Fiction Commentary: Reading and Writing Between the Lines. 68. In both Shuihu ziliao huibian (Ma Tiji, ed.) and Shuihu zhuan ziliao huibian (Zhu Yixuan and Liu Yuchen, eds.), as well as Shuihu zhuan huiping ben (Chen Xizhong et al., eds.), the preface is attributed to the monk Huai Lin. But as Ye Lang suggests in his “Ye Zhou pingdian Shuihu zhuan kaozheng” (Zhongguo xiaoshuo meixue, 280–302), the true author may have been Ye Zhou (?-1625), a shadowy figure who has been both hailed for his critical insights and condemned for his alleged literary forgeries. According to the early Qing scholar Zhou Lianggong, Ye Zhou was a talented critic particularly known for his “cunning and eccentric behaviors” (Yin shuwu shuying ). For another discussion of Ye Zhou’s possible authorship of the preface and other commentaries and for numerous references to Ming-Qing sources on the issue, see Andrew Plaks, The Four Masterworks, 515–517. 69. Shuihu zhuan huiping ben, 26. The preface is also included in Li Zhuowu piping Zhong yi Shuihu zhuan. 70. “How to Read the Fifth Book of Genius,” John C. Y. Wang, trans. and David Rolston, annotator, in Rolston, ed., How to Read the Chinese Novel, 134; “Du Diwu caizi shu fa,” 17. 71. The comment occurs at the end of chapter 24 of the Rongyutang edition. See Shuihu zhuan huiping ben, 470. The purported commentator in the Rongyutang edition is Li Zhi (styled Li Zhuowu), as the full title of the edition, Li Zhuowu xiansheng piping Zhong yi Shuihu zhuan, would suggest. The authenticity of Li’s authorship of the commentary has become controversial, however, entangled with other issues surrounding such names as Huai Lin and Ye Zhou. See Andrew Plaks, “The ‘Li Zhuowu’ Commentary Editions,” The Four Masterworks, 513–517. 72. The comment occurs at the beginning of chapter 11 of the Guanhuatang edition. See Shuihu zhuan huiping ben, 232. 73. For the relationship between pingdian criticism of the narratives and the Eight-Legged Essay commentary, see David L. Rolston’s essay, “Sources of Traditional Chinese Fiction Criticism,” in Rolston, ed., How to Read the Chinese Novel, 3–34, 17–29. 74. “How to Read the Fifth Book of Genius,” in How to Read the Chinese Novel, 131; “Du Diwu caizi shu fa,” 15.

Notes to Pages 53–55

217

75. “How to Read the Fifth Book of Genius,” 132; “Du Diwu caizishu fa,” 15. 76. “How to Read the Fifth Book of Genius,” 133; “Du Diwu caizishu fa,” 16. 77. “How to Read the Fifth Book of Genius,” 135; “Du Diwu caizishu fa,” 17. 78. Jin Shengtan, “Shuihu zhuan xu san,” in Shuihu zhuan huipingben, 1: 10–11. 79. “Metacommentary,” in Robert Con Davis, ed., Contemporary Literary Criticism, 114. 80. For an overview of the recent study of traditional Chinese fiction, both in China and the West, see Robert E. Hegel, “Traditional Chinese Fiction: The State of the Field,” Journal of Asian Studies 53:2 (May 1994): 394–426. Hegel’s article touches on different subgenres of fiction but focuses on the full-length vernacular narratives, or zhanghui xiaoshuo, as they were traditionally called. For a comprehensive bibliography of Chinese vernacular fiction spanning from the Tang period to the early twentieth century, see Zhongguo tongsu xiaoshuo zongmu tiyao, Jiangsu sheng shehui kexueyuan Ming-Qing xiaoshuo yanjiu zhongxin, ed. (Beijing: Zhongguo Wenlian Chuban Zhongxin, 1990). For a bibliography of recent scholarship—in Chinese and other languages—on Ming-Qing full-length vernacular fiction, see David L. Rolston, ed., How to Read the Chinese Novel, 367–484. 81. John L. Bishop states in his famous article, “Some Limitations of Chinese Fiction,” that “the traditional colloquial fiction of China is limited in two respects: the one a limitation of narrative convention, the other a limitation of purpose” (238). 82. C. T. Hsia, while contributing significantly to making Chinese vernacular fiction a respectable discipline of literary study in his pioneering The Classic Chinese Novel, feels compelled to say that Chinese vernacular fiction appeals to the uncultivated tastes of young people before they become elevated to sophistication by reading Western novels: People may be “very fond of traditional Chinese fiction in their early youth, but once exposed to Western fiction, they could only acknowledge, implicitly if not explicitly, the greater technical proficiency of the latter along with its far greater moral seriousness” (4). 83. This view is best illustrated by a quote from C. T. Hsia: “We cannot expect colloquial Chinese fiction, with its humble oral beginnings, to have been designed for the cultivated modern taste” (The Classic Chinese Novel, 6). 84. John Bishop, “Some Limitations,” 239. 85. C. T. Hsia, The Classic Chinese Novel, 101. Hsia’s unfavorable attitude toward the formal influence from storytelling is also explicit in his ranking of the early Chinese novels, where his preference for a certain work is in reverse proportion to his assessment of its connection to the oral tradition. While admitting that Shuihu zhuan in some respects marks an advance in the artistic evolution of Chinese narrative literature, he hastens to add that such gains “are offset by concomitant liabilities stemming from the work’s greater indebtedness to professional storytellers,” and therefore “it must be read as a lesser work of art” than Sanguo yanyi (75–76). 86. John Bishop, “Some Limitations,” 240. 87. Ibid., 243. 88. Xia Zhiqing (C. T. Hsia), “Sui shi yi wen chongkan xu,” in Zhongguo gudian xiaoshuo lunji, 2: 124–125. In his Classic Chinese Novel, C. T. Hsia states that there is in Chinese vernacular fic-

218

Notes to Pages 56–61

tion an almost morbid obsession with the darker side of human nature, which is again associated to oral storytelling: “In spite of the severe moralism it ostensibly upholds, Chinese fiction is notable for its lack of Victorian prudery. But this tolerance, originally stemming from the low culture of storyteller and audience alike, indicates not so much modern broadmindedness as an unsqueamish delight in sex, filth and disease” (21). 89. W. L. Idema, Chinese Vernacular Fiction, 56. 90. “A Taste for Apricots,” 55. 91. The Iliad, 445–446. 92. H. M. Chadwick and N. K Chadwick, The Growth of Literature, 2: 488. 93. The Spectator, 4: 287. 94. Walter Ong, Presence of the Word, 131 and 133. 95. Ibid., 133. 96. Walter Ong, Orality and Literacy, 70. 97. “Full-Length Xiaoshuo,” 165, 167. 98. The Four Masterworks, 50ff. 99. Chinese Vernacular Fiction, x-xxii. 100. The theory that identifies a huaben with a diben for a storyteller in the Song period started with Lu Xun’s A Brief History of Chinese Fiction. Masuda Mataru has challenged the theory by arguing that the word “huaben” should be better defined simply as “story.” See Masuda Mataru, “‘Wahon’ to iu koto ni tsuite.” Masuda, however, does not say that a huaben text could not be a diben for a storyteller, for his paper focuses on the term “huaben” itself rather than the nature of the genre. While his article was a significant step toward clarifying the once cloudy issue, it is the more recent scholarship, particularly by Patrick Hanan and W. L. Idema, that has decisively rectified the prevalent misconceptions of huaben and pinghua. See Hanan’s Chinese Short Story and Idema’s Chinese Vernacular Fiction: The Formative Period. 101. In an article rejecting the theory that huaben was once the storyteller’s promptbook, Zhou Zhaoxin suggests that twenty-six of the early huaben stories contain too detailed a narrative discourse to have ever served as promptbooks in storytelling. Rather, they are more likely to have been men of letters’ edited notations of storytelling performances. See Zhou Zhaoxin, “ ‘Huaben’ shiyi,” 202. 102. The ironic or allegoric interpretation of vernacular fiction has a history that can be traced back to seventeenth- and eighteenth-century commentators. Jin Shengtan, in particular, was infatuated by what he calls “chunqiu bifa” (writing techniques of The Spring and Autumn Annals) in Shuihu zhuan, by which he means the rhetorical manipulation on the part of the writer to encode some hidden and deeper meaning beneath the surface layer of the text. For instance, he considers the description of Song Jiang’s respect for Chao Gai and his reluctance to let Chao be exposed to dangerous fighting as betraying Song’s undeclared strategy of reducing Chao to a mere figurehead, and Song’s later unusual acquiescence in Chao’s expedition to the Zengtoushi as evidence of his latent wish to see Chao dead. Jin refers to the textual irony that he has uncovered as “shenwen qubi” (roundabout writing with deep-seated meanings), stressing the incongruity or discrep-

Notes to Pages 61–66

219

ancy between the surface diction and a meaning that might lie hidden (Shuihu zhuan huiping ben, 2: 1084). Since Jin Shengtan, the tradition of interpreting Shuihu zhuan in terms of irony or allegory has continued unabated. Sun Shuyu, for example, has offered a reading of the narrative in which the exchange of letters and subsequent battles between the Liangshanbo and the Zengtoushi are interpreted as an allegory of the war and diplomatic negotiations between the Song and its belligerent northern neighbor, the state of Jin, and the death of Chao Gai after the Zengtoushi campaign is said to represent the humiliating abduction of the two Song emperors, Huizong and Qinzong, by the Jin invaders (Shuihu zhuan de laili, xintai, he yishu, 156–160). In the English language, Andrew Plaks’ Four Masterworks of the Ming Novel is the most systematic and powerful study of the rhetorical stance of irony in the vernacular narratives, including Shuihu zhuan. In the fanben text of Shuihu zhuan, Plaks sees the heroic codes to which the major characters (notably Wu Song, Li Kui, and Song Jiang) apparently adhere constantly undercut by the narrative itself. As a result, the plot and characterization of the novel are rendered more ambiguous and problematic (319–358). 103. While acknowledging that for Shuihu zhuan there had been “a variety of historical sources and a wide range of treatments of similar narrative material in such forms as oral storytelling, popular song narratives (cihua) and drama,” ( Four Masterworks, 279), Plaks suggests that the narrative had distanced itself from its source material and put itself in a totally different category: “This is not to deny the parallel existence of the traditional popular images of the ‘watermargin’ heroes. It is only to insist that the fanben novel is simply no longer working along these lines” (358). 104. The Odyssey, Book 13, 232–233. 105. See Porphyry, The Cave of the Nymphs, Thomas Taylor, trans. (Grand Rapids, MI: Phanes Press, 1991). 106. 2 Corinthians, 3, 6. 107. Origen, An Exhortation to Martyrdom Prayer, Rowan A. Greer, trans. (New York: Paulist Press, 1979), 218. 108. On Christian Doctrine, 8–13. 109. Ibid., 84. 110. Shuihu zhuan huiping ben, 2: 1084.

Chapter 3: The Narrative Pattern: The Uniform versus the Multiform 1. The Classic Chinese Novel, 101. 2. W. L. Idema, Chinese Vernacular Fiction, 71. Original italics. 3. “The Nature of Ling Meng-chu’s Fiction,” 87. 4. Y. W. Ma and Joseph S. M. Lau have challenged the usefulness of the generic label “nihuaben,” which, according to them, can lead to a misconception of all vernacular stories using Ming setting as imitations of the Song stories. See Ma and Lau, eds., Traditional Chinese Stories: Themes and Variations, xxii-xxiii. 5. Albert B. Lord has proposed the concept of the “degree of orality” in his “Perspective on Recent Work on the Oral Traditional Formula.” Discussing the Anglo-Saxon tradition, he seems

220

Notes to Pages 66–76

to be somehow modifying his previous belief in the polarization of the oral and the literate cultures: “We have indicated that what we were measuring in our analysis of formula density was not orality in the absolute, not whether any given text was a fully oral-traditional poem or not, although this was sometimes the case, but the degree of ‘orality’ of that text, judging by its use of traditional formulas, which all seem to agree are characteristic of oral-traditional composition” (396). 6. Zhao proposes that the early stage of vernacular fiction was characterized by “a continuous rewriting.” But he does not consider the rewriting in terms of the gradual vernacularizing process. Instead, he regards it as the result of anonymous authorship that was typical of early vernacular fiction: “Because of the lack of authorship, editorship asserted itself through repetition.” See Henry Y. H. Zhao, The Uneasy Narrator, 10. 7. The Making of Homeric Verse, 405–406. 8. The Singer of Tales, 68. 9. Ibid., 4. 10. Some scholars have challenged Milman Parry’s claim that the function of formulas is to serve the need of the metrical pattern. They argue that in oral prose, where there is no such need, formulaic diction is the vehicle for verbal repetitions. Bennison Gray, for example, goes so far in his article “Repetition in Oral Literature” as to say: “As far as formulas as a whole are concerned . . . they occur just as frequently in oral narrative prose as they do in oral narrative verse” (293). 11. “The Composition of P’ing yao chuan,” 217. 12. Peter Li, “Narrative Patterns in Sanguo and Shuihu,” 80–81. Li’s description of what he calls a “general pattern” for the individual cycles is as follows: “The hero first demonstrates his gallantry with a feat or bravery that, unfortunately, is a violation of the law, and he takes flight. But his flight is short-lived; he is captured and branded. Then, he is sentenced and exiled to a distant region to serve his term. Usually, being sent into exile is the beginning of the hero’s adventures. He is escorted by two incompetent and unscrupulous guards who invite trouble. But the hero does arrive at his destination and begins to serve his sentence. Soon, however, unbearable conditions force him to escape again, and this time he becomes a permanent fugitive from the law” (81). Li’s “general pattern” for the most part coincides with the “exile and imprisonment sequence” that I will discuss in this chapter, but it is not the overarching pattern for all the heroes’ individual adventures in the narrative. 13. The Singer of Tales, 68. 14. Typical Battle Scenes in the Iliad, 5. 15. The Conference Sequence: Patterned Narration and Narrative Inconsistency in the Odyssey, 2–3. 16. In his “Composition of P’ing yao chuan,” Patrick Hanan considers the Lin Chong story and the Lu Junyi story a repetition of the same stuff material: “The incident of the two escorts appears twice in the Shuihu, and the author’s lame excuse is that the two have been given a second chance to show their competence as murderers” (215). 17. Samuel Hung-Nin Cheung has noticed a parallel in the three episodes of killing an adulterous woman, involving respectively Song Jiang, Wu Song, and Shi Xiu and Yang Xiong. Cheung calls this narrative parallel “structural cyclicity.” See his article, “Structural Cyclicity in Shuihu

Notes to Pages 78–91

221

zhuan: From Self to Sworn Brotherhood.” Yet despite the similarities among the three women and their bloody deaths, as well as the strong resemblance between the two young street vendors— Tang Niuer in the Song Jiang story and Yunge in the Wu Song story—Song Jiang’s killing of Yan Poxi is considerably different from the other two murders. For one thing, it does not involve the moral values of brotherhood or sworn brotherhood, a point that Cheung notes in his article. 18. The Singer of Tales, 68. 19. Ibid., 191. 20. Ibid., 192. 21. Vladimir Propp, Morpholog y of the Folktale, 19. 22. “From Etic to Emic Units in the Structural Study of Folktales,” 101. 23. Ibid. 24. “Structural Typology in North American Indian Folklore,” 208. 25. Si Su, “Shuoshu youwu jiaoben?” 44ff. 26. Vena Hrdlickova, “The Professional Training of Chinese Storytellers and the Storyteller’s Guilds,” 225ff. 27. Mark Bender, “Keys to Performance in Kunming Storytelling,” 27. 28. Vibeke Børdahl, “A ‘Poetics’ of Chinese Storytelling,” 249. 29. Endel Tulving, Elements of Episodic Memory, 9. 30. Ibid., 11. 31. Ibid., 171. 32. Ibid., 44. 33. Michael J. Howe, Introduction to Human Memory: A Psychological Approach, 57. 34. Ibid., 52. 35. Similar triplets occur frequently in other early vernacular narratives as well. In Sanguo yanyi and Xiyou ji, for instance, there are such narrative stretches as san rang Xuzhou (Tao Gongzu yields Xuzhou three times), san gu maolu (Liu Bei visits Zhuge Liang’s cottage three times), Liu Biao san qiuji (Liu Biao thrice asks Zhuge Liang for a strategy), san qi Zhou Yu (Zhuge Liang infuriates Zhou Yu three times), Shimo san xi Tang Sanzang (Skeleton Demon tricks Tripitaka three times), and san diao bajiaoshan (Sun Wukong tries three times to borrow the palm-leaf fan). This is a phenomenon that Andrew Plaks has also noted in his Four Masterworks of the Ming Novel, 383–384. 36. “Epic Laws of Folk Narrative,” 133. 37. The Four Masterworks of the Ming Novel, 309. Plaks refers here to the narrative pattern in the opening chapters of Shuihu zhuan, “where we follow the course of one figure until he runs into another, whereupon the narrative then follows the new trajectory of the latter’s adventures, leaving the former in his tracks, perhaps to be reintroduced at a later juncture.” 38. Gérard Genette, Narrative Discourse, 27. 39. Quoted in Andrew Plaks, “Terminology and Central Concepts,” in David L. Rolston, ed., How to Read the Chinese Novel, 87. 40. “How to Read the Fifth Book of Genius,” John C. Y. Wang, trans. and David L. Rolston, annotator, in David L. Rolston, ed., How to Read the Chinese Novel, 135; “Du Diwu caizi shu fa,” 17.

222

Notes to Pages 91–93

41. The comment occurs at the beginning of chapter 11 of the Jin Shengtan edition. See Shuihu zhuan huiping ben, 232–233. 42. “How to Read the Fifth Book of Genius,” 143; “Du Diwu caizishu fa,” 21. 43. “How to Read the Fifth Book of Genius,” 143–144; “Du Diwu caizishu fa,” 21. 44. The Four Masterworks, 102. 45. Andrew Plaks judiciously treats the recurrent patterns in Jin ping mei differently from those in the other three “masterworks,” each of which was preceded by an oral tradition. He attributes the “repetitiveness” in Shuihu zhuan, Sanguo yanyi, and Xiyou ji “partially to their adaptation of conventional narrative topoi from their source materials,” a statement that amounts to an acknowledgment of the impact of the oral tradition on the story making. As for Jin ping mei, he considers the recurrence “as an integral part of the author’s literary design,” on the grounds that the work “departs from the other novels in its relative independence from antecedent narratives.” See his Four Masterworks, 102–103. Although the theory that Jin ping mei was written by an individual writer is widely accepted, there are dissenting voices. Xu Shuofang, for example, argues that Jin ping mei, like the other three masterworks in the Ming vernacular fiction, should be considered as a work rooted in the oral and popular tradition. See Xu Shuofang, “Jin ping mei de xieding zhe shi Li Kaixian,” in his Lun Tang Xianzu ji qita, 133–147. 46. One area of early vernacular literature in which thematic parallels to Shuihu zhuan are particularly copious is the short stories. I return to this issue in chapter 4, where the kinship between Shuihu zhuan and the stories are used to justify the stylistic and linguistic analyses of Shuihu zhuan for which the stories serve as a context. 47. See Hanan, “The Composition of P’ing yao chuan,” 213. Luo Ergang has also noted the parallel between passages and narrative sequences in Shuihu zhuan and San Sui ping yao zhuan, including what I have called the “exile-imprisonment sequence.” Instead of considering such recurrences as determined by a shared storehouse of storytelling material, however, Luo uses them as evidence for a shared authorship. See Luo Ergang, “Cong Luo Guanzhong San Sui ping yao zhuan kan Shuihu zhuan zhuzhe he yuanben.” Ouyang Jian has also commented on the extensive parallels between Shuihu zhuan and San Sui ping yao zhuan. He argues that such parallels could be partly because Luo Guanzhong may have been involved in the compilation of both works, but a more fundamental reason is that both narratives may have gone through a long process of evolution in popular orality and shared similar story-making material. See Ouyang Jian, “Ping yao zhuan: Shuihu zhuan de zimei pian.” 48. Ming Chenghua shuochang cihua congkan, 314ff. 49. Ibid., 361ff. 50. Qingping shantang huaben (Shanghai: Shanghai Guji Chubanshe, 1992), 8. David L. Rolston has also noticed the ubiquity of these two bailiffs in early vernacular literature, including Shuihu zhuan and the story Jiantie heshang. Rolston compares these two names to two other equally stereotyped ones in popular drama, the names of Meixiang and Zhang Qian, which are almost invariably used to refer to maids and servants, respectively. See Rolston, Traditional Chinese Fiction and Fiction Commentary, 200–201, note 23.

Notes to Pages 94–97

223

51. In “Bao Longtu duan baigujing zhuan,” however, it is Yang Guang, instead of Di Qing, who is praised as the best military officer during Renzong’s reign. 52. Again, whatever the origins of the yinshou, it does not block the way for different interpretations of it as part of a literary text. Deborah Porter, for instance, argues cogently that the imprecise references to historical facts and numbers of years in the yinshou may be examples of a deliberate and subversive playfulness challenging a conventional discourse to delimit history. See her article, “Setting the Tone: Aesthetic Implications of Linguistic Patterns in the Opening Section of Shui-hui chuan.” 53. Ming shuochang cihua congkan, 4ff. 54. Xuanhe yishi, 51–53. This parallel between Hua Guan Suo and the Shuihu cycles has been noted by Anne E. McLaren in her Chinese Popular Culture and Ming Chantefables, 242–243. McLaren observes, “The Taihang Mountains rapidly became a storyteller’s conventional term for a bandit lair,” occurring also in Sanguo zhi pinghua. 55. Song Yuan pinghua ji, 1: 39. 56. Ibid., 1: 191. 57. Chu Renhuo, Sui Tang yanyi, 40–44. The extant text is based on the edition dated 1675 (the fourteenth year of the Kangxi period), which was revised and edited by Chu Renhuo. In his preface to Sui Tang yanyi, Chu Renhuo attributes the narrative’s alleged antecedent, Sui Tang zhizhuan, to Luo Guanzhong. Even if we do not accept the attribution at its face value, it is still quite possible that the basic structure of the narrative is based on a prototype dated much earlier. 58. The similarity between the Lu Zhishen-Zhou Tong episode in Shuihu zhuan and the Sun Wukong-Zhu Bajie episode in Xiyou ji is also pointed out by Andrew Plaks in The Four Masterworks, 200. Plaks also suggests a more subtle parallel between the episode of the “centipede spirits” in Xiyou ji and Wu Song’s killing of a lascivious priest at the Centipede Ridge. 59. For a late Ming edition of Hei Xuanfeng zhang yi zhucai, see Shuihu xiqu ji, 1: 95–114. 60. Ming Chenghua shuochang cihua congkan, 338ff. 61. Texts for both cihua appear in Ming Chenghua shuochang cihua congkan, 463–550 and 659–720. The latter piece is also entitled “Shangyuan shiwu ye kandeng zhuan.” The plots of these two cihua are strikingly similar. The villain in each is from the royal family, as indicated in the title. He forcibly seizes a commoner’s wife, kills the husband (in the latter cihua, he slaughters the husband’s entire family except his brother, his small son, and an old servant), and keeps the woman as his concubine. In both works, the villain has an accomplice: In the former cihua his older brother and in the latter his friend. Each cihua ends with Judge Bao’s execution of the powerful culprit and his accomplice. Indeed, the two cihua display such a close correspondence with each other that they might have evolved from an identical source. 62. The text of Zhang Wengui zhuan appears in Ming Chenghua shuochang cihua congkan, 551–626. For the episode in question, see 565–580. 63. Xue Rengui zheng Liao shilue was collected in the Yongle dadian and is reprinted in Ming Chenghua shuochang cihua congkan, 209–288. For the fighting scene in question, see 276–277. 64. Ibid., 218–219.

224

Notes to Pages 98–103

65. Only the latter part (houji) of Qiguo Chunqiu pinghua is extant, which is included in Song Yuan pinghua ji, 2: 483–560. 66. Song Yuan pinghua ji, 2: 533 ff. 67. The goddess is also a key figure in San Sui ping yao zhuan, where she offers both her blessing and counseling to the rebels, as she does in Shuihu zhuan. 68. Song Yuan pinghua ji, 2: 552 ff. 69. The Four Masterworks, 499. 70. “The Composition of P’ing-yao chuan,” 215. 71. The Chinese Short Story, 197. 72. Vladimir Propp, Morpholog y of the Folktale, 70. 73. Literate Revolution in Greece and Its Cultural Consequences, 177–178.

Chapter 4: From Voice to Text: The Orality Writing Dynamic 1. Zhu Qing and Li Yonggu have noted that the term “diben” was commonly used in the late Yuan period to advertise textual authenticity but became largely out of vogue in the Ming. If that was the case, the term might suggest that there had been a Shi Nai’an text—possibly a manuscript— in the late Yuan as an early link in the chain of the textual evolution of the narrative. See Zhu Qing and Li Yonggu, “Shuihu zhuan zuben ji ‘Guo Wu dingben’ wenti xinyi.” 2. Shuihu zhuan ziliao huibian, Zhu Yixuan and Liu Yuchen, eds., 130–132. 3. In Jia Zhongming’s Lu gui bu xubian, titles of three plays are listed under Luo Guanzhong’s name: Feng yun hui, Lianhuan jian, and Fei Huzi (281). In recent scholarship, the authenticity of Luo’s compilership for some of the narratives traditionally attributed to him has been called into question. Liu Ts’un-yan suggests that Luo is likely to have been the compiler of Sanguo zhizhuan, part of an early version of Da Tang Qinwang cihua, and the Tian Hu, Wang Qing sections of the jianben Shuihu zhuan. He proposes that neither the fanben Shuihu zhuan nor most of the twentychapter version of Ping yao zhuan should be attributed to Luo. See Liu Ts’un-yan, “Luo Guanzhong jiangshi xiaoshuo zhi zhenwei xingzhi,” 112–158. 4. Lu gui bu xubian, 281. 5. Xihu youlan zhiyu, 468. Some modern scholars such as Xu Shuofang have tried to explain away this contradiction in different accounts of Luo Guanzhong’s nativity with a hypothesis that Taiyuan might be Luo’s ancestral town while Hangzhou was his own place of residence. See Xu Shuofang, “Cong Song Jiang qiyi dao Shuihu zhuan chengshu,” 193ff. 6. Shuihu zhuan ziliao huibian, 130–132. 7. Ibid., 131. 8. Ming Qing xiaoshuo ziliao xuanbian, 69. Dongyuan, according to Cihai, was an ancient name for the area in Shandong that became Dongping Prefecture during the Han. 9. See, for example, Li Weishi, “Shuihu zhuan chengshu yu Yuan-mo Ming-chu zhi shuo buneng chengli.” Yan Dunyi made a similar suggestion in Shuihu zhuan de yanbian (238). 10. Shuihu zhuan ziliao huibian, 130 and 131. 11. Shuihu zhuan de yanbian, 211.

Notes to Pages 103–104

225

12. A Brief History of Chinese Fiction, 191. 13. These articles are: Bao Yu, “Shi Nai’an zai Baiju Chang,” Shen bao, October 29, 1946; Tao Liang, “Baiju Chang zhuiyu,” Shen bao, November 4, 1946; Wang Zhexing, “Shi Nai’an yu Zhang Shicheng,” Shen bao, November 4, 1946; and Yu Heng, “Shi Nai’an shiji zhi xin shangque,” Shen bao, May 31, 1947. All these articles are reprinted in Shi Nai’an yanjiu. 14. Liu Dong and Huang Qingjiang, “Shi Nai’an yu Shuihu zhuan,” Wenyi bao 21, 1952. 15. Nie Gannu, “Shuihu wulun,” 15. 16. Based on the information from the epitaph and from some other sources, one scholar comes to the conclusion that the dates for Shi Nai’an’s life should be 1332–1406. See Chen Liao, “Quwei cunzhen, Shi Nai’an zhi mi ke jie.” In a similar vein, Huang Ticheng in his/her “Shi Nai’an muzhi kaosuo” and “Shi Nai’an jiguan kaobian” supports the theory that Shi was a native of Xinghua who lived at the end of the Yuan and the beginning of the Ming. Zhang Peiheng also agrees that Shi Nai’an was another name for Shi Yanduan, although he considers some of the evidence used by others to prove that the case is spurious. See his “Shi Yanduan shifou Shi Nan’an.” In another of his articles, “ ‘Shi Nai’an muzhi’ bianwei ji qita,” Zhang Peiheng dismisses the Shi Nai’an tombstone inscription attributed to Wang Daosheng of the Ming as a forgery by a Hu Ruiting in 1928. 17. For the controversy among contemporary Chinese scholars on the life of Shi Nai’an and on the authenticity of the Xinghua-Dafeng evidences in relation to Shi Nai’an’s nativity, see Shi Nai’an yanjiu. 18. Wu Mei, Gu qu zhu tan, in his Wu Mei xiqu lunwen ji, 98. 19. Both Wang Liqi and Huang Lin have made the suggestion. See Wang Liqi, “Shuihu quanzhuan shi zenyang zuanxiu de,” and Huang Lin, “Song-mo Yuan-chu ren Shi Nai’an ji ‘Shi Nai’an di ben.” 20. Oral Tradition in Literature, 8. 21. The fragment was first identified to be a Zhengde-Jiajing print by Gu Tinglong and Shen Jin in their article “Guanyu xin faxian de Jingben Zhong yi zhuan canye.” Most Shuihu scholars have accepted that theory. Zhang Guoguang, however, believes the fragment to be a jianben text, and therefore necessarily later than the Wuding edition. See his “Yijiubaer nian de Shi Nai’an re cong he er lai.” Both Zhang’s premise and conclusion seem debatable. The Jingben fragment is somewhat shorter than the corresponding section in the Jiajing canben or Rongyutang edition, but that does not necessarily put it in the category of jianben, for all “full” editions ( fanben) do not have to be equally “full.” Still, even if we agree with Zhang Guoguang in accepting the Jingben as a jianben text, it does not necessarily establish its later dating than the Wuding edition. That all extant jianben versions are dated later than the Jiajing period does not allow one to assume any newly uncovered jianben edition to be post-Jiajing a priori. Liu Shide also thinks that Jingben Zhong yi zhuan should be categorized as a jianben, but he agrees that it is probably earlier than any other extant editions. See his article, “Lun Jingben Zhong yi zhuan de shidai, xingzhi he diwei.” 22. For the Wuding edition, I personally have not come across any evidence that would help pinpoint the date of publication. One scholar, however, has suggested that the Wuding edition

226

Notes to Pages 104–106

was published by Guo Xun in 1539 or 1540 (the eighteenth or nineteenth year of the Jiajing reign) without citing the source. See Yi Yongwen, “Zai lun Shuihu zhuan shi fanying shimin jieceng liyi de zuopin.” 23. Zheng Zhenduo, once in possession of some of the fragments, suggests in his “Jie zhong de shu xuji” that they were very likely parts of a copy of the Jiajing edition Gao Ru referred to in Bai chuan shu zhi, but he does not explicitly identify it with the Guo Xun edition (Shuihu ziliao huibian, 184–185). Later, however, he asserts in his preface to Shuihu quanzhuan that the fragments were indeed from a copy of the Wuding version. This unexplained flip-flop is challenged by Y. W. Ma in his “Huyu yanjiu jianben Shuihu yijianshu.” 24. Shuihu ziliao huibian, 1. 25. Four Masterworks, 287–289. 26. Shuihu ziliao huibian, 1. For discussions of Guo Xun’s “reprint” of an earlier prototype, see Zheng Zhenduo, “Shuihu zhuan de yanhua,” 117–222; Nie Gannu, “Shuihu wu lun,” 147–148; Huang Lin, “Yizhong zhide zhumu de Shuihu guben;” and Shiroki Naoya, “Kaku Butei hon shikò.” 27. Shuihu zhuan ziliao huibian, 148. 28. According to Andrew Plaks, Jingben Zhong yi zhuan, which is considered by most scholars as another version in the fanben tradition, may have belonged to “a separate line of textual development.” See Four Masterworks, 286. 29. This is a point that Andrew Plaks has also made in Four Masterworks. He states that Jin Shengtan’s edition “does not depart from the basic fanben text,” but he concedes in the note that “this is obviously a matter of subjective judgment,” acknowledging dissenting views of other scholars (293). 30. See Ouyang Jian, “Jingben Zhong yi zhuan canye pingjia shangdui.” Ouyang Jian attributes the feat of counting all the appearances of the suffix of plural personal pronouns to Ding Zhenghua. According to Lü Shuxiang, “men Ã,” also in the variant forms of “men ˘” and “men V,” actually began to be used as a suffix for plural nouns/pronouns in the Song period, earlier than “mei.” However, “men” was largely replaced by “mei” during the Yuan and early Ming, before “men” regained its dominance around the middle of the Ming. See Lü Shuxiang, “Shuo men,” 145–148. Jin ping mei, as Lü points out, is the conspicuous exception to the rule, where “mei” is still used despite the work’s late Ming dating. “Mei” is the word found in writings with indisputable Yuan dating, such as Yuan dianzhang and Yuankan zaju sanshi zhong. It remains to be used in Zang Maoxun’s Yuanqu xuan and the late Ming manuscript of zaju plays such as those from the Maiwangguan Library. In Zhu Youdun’s plays collected in Shemotashi qucong, “mei” is used on all occasions except for one instance of “men” in Qujiangchi. Among the three xiwen texts from the Yongle dadian, the ratio of the occurrences of “mei” to those of “men” is 3 to 0 in both Xiao Sun tu and Huanmen zidi cuo lishen, but 11 to 10 in Zhang Xie zhuang yuan. In the 1498 edition of Xixiang ji, again only “mei” is used. In the 1480 revised editions of Nog oltae ònhae and Pak t’ongsa ònhae (both in Nog oltae Pak tongsa ònhae), only “men” is used. The use of the two suffixes is complicated in early vernacular literature, and no clear-cut lines can be drawn. In general terms we can probably say that there were two shifts from the Song to the Ming, the first from men” to “mei” and the second from “mei” back to “men.” Obviously, the second shift is more relevant to what happened in the evolution of Shuihu zhuan.

Notes to Pages 106–109

227

31. Shen Guoyuan, Huang ming cong xin lu, excerpted in Shuihu zhuan ziliao huibian, 552. 32. Qi xiu lei gao, exerpted in Shuihu zhuan ziliao huibian, 130. 33. A Brief history of Chinese Fiction, 186. 34. There are, of course, exceptions. Zhang Guoguang has argued that Shi Nai’an could be the pseudonym of one of Guo Xun’s hangers-on. See Zhang Guoguang, “Shuihu zuben tankao.” 35. According to Y. W. Ma, the earliest among all the extant jianben texts, complete or fragmentary, is the so-called Zengcha ben—namely, the edition with the parts on Tian Hu and Wang Qing inserted. Of the fragments belonging respectively to two different versions of the edition, one can be dated between 1573 and 1594 and the other between 1573 and 1603. Both could possibly be earlier than the Pinglin ben, also known as the Shuangfengtang edition, which was dated 1594 and sponsored by the Fujian publisher Yu Xiangdou (1588–1609). See Y. W. Ma, “Xiancun zuizao de jianben Shuihu zhuan.” 36. On the complicated issue of the priority between the two recensions, to divide the scholars’ opinions into two clear-cut camps would necessarily be an oversimplification. But since the focus of my study lies elsewhere, I will not go further than this brief summary. For more detailed accounts of the controversy, see He Xin (32–73), Nie Gannu (140–204), Yan Dunyi (Shuihu zhuan de yanbian, 140–205), Irwin’s Evolution (61–75), Liu Ts’un-yan (112–158), and Y. W. Ma, “Huyu yanjiu jianben Shuihu yijianshu.” 37. Four Masterworks, 301. 38. This is different from but has something in common with the lineage of textual evolution suggested by Ouyang Jian. There were, as Ouyang Jian argues, three stages of development: (1) from the jianben with Tian and Wang but without Liao to the fanben with Tian and Wang trimmed off but Liao added; (2) from the fanben with Liao but without Tian and Wang to the abridged jianben also with Liao but without Tian and Wang; and finally, (3) roughly concurrent appearances of the fanben with transformed parts of Tian and Wang and the jianben with an old version of Tian-Wang inserted. See Ouyang Jian, “Shuihu jianben fanben dishan guocheng xinzheng.” 39. The Classic Chinese Novel, 78. 40. This hypothesis may be enforced by the fact that some of the jianben editions, especially those that were printed relatively earlier such as the so-called Pinglin ben and the Zengcha ben, contain pictures in the upper portion of each page. These pictures might have been meant for showing to the audience, which may testify to the function of the ancestral version of the jianben recension as a performance aid for professional raconteurs. This certainly agrees with the fact that there existed in China a long tradition of “picture-storytelling,” starting with the Tang performance of bian and continuing probably with pinghua and baojuan recitation in later times. For “picture-storytelling” in China, and in other parts of Asia, see Victor Mair, Painting and Performance: Chinese Picture Recitation and Its Indian Genesis. 41. In the late Ming Maiwangguan manuscript copies of the plays, however, the Liangshan bandits are referred to as “thirty-six major chieftains and seventy-two minor ones.” 42. See George A. Hayden, “A Skeptical Note on the Early History of Shui-hu chuan” and Y. W. Ma, “Cong Zhao’an bufen kan Shuihu zhuan de chengshu guocheng.”

228

Notes to Pages 109–119

43. It was Huang Lin who uncovered the relevance of Wu Congxian’s Xiaochuang ziji to the study of the textual evolution of Shuihu zhuan. In his “Yizhong zhide zhumu de Shuihu guben,” Huang Lin considers the edition that Wu Congxian referred to as a very early version, possibly from the early Yuan period. 44. Shuihu zhuan ziliao huibian, 218. 45. Wu Tongrui et al., Zhongguo suwenxue gailun, 145–146. 46. “Piping Shuihu zhuan shuyu,” Li Zhuowu piping Zhong yi Shuihu zhuan, 1: 3–4. 47. Shuihu zhuan huiping ben, 32. 48. T’ang Transformation Texts, 119. 49. Stephen West, “Translation as Research: Is There an Audience in the House?” 141. 50. Ming Chenghua shuochang cihua congkan, 96–113. The cihua tells the story of Shi Jingtang (892–942), the first emperor of the Latter Jin (936–946). 51. “Text as Interpretation: Mark and After,” 166. 52. Yuan-Ming zaju, 27. 53. In his preface to Yuanqu xuan, Zang Maoxun claimed that he had borrowed from his friend two hundred zaju texts that were said to be originally from the Office of the Imperial Theater ( Yuxijian). 54. The play is about Liu Xibi’s (whose other name is Wenlong) separation from and eventual reunion with his wife. In the Yongle dadian, a xiwen play with the title of Liu Wenlong was listed, but the play is nonextant. The manuscript unearthed in 1975 is thus the only extant version of the play, which is now available in a modern typeset edition with the title Xuande xieben Jinchai ji (Guangzhou: Guangdong Renmin Chubanshe, 1985). 55. Xuande xieben Jinchai ji, 143. 56. See Chen Liming, “Li Xibi Jinchai ji de faxian jiqi gaimao,” Xuande xieben Jinchai ji, 141–145. 57. Ibid., 145. 58. See The Chinese Short Story, 198–199. 59. Xingshi heng yan (Beijing: Renmin Wenxue, 1987), 1: 264. 60. Xingshi heng yan, 1: 264. 61. Yushi ming yan (Shanghai: Shanghai Guji, 1992), 373. 62. Jingshi tong yan, 1: 457. 63. Xingshi heng yan, 2: 722. 64. Yushi ming yan, 374. 65. For the titles of the early stories whose source material may fit into different categories of xiaoshuo as classified in Zuiweng tanlu, see Patrick Hanan, The Chinese Short Story, 171ff. 66. The Chinese Short Story, 204. 67. Luo Ye, Xinbian Zuiweng tanlu, 4b. 68. The Chinese Short Story, 21. 69. For further details about the method and result of Hanan’s dating of the stories, see his Chinese Short Story. 70. This belief is still common, especially among Chinese scholars. In one of the most re-

Notes to Pages 120–125

229

cently published histories of Chinese literature, Shuihu zhuan, along with Sanguo yanyi, is categorized as Yuandai xiaoshuo (fiction of the Yuan period). See Zhongguo wenxue shi, Zhang Peiheng and Luo Yuming, eds. 3 vols. (Shanghai: Fudan Daxue Chubanshe, 1996), 3: 173. 71. Kòsaka Jun’ichi calls jiubian “a synonymous-compound adverb,” as it has basically the same meaning as either “jiu” or “bian.” See Kòsaka, Shuihu cihui yanjiu (xuci bufen), 198. Kòsaka is, however, erroneous in asserting that the compound appears in Shuihu zhuan only in the form “jiubian” and never in the reversed order of “bianjiu.” In fact, “bianjiu” does appear in the Rongyutang edition. Here are two examples: “bianjiu candeng xia shaole” (and burned it under the expiring lamp) (2: 658); “Duode, Ganniang bianjiu shoule” (If there is any change, please just keep it) (2: 786). This is another piece of evidence for the Shuihu zhuan’s linguistic heterogeneity. In my scanning of the text, however, I have not counted “bianjiu” as a variant of “jiubian.” 72. The “Prologue” ( Yinshou) is not included in the table, because it is inordinately shorter than the average length of the chapters. It should be noted, however, that the “Prologue” contains one occurrence of the stylistic feature “daoshi” from the early period, while no features from the middle period are found. 73. In his Shuihu cihui yanjiu (xuci bufen), Kòsaka Jun’ichi observes that by the time of Zhou Deqing’s Zhong yuan yinyun (Sounds and Rhymes of the Central Plain), the pronunciations of “qia” and “que” “might have become extremely close to each other and therefore interchangeable” in early vernacular fiction. Kòsaka notes that, in addition to “quedai” and “qiadai,” another similar pair, “quehao” and “qiahao,” also appear alternately in Shuihu zhuan and at a high frequency (136). 74. Kòsaka Jun’ichi has noted in his Baihua yuhui yanjiu that jiao – and jiao Ê also appear in Shuihu zhuan as two variant writing forms for the same preposition. In Jin Shengtan’s Guanhuatang edition, some occurrences of jiao – are replaced by another homophonic, jiao s (7–8). In Shuihu cihui yanjiu (xuci bufen), Kòsaka cites several pre-Ming examples to suggest that jiao Ê came into use earlier than jiao – (333), which corroborates Hanan’s classification of zhijiao Ω– and youfenjiao ≥¿– as middle-period stylistic features. In the twenty-chapter version of San Sui ping yao zhuan, youfenjiao appears thirteen times in chapter-ending paragraphs, all invariably written as ≥¿Ê. If indeed jiao Ê and jiao – belonged to two different periods, the coexistence of youfenjiao ≥¿– and youfenjiao ≥¿Ê in Shuihu zhuan may also, like many other things, suggest the cumulative nature of the textualizing process. 75. The character “you —” is obviously a homophonic substitute for “you S,” which means “still” or “yet,” but there is a possibility that in early vernacular literature shuo you wei liao °— ºF came into use earlier than the orthographically correct form shuo you wei liao °SºF. In Da Tang Sanzang qujing shihua, the expression appears twice, both written as °—ºF. 76. Patrick Hanan considers such stylistic features, which he heavily depends on in his periodization of the vernacular stories, “at a higher level, analytically,” than other language elements ( The Chinese Short Story, 156). 77. The Chinese Short Story, 157–158. 78. Ibid., 158.

230

Notes to Pages 126–133

79. Hu Zhu’an, “Song Yuan baihua zuopin zhong de yuqi zhuci,” 273ff; Patrick Hanan, The Chinese Short Story, 157. 80. The Chinese Short Story, 158, note 22. 81. “Xiu” occurs nine times in Shuihu zhuan by Hanan’s counting (The Chinese Short Story, 158, note 22). 82. Such features containing “wu” occur very frequently in Shuihu zhuan but are rarely seen in Sanguo yanyi or Xiyou ji. According to Yang Tiange, apart from Shuihu zhuan and the early vernacular stories, these features are seen mostly in dramatic texts attributed to Jin-Yuan writers. Their use seems to have decreased during the Ming and faded away after the Ming. See Yang Tiangge’s article, “Shuo ‘wu.’ ” 83. Lü Shuxiang, Hanyu yufa lunwen ji, 120; Kòsaka Jun’ichi, Baihua yuhui yanjiu, 259. 84. Shuihu cihui yanjiu [xuci bufen], 266. 85. Y. W. Ma has suggested the affinity between the Yang Xiong-Shi Xiu story and the plot in the Yuan Shuihu play, Yan Qing bo yu, by Li Wenwei. See his “Cong zhaoan bufen kan Shuihu zhuan de chengshu guocheng,” 165. 86. Zhong Sicheng, Lu gui bu, 113. 87. Y. W. Ma has noted that, among different parts of Shuihu zhuan, only the portion between the ranking of the heroes (chapter 71) and the capitulation to the court (chapter 82) features actions that parallel the plots in the early Shuihu zaju. From there Ma goes on to suggest that this portion of Shuihu zhuan possibly assumed its textual existence earlier than any other part. The frequent occurrences of “zege” and “wuzi” in section S seems to confirm that chapters 72, 73, and 74 may indeed have had a long textual history, if not necessarily the longest among all the sections. The same, however, cannot be said about section T (chapters 75–82), where the frequency of our linguistic criteria is considerably lower. 88. None of the extant Shuihu plays attributed to Yuan writers feature Lin Chong as a character. In Zhu Youdun’s Baozi heshang zi huansu, Lin Chong is listed among the names of the thirtysix chieftains, but he is still not a speaking character in the play. According to Fu Xihua, the earliest play from the Ming period that was based on the Lin Chong story was probably a chuanqi by Tan Yu, possibly entitled Lin Chong baojian ji, which may have served as the blueprint for Li Kaixian’s Baojian ji. Li Kaixian’s play was in turn followed by Shen Chucheng’s chuanqi, also entitled Baojian ji, and Chen Yujiao’s Ling baodao. See Ma Tiji, Shuihu shulu, 454. Both Li Kaixian’s Baojian ji and Chen Yujiao’s Ling baodao are collected in Fu Xihua, ed., Shuihu xiqu ji, vol. 2. The earliest mention of a nanxi play featuring Lin Chong is probably in the late Ming fiction Guzhang juechen. In chapter 33, several nanxi plays are staged in Minister Dong’s house for the celebration of the Lantern Festival, including one called Lin Chong ye shang Liangshanbo. The publication of Guzhang juechen cannot have been earlier than 1631, as that is the year one of the prefaces to the narrative is dated. See Zhongguo tongsu xiaoshuo zongmu tiyao, 228. 89. At the end of section K, Song Jiang triumphs in the campaign against the village of the Zhu family. Back in Liangshan, he receives a visitor who turns out to be Lei Heng. Infuriated by an insult on his mother, Lei Heng kills a courtesan and is arrested; Lei is released by his friend

Notes to Pages 134–135

231

Zhu Tong, who is consequently exiled. In an effort to force Zhu Tong to join the rebellion, Li Kui kills the prefect’s four-year-old son entrusted in Zhu Tong’s custody, which causes a bitter animosity between Zhu and Li Kui. To appease the wrathful Zhu Tong, Li Kui is ordered to stay temporarily with Chai Jin, and with that, section M—in which Chai Jin is the major character— begins. 90. In Xuanhe yishi, Zhang Qing is one of the twelve escorts for the shipment of the exotic plants and rockery ( huashi gang ) from the south to the capital of Bianliang, while Dong Ping is the constable in Yuncheng County. Dong Ping is the one who is sent by the county magistrate to arrest Chao Gai and his cohort, suspects in a robbery case, only to find them gone. In Shuihu zhuan, it is Zhu Tong and Lei Heng who are the police chiefs in Yuncheng County, and it is Zhu and Lei who are sent to arrest Chao Gai and, later, Song Jiang. 91. The fanben edition of Shuihu zhuan is, in general, “dialogue-heavy,” considerably more so than Sanguo yanyi and Xiyou ji. Some sections, however, may feature fewer dialogues than others. The relatively “dialogue-light” sections in the fanben Shuihu zhuan are those about military campaigns, such as the ones on the expeditions against the state of Liao and the rebellion of Fang La. 92. Here are two examples, both from chapter 2 of the Rongyutang edition of Shuihu zhuan, for the uses of “bian” and “jiu” as adverbs suggesting temporal or logical immediacy: “Ni bian xiachang lai ti yihui” (Come and join the game now); “jiu liuzai gongzhong guole yiye” (then kept [him] overnight in the palace) (1: 36; 1: 37). 93. “Jiu” is polysemous. It can be used as a conjunction to preface a concessive clause, as in the case “Jiu ni buqu, wo ye yao qu” (Even if you don’t go, I will still go). In modern Mandarin it is most often replaced by “jiushi.” “Jiu” can also be used as a preposition or a verb indicating spatial proximity, as in such phrases as “jiu di” (on the spot) and “jiu zuo” (to take the seat). For the polysemy of “jiu,” which has more uses than those mentioned here, see Lü Shuxiang, Xiandai hanyu babai ci, 280–286. In his article, “Cong yuyanshi kan jiben Yuan zaju binbai de xiezuo shiqi,” Mei Tsu-lin affirms that the use of “jiu” as a conjunction as mentioned above appears in such early texts as Hou Han shu (Book of the Latter Han), while “jiu” as a preposition is found in Zhuzi yulei. “Jiu” as an adverb, according to Mei Tsu-lin, began to be used only occasionally during the Yuan period, appearing once in Zhong yong zhijie and twice in Xiaojing zhijie (where one of the two occurrences can also be considered as a preposition). In the thirty Yuan-edition zaju plays, only “bian” is used, with only one instance of “jiu” that, according to Mei Tsu-lin, can be arguably counted as an adverb. The shifting process from “bian” to “jiu” may therefore have started in the Yuan period. My own findings in early vernacular literature support this hypothesis. “Jiu” as an adverb does not appear in the Chan yulu text Bi yan lu, where the adverb “bian” is used in alternation with “sui” or “ji,” its wenyan equivalents. In juan 1 of Bi yan lu, there are fifty-seven instances of the adverb “bian” as compared to twelve occurrences of “sui.” The character “jiu” appears three times, all as prepositions. See Liu Jian and Jiang Shaoyu, eds., Jindai Hanyu yufa ziliao huibian—Songdai juan, 48–77. In Da Tang Sanzang qujing shihua, “bian” occurs twenty-four times while “sui” and “ji” appear twenty-five times. “Jiu” occurs only once, and again as a preposition ( Jindai Hanyu yufa ziliao huibian—Songdai juan, 235–259.) In “Xue Rengui zheng Liao shilue,” both “bian” and

232

Notes to Pages 135–136

“sui” are used. “Jiu” appears six times, but in none of the cases is it an adverb suggesting temporal or logical immediacy. The same can be said of the pinghua texts. In Xuanhe yishi, however, there are two occurrences of “jiu” where it is used indisputably as an adverb interchangeable with “bian”: “jiu mai xie jiu qu zhaoxiang” (54); and “jiu ci jinzhan yu zhi” (105). It should also be noted that a particular form, “yijiu,” occasionally occurs in early vernacular literature. It appears, for instance, once in Wudai Liang shi pinghua and once in the Yuan-edition zaju Li Taibai bian Yelang. As it is interchangeable with “bian,” it may well be a variant form for “jiu” in the early stage of the transitional process from “bian” to “jiu.” 94. Mei Tsu-lin, “Cong yuyanshi kan jiben Yuan zaju binbai de xiezuo shiqi,” 143–144. Mei’s criteria are challenged by Cao Guangshun in his article, “Shishuo ‘jiu’ he ‘kuai’ zai Songdai de shiyong ji youguande duandai wenti.” Cao argues that, similar to the change from “mei” to “men” as the standard plural suffix, there might be two phases in the shift from “bian” to “jiu”: “jiu” began to be used in the Song but became basically out of use during the Yuan before it came back in the Ming. Cao’s description may not be accurate, as the uses of “jiu” in the Song writings that he cites as evidence are more exceptional than regular, as he concedes (290–291). Furthermore, even if “jiu” had been used more often in the Song, the frequency ratio between “bian” and “jiu” in Shuihu zhuan remains useful for our purpose, because only the second phase of the change, the change from the Yuan to the Ming, was relevant to Shuihu zhuan. 95. Here are two examples for “jiang” and “ba” used as verbs for the action of carrying and holding, both from chapter 3 of the Rongyutang edition of Shuihu huan: “Ni fuzi liangge jiang qu zuo panchan” (You, father and daughter, take this to cover your traveling expenses); “Linzi li shide Wang Si de huishu, ba zai xian qian kan” (I picked up the letter Wang the Fourth was carrying in the woods and took it to the county seat to have it read) (1: 94; 1: 80). 96. Wang Li, Hanyu shigao, 410ff. 97. For the evolution of “jiang” and “ba” from pure verbs to “dispositive” prepositions, see Wang Li, Hanyu shigao, 410–418, and Lü Shuxiang, “Ba zi yongfa de yanjiu.” More recently, the different uses of “jiang” as an auxiliary word in premodern Chinese is discussed in Jindai hanyu xuci yanjiu by Liu Jian et al. and in Cao Guangshun’s Jindai hanyu zhuci. 98. Here are two examples for “jiang” and “ba” used as prepositions, both from chapter 1 of the Rongyutang edition of Shuihu zhuan: “jiang danzhao shoucang yu yushujia nei” (placed the imperial edict in a casket for royal documents); “Zhongren zhide ba shiban yiqi kangqi” (The men had no alternative but to join their efforts to lift the stone slab) (1: 17–18; 1: 22). 99. “Qu” has a long history of use in classical literature, meaning “to take” and “to possess.” Obviously, “qu” used in the vernacular is a continuation of its use in classical literature, but the semantic range is somehow expanded. While retaining its original sense, it also means “to fetch” or “to carry” in vernacular usage. In early vernacular literature, it appears in Da Tang Sanzang qujing shihua. While the title itself contains “qu” in its original sense, it occurs in the text in the newer sense, as in the sentence: “Zhangzhe qu dao duyu fashi ( The old man got the knife and passed it to the master) (254). It also appears in Zhuzi yulei: “Ni jiang shao qu qu, zhi de yishao” (If you fetch it with a spoon, you will have only one spoonful) (22). The other words in the group might be of

Notes to Pages 136–149

233

later origins, but we are not sure how late they were. “Na,” which is used in Shuihu zhuan probably less often only than “qu” but more frequently than the others in the group, also appears in the Yuan-edition zaju plays. 100. In my survey, I used the rule of interchangeability to determine whether an instance of “jiu” is an adverb or not. If it can be substituted by “bian” without causing any semantic change to the sentence, it is counted as an adverb and is therefore included in my statistics; otherwise it will be excluded. The same rule applies to the occurrences of “jiang” and “ba”: They will be counted as verbs only if they can be substituted by a verb from the “qu” group. Some occurrences of the verbs from the “qu” group are also excluded when they are not used in the sense of “to hold” or “to carry.” “Qu,” for instance, sometimes appears to mean “to occupy (a city),” and “na” can mean “to arrest” or “to capture.”

Chapter 5: The Engine of Narrative Making: Audience, Storytellers, and Shuhui xiansheng 1. Walter Ong, “Text as Interpretation: Mark and After,” Oral Tradition in Literature, John Miles Foley, ed. (New York: Columbia University Press, 1986), 160. 2. Ibid., 164. 3. Ibid., 4. 4. Georges Poulet, “Phenomenology of Reading.” Reprinted in Contemporary Literary Criticism, Robert Con Davis, ed. (New York: Longman, 1986), 350. 5. “What Is a Text?” In Mythic-Symbolic Language and Philosophical Anthropolog y, David Rsamussen, ed. (The Hague: Martinus Nijhoft, 1971), 139. 6. Eugene Eoyang, “A Taste for Apricots,” 64. 7. See Wolfgang Iser, The Implied Reader. 8. Walter Ong, Orality and Literacy, 78. 9. Paul Ricoeur, “What Is a Text?” 137. 10. See Alan Dundes, “Metafolklore and Oral Literary Criticism.” 11. Sandra Dolby, Literary Folkloristics and the Personal Narrative, 6ff. 12. Eugene Eoyang, “A Taste for Apricots,” 69. 13. See Alexander Woodside and Benjamin A. Elman, “The Expansion of Education in Ch’ing China,” in Elman and Woodside, eds., Education and Society in Late Imperial China, 1600–1900, 525–560, 529. 14. John DeFrancis, Nationalism and Language Reform in China ( Princeton: Princeton University Press, 1950), 222. 15. Richard Solomon, Mao’s Revolution and the Chinese Political Culture, 82. 16. Frederick W. Mote, “China’s Past in the Study of China Today,” 110. 17. See Barbara E. Ward, “Reader and Audiences: The Spread of Chinese Culture,” 189. 18. Evelyn Sakakida Rawski, Education and Popular Literacy in Ch’ing China, 140. 19. “Readers and Audience,” 193. 20. Zeng Zuyin et al., eds., Zhongguo lidai xiaoshuo xu ba xuan zhu, 71.

234

Notes to Pages 149–152

21. This is also evidenced by the ways in which technological information was spread in traditional China. As Barbara Ward observes: “The ways the methods of agriculture, for example, were spread, or other types of technology, or ways of doing business, usually had little to do with reading in China . . . as they had to do with listening to verbal instructions elsewhere” (“Readers and Audiences,” 192). 22. Preface (“Xu”) to Gujin Xiaoshuo. Yushi ming yan, 1. 23. Despite the much-celebrated economic prosperity of the Tang period, there is a general consensus that it was the Song that marks the beginning of fast urbanization and commercialization. In his Commercial Development and Urban Change in Sung China (960–1279), Laurence Ma states: “Economically China during this period experienced a speedy development in her commercial activities whose intensity and scale had been unknown in previous dynasties. There had been, admittedly, similar development during times of disunion, but the degree of commercialization of the Song was far greater than that of any previous age. . . . Commerce, together with the forces of other still less well understood socioeconomic factors, gave rise to a host of new cities” (3–4). 24. Naide Weng, Ducheng jisheng, 100. Xihu Laoren’s Xihu Laoren fansheng lu also gives the figure of one million—which, however, includes households in the suburbs as well (111). In Xianchun Lin’an zhi (Song Yuan difang zhi sanshiqi zhong, no. 7), the number for the households in Hangzhou during the Xianchun period (1265–1274) is given as 391,259 and the number for the population as 1,240,760 (4409). 25. No other city in the Southern Song period exemplified the tendency of commercialization better than the capital Lin’an (Hangzhou). For detailed accounts of the prosperity and various commercial activities in Lin’an during the Southern Song period, see Naide Weng’s Ducheng jisheng, Xihu Laoren’s Xihu Laoren fansheng lu, Wu Zimu’s Meng liang lu, and Zhou Mi’s Wulin jiushi. 26. Marco Polo, The Travels, 213ff. 27. Michael Marmé, “Heaven on Earth: The Rise of Suzhou, 1127–1550,” 25. 28. Marco Polo, The Travels, 212. 29. Quoted in Paolo Santangelo, “Urban Society in Later Imperial Suzhou,” 83. 30. “The Beginning of Popular Chinese Literature: Urban Centers—The Cradle of Popular Fiction,” Archiv Orientální 36 (1968): 67. 31. Xihu Laoren fansheng lu, 123–124. 32. Wu Zimu, Meng liang lu, 298. 33. Zhou Mi, Wulin jiushi, 440–441. 34. Tao Zongyi, Chuo geng lu, 27/18b. 35. Yuan dianzhang 57:2, 780. 36. Li Dou, Yangzhou huafang lu, 251–267. 37. Ibid., 257ff. 38. Storytellers in China were called, variously, shuohuaren, shuohuade, and more recently, shuoshude. For a discussion of such names, see André Lévy, “About the Chinese Storyteller’s Change of Name.”

Notes to Pages 152–156

235

39. Su Shi, Dongpo zhilin. Excerpted in Ming Qing xiaoshuo ziliao xuanbian, 49. 40. “Xu” (preface) to Jingshi tong yan, 1: 7. 41. “Xu” (preface) to Gujin xiaoshuo. In Yushi ming yan, 1. 42. For a brief account of the influence of the popular pro-Shu sentiments on the making of the Three-Kingdom stories, see Wu Zuxiang, “Guanyu Sanguo yanyi.” 43. See Eugene Eoyang, “A Taste for Apricots,” 65ff. 44. Ma Nancun, “Sanzhong Zhuge Liang,” in his Yanshan yehua, 315–318. 45. Bai Xiuying’s story is entitled “Yuzhangcheng Shuangjian gan Su Qing” (Shuangjian Pursues Su Qing to Yuzhangcheng). The love affair between Shuangjian, a young scholar, and Su Xiaoqing, a famous courtesan, was a popular subject matter in different forms of performative art during the Song-Jin-Yuan periods. Tao Zongyi’s Chuogeng lu, for example, lists Tiao Shuang jian among the titles of yuanben (25/8b). Several plays in zaju and nanxi, according to Zhao Jingshen, were based on the same story, including the anonymous Yuzhangcheng ren yue liang tuanyuan. See Zhao Jingshen, Yuan Ming nanxi kaolue, 6. In Zhong Sicheng’s Lu gui bu, the Yuan playwright Wang Ye is said to have composed Shuang jian Xiaoqing wenda, in collaboration with Zhu Shikai (135). In Dong Jieyuan’s Xixiang ji zhugongdiao there is a disclaimer that “This is not the story of Shuangjian Yuzhangcheng either” ( Ming Jiajing ben Dong Jieyuan Xixiang ji, 1/2b), from which we may assume that there might have been a zhugongdiao version of the story sharing the same formal features, especially the alternation of speech and singing, as Xixiang ji zhugongdiao. Indeed, Bai Xiuying’s story in Shuihu zhuan may well be a zhugongdiao, as she is said to be good at zhuban pindiao. This is also confirmed by the prosimetric form of her performance, as she is said to be “singing and speaking, and then singing again.” 46. Shuihu zhuan huiping ben, 934. 47. Qian Cai, Shuo Yue quanzhuan, 1: 191–196. 48. Mark Bender, “Key to Performance in Kunming Storytelling,” 26. 49. Vibeke Børdahl, “A ‘Poetics’ of Chinese Storytelling,” 233. 50. Nie Gannu, “Shuihu wulun,” 10–17. That the Song Jiang bandits were said to be based in the Taihang Mountains in an early stage of the oral complex is evidenced by Gong Shengyu’s “Song Jiang sanshiliu zan.” In the panegyrics on Lu Junyi, Yan Qing, Zhang Heng, Dai Zong, and Mu Hong, Gong mentions Taihangshan five times, while the name of Liangshanbo never appears. Xuanhe yishi may mark the start of the transition of the locale from the Taihang Mountains to Liangshanbo, for the two places—hundreds of miles apart from each other—are sometimes mentioned side-by-side in the curious combination “Taihangshan-Liangshanbo” as the base of the rebellion. 51. See Gong Shenyu, “Song Jiang sanshiliu zan,” in Zhou Mi, Guixin zashi xuji, 276–287. 52. Xuanhe yishi, 55. 53. Ducheng jisheng, 97; Meng liang lu, 309. 54. “Zhuangjia bushi goulan,” in Zhongguo wenxue xinshang quanji (Taipei: Zhuangyan Chubanshe, 1983), 25: 8–12. 55. Shuo Yue quan zhuan, 1: 194–195.

236

Notes to Pages 156–163

56. Xihu Laoren, Xihu Laoren fansheng lu, 120. 57. Shuihu xinyi, 40. 58. Robert E. Hegel comments insightfully on the creation of the bandit heroes in Shuihu zhuan: These characters had their genesis in popular entertainment; no audience, literate or otherwise, wishes to hear real events recounted without selectivity or embellishment. If the aim of the entertainer is to satisfy his audience, he must side with them, not attack them; he must confirm their ideals and perceptions of life. Water Margin heroes do just that, at least in part, by confirming the image that was popular—and most likely accurate—through time, of an administrative and legal system less concerned with matters of right and wrong than with personal privilege and preserving the status quo. (The Novel in Seventeenth-Century China, 75) 59. The Classic Chinese Novel, 81. 60. Jiao Xun, annotator, Mengzi zheng yi, 3: 5, 94; James Legge, trans., The Works of Mencius, 125–126. 61. E. A. Kracke, “Region, Family, and Individual in the Chinese Examination System,” 253. 62. Ping-ti Ho, Ladder of Success, 11. 63. A conspicuous instance is to be found in the early Ming period. In 1370, the emperor of Hongwu decreed that examinations be held on a regular triennial basis, but the tension between the emperor and his top civil officials was such that he had the system suspended in 1373; he did not restore it until 1384. During the suspension of the examinations, recommendation became the rule, which, however, cost the life of a prime minister, who was executed by the emperor on the charge of cronyism. See Edward Dreyer, Early Ming China, 98–101 and 131–133. 64. Ping-ti Ho, Ladder of Success, 46. 65. The jinshi examination, conducted at the highest level, was usually composed of the metropolitan examination supervised by top officials and the palace examination set by the emperor himself. But in order to participate in the jinshi examination, a scholar had to acquire the preliminary degree of sheng yuan and the intermediate degree of juren by passing the examinations at the local and provincial levels. For a more detailed account of the three levels of the examination system, see Ping-ti Ho, The Ladder of Success in Imperial China, 26ff. 66. Wang Daocheng, Keju shihua, 20. 67. Tian Rucheng, Xihu youlan zhiyu, 402. 68. See Ping-ti Ho, Ladder of Success, 171. For the coordination and alignment between the curriculum at the public schools and the contents of civil service examinations in the Ming period, see Benjamin A. Elman, “Changes in Confucian Civil Service Examinations from the Ming to the Ch’ing Dynasty.” 69. See Ping-ti Ho, Ladder of Success, 175. Benjamin A. Elman has noted perhaps an even more

Notes to Pages 163–166

237

dramatic increase of the number of licentiates at the national level, which grew from an estimate of thirty thousand for the year of 1400 to approximately five hundred thousand for the year of 1700 (“Changes in Confucian Civil Service Examinations,” 117). 70. See Ping-ti Ho, Ladder of Success, 175–183. 71. Ju-k’ang T’ien, Male Anxiety and Female Chastity, 84. Benjamin A. Elman’s numbers are slightly different, but they present a picture equally grim: Only 1.5 percent of participants in the country examinations achieved licentiate status; only 5 percent of licentiates could pass the provincial examinations and receive a juren degree; and only 20 percent of the juren would be able to pass the metropolitan examination and become jinshi. The chance for success in all three levels of examination was therefore one in six thousand. To be sure, Elman’s numbers refer to the situation in the mid-nineteenth century, but the examinations in the Ming period, with fewer jinshi graduates per annum than in the Qing, could not be much less competitive. See Elman, “Changes in Confucian Civil Service Examinations,” 117. 72. Ping-ti Ho, Ladder of Success, 27–28. 73. Male Anxiety and Female Chastity, 101. 74. For a full account of the issue, see Ju-k’ang T’ien, Male Anxiety and Female Chastity, 90–113. 75. In Meng liang lu, seven names are listed for xiaoshuo, four names for jiang jing, and seven names for jiangshi (312–313). The list in Wulin jiushi is more complete, giving twenty-three names for jiangshi, seventeen names for jiang jing, and as many as fifty-two names for xiaoshuo. It also lists raconteurs’ names in other oral genres (453–466). 76. See Zhou Mi, Wulin jiushi, 454. 77. See Hu Shiying, Huaben xiaoshuo gailun, 65–68; also Yan Dunyi, Shuihu zhuan de yanbian, 223. 78. Zhou Mi, Wulin jiushi, 454. 79. Zhong Sicheng, Lu gui bu, 252, note 1206. 80. Ibid., 204, note 605. 81. Xiao Sun tu deng sanzhong, 1a, 15a, and 54b. 82. Ming Chenghua shuochang cihua congkan, 817. 83. Zhu Youdun, Tao Yuanjing (in Shemotashi qucong), 1a. 84. Jingshi tong yan, 2: 1121. 85. Qingping shantang huaben, 11. 86. Ming Chenghua shuochang cihua congkan, 393. 87. Sentences that acknowledge the contribution of the laolang to previous versions of the narrative are found in several vernacular stories, such as: “This is the shuohua passed down from the laolang” in “Chen Yushi qiao kan jin chaidian” ( Yushi ming yan, 24) and in “Zhang Xiaoji Chenliu ren jiu” ( Xingshi heng yan, 1: 343); and “This huaben originally circulated among the laolang in the capital” in “Shi Hongzhao longhu fengyun hui” ( Yushi ming yan, 152) and in “Kan pixue dan zheng Erlangshen” ( Xingshi heng yan, 1: 273). 88. Shuihu zhuan de yanbian, 129. 89. Huaben xiaoshuo gailun, 57.

238

Notes to Pages 166–173

90. The Evolution of a Chinese Novel, 43. 91. It is interesting to note that the mention of shuhui is deleted in the Guanhuatang edition, as another piece of evidence for Jin Shengtan’s deliberate attempt to distance the narrative from its origins in popular orality. 92. The Evolution of a Chinese Novel, 49. 93. Ibid., 59. 94. Shuihu zhuan de yanbian, 223. Yan Dunyi goes on to suggest the possibility that Shi Nai’an could be an offspring of Shi Hui, a famous storyteller in the Southern Song period. 95. Jack Goody and Ian Watt, “The Consequence of Literacy,” 37. 96. Mao Dun, “Zhongguo wenxue buneng jianquan fazhan zhi yuanyin,” in Mao Dun gudian wenxue lunwen ji, 157. 97. In juan 70 of Ming shi, it is stated: “The examination system follows the rules of the Tang and Song periods, with minor adaptations. Topics are to be taken exclusively from the Four Books and Five Classics. This was decided by Emperor Taizu and Liu Ji. The examination essays are to imitate the Song people’s exegeses of the classics, but to be written in the style of the ancients. The essays are to be in a paralleled structure, which is called bagu.” See Zhang Tingyu et al., Ming shi, 1: 722. 98. The prescribed essay form consisted of eight parts, which were set in a symmetrical structure like four pairs of legs—hence the name Eight-Legged Essay. In the first leg, technically described as the part to “break open the topic,” the candidate was supposed to introduce his topic by citing the quotation from the classic text. The second leg was the part to “accept the topic”— that is, to declare the general treatment of the topic in the essay. The third and fourth legs were two introductory paragraphs, parallel to each other both in structure and in diction, which were followed by another pair of legs, also in the parallel pattern, which were to elaborate on the topic. The final two legs were supposed to wind up the composition and bring about a completeness to the whole piece. 99. Wu Jingzi, Rulin waishi, 193. 100. The word “gongming” literally means “achievement” and “fame,” but in the context of the Ming-Qing society, the term was almost always associated with entrance into officialdom through success in the civil service examinations. 101. “Zhong yi Shuihu zhuan xu,” Shuihu zhuan ziliao huibian, 192. The prefatory piece claims to be composed by Li Zhi. A photocopy of it is appended to Li Zhuowu piping Zhong yi Shuihu zhuan. It is also collected in Li Zhi’s Fenshu ( Fenshu, Xu Fenshu, 109–110). 102. In the 1970s, it was, of course, Mao Zedong’s interest in Song Jiang’s surrender to the emperor that triggered in the People’s Republic of China a critical reevaluation of Shuihu zhuan and the character of Song Jiang. It quickly escalated into a political campaign with strong implications for the power struggle within the leadership of the Communist Party. 103. For example, one of those moments occurs in chapter 32, where Song Jiang asks Wu Song, who is on his way to join the band led by Lu Zhishen and Yang Zhi, to persuade the chieftains to surrender when the opportunity comes: “If the imperial offers amnesty for you, urge Lu

Notes to Pages 173–174

239

Zhishen and Yang Zhi to surrender, too. Some day you may be sent to the frontier. By fighting one battle after another you’ll earn honors for your family and leave behind a good name in history so that you won’t have lived in vain. I myself am a man of no talents. Though I’m loyal to the emperor, I cannot advance. But you are so courageous, you will surely become a high-ranking official” (2: 1021). 104. As one who recognizes and appreciates Yang Zhi’s worth, the commander, Liang Zhongshu, is eager to see Yang Zhi win, but his partiality does not in any way facilitate Yang Zhi’s fight with Zhou Jin. Actually, for the contest on archery, Yang Zhi voluntarily puts himself in a disadvantageous position by letting his adversary shoot at him first. Before Yang Zhi’s contest with Suo Chao, Liang Zhongshu lends his battle charger to Yang Zhi, but that favor is soon more than counterbalanced as Li Cheng, Liang Zhongshu’s chief general, lends to Suo Chao both his battle charger and a set of armor. 105. The ranking of the bandit chieftains in Liangshan takes place five times in the narrative: (1) After Lin Chong joins the band, the chieftains are ranked, with Wang Lun as the head (chapter 12); (2) when Chao Gai and his men come to Liangshan after their robbery of the birthday gift, Lin Chong murders Wang Lun, and the chieftains are re-ranked, with Chao Gai as the new head (chapter 20); (3) after the arrival of the group led by Qin Ming and Hua Rong, the chieftains are again re-ranked (chapter 35); (4) when Song Jiang finally joins the band, the chieftains are once again re-ranked, with Chao Gai and Song Jiang in the first and second places (chapter 41); and (5) the heroes are ranked for the last time after their number reaches the full 108, with Song Jiang and Lu Junyi in the first two places. Their rank is said to be “decided by Heaven, with a general division into higher and lower” (chapter 71). 106. It is interesting to note here this parallel between Shuihu zhuan and the Qing fiction Jinghua yuan by Li Ruzhen (ca. 1763–1830). In Shuihu zhuan, the ranking of the 108 chieftains is apparently determined by the celestial destinies of the 108 demons that Minister Hong accidentally released at the beginning of the narrative. In chapter 71, a stone tablet is discovered with “tadpole script” on it. The script, as interpreted by a Daoist priest, turns out to be a list of the names of the chieftains ranked in the order of the thirty-six stars of Heavenly Spirits and seventytwo stars of Earthly Fiends, which are the preincarnate identities of the 108 warriors. In chapter 48 of Jinghua yuan, a white jade tablet is discovered with “tadpole script” on it, which, as decoded by Tang Xiaoshan, is a list of the names of the one hundred girls, who are human incarnations of celestial flower spirits. Then, in chapter 67, when the list of the successful candidates of the examination is publicized, the ranking of the one hundred selected girls turns out to be identical with the one presented in the inscription on the tablet. 107. Song Jiang is buried in Liaoerwa in Chuzhou, where he served as magistrate. Although in previous chapters Liaoerwa is regarded as part of Liangshanbo, in the end of the narrative it becomes a separate locus hundreds of miles away from there. As three of Song Jiang’s confidants— Wu Yong, Hua Rong, and Li Kui—are interred there to join their beloved leader, it is possible to consider Liaoerwa as a double for Liangshanbo, or the place for another gathering of disgruntled heroes, even after they have received official appointments. The circular pattern of the narrative

240

Notes to Pages 174–179

may suggest the futility of their quest for gongming. As one scholar proposes, the name of the place, Liaoerwa, is a pun on “liao” (“to finish”), which may suggest that this is the place the heroes’ quest, and indeed their lives, must end. See Yenna Wu, “Outlaws’ Dreams of Power and Position in Shuihu zhuan,” Chinese Literature: Essays, Articles, Reviews 18 (1996): 45–67. 108. Sima Qian (Hong Kong, 1975), 294. 109. Ibid. 110. “Zhong yi Shuihu zhuan xu,” Shuihu zhuan ziliao huibian, 192; Fenshu, Xu Fenshu, 109. 111. The edition is dated around 1620, published by Xiongfeiguan, with prefaces by Xiong Fei and Yang Minglang. 112. Yang Minglang, “Yingxiong pu juanshou,” reprinted in Shuihu zhuan ziliao huibian, 231. 113. Wang Tao, “Diwu caizi shu juanshou,” reprinted in Shuihu zhuan ziliao huibian, 374. 114. “How to Read the Fifth Book of Genius,” John C. Y. Wang, trans. and David Rolston, annotator, in How to Read the Chinese Novel, David Rolston, ed., 131; “Du Diwu caizi shu fa,” in Shuihu zhuan huiping ben, 15. 115. Hu Shi, “Shuihu zhuan kaozheng,” in his Zhongguo zhanghui xiaoshuo kaozheng, 58. 116. “How to Read the Fifth Book of Genius,” 131, note 5. 117. Shuihu zhuan huiping ben, 38. 118. Ibid., 342. 119. A well-known example of venting personal grievances indirectly in literary writings is the meiren xiangcao (beautiful women and fragrant flowers) tradition originated with Qu Yuan (340?278? b.c.). In his long poem, Lisao (Encountering Sorrow), Qu Yuan uses the imagery of beauties and plants to refer allegorically to his own lofty personality under the siege of political calumnies. This practice became a literary convention for writers of later ages. For instance, Hellmut Wilhelm has found that in a certain type of fu—a peculiar semipoetic and semiprosaic literary form that flourished during the Han Dynasty (206 b.c.–a.d. 220)—the stereotyped content of the complaints of a neglected wife is often the indirect expression of an official out of favor with the sovereign. See Wilhelm, “The Scholar’s Frustration: Notes on a Type of Fu.” 120. Li Zhi, “Za shuo,” in Fenshu, Xu Fenshu, 97.

Chapter 6: Literary Vernacular and Novelistic Discourse 1. On the rise of European literary vernaculars following the end of Roman times, Robert Scholes and Robert Kellogg write in The Nature of Narrative: In the European vernacular languages an oral literature developed along the usual lines, gradually becoming aware of the written classical literature and gradually becoming itself a written literature. The existence of a variety of oral vernaculars, in the presence of the ubiquitous Latin, established conditions for literary development in Western Europe which were different from those that had prevailed in Greece and Rome. The awareness of the existence of writing as a medium for conveying narrative almost certainly resulted in Western European epic being reduced to writing. (247)

Notes to Pages 180–183

241

2. Interface of the Word, 257. 3. For the function of writing as a means to separate the knower from the known and its relationship to the rise of analytical learning, see E. A. Havelock, Preface to Plato, 197–233. 4. Feng Menglong, “Xu” (Preface) to Gujin xiaoshuo, in Yushi ming yan, 1. 5. Northrop Frye, Anatomy of Criticism: Four Essays, 33–34. In one of his early articles, Patrick Hanan finds Frye’s “modes” in narrative forms useful and substitutes them for the notions of “lifesized character” and “larger-than-life character” that he had previously adopted in a study of Chinese vernacular fiction. See Patrick Hanan, “The Early Chinese Short Story.” 6. Georg Lukacs, The Theory of the Novel, 66. 7. M. M. Bakhtin, “Epic and Novel: Toward a Methodology for the Study of the Novel,” in his Dialogic Imagination, 15. 8. Sheldon Hsiao-peng Lu, From Historicity to Fictionality, 39. 9. The Chinese conception of creation and originality differs significantly from that in the West. The classical exemplar of the Chinese preference for participation and transmission to invention and innovation is of course the often-cited Confucian aphorism in Lunyu (The Analects): “Shu er buzuo” (Transmission/elaboration but not creation/origination). And this Chinese skepticism on the power of willing something into existence out of nothing is often linked to the lack of a Chinese creation myth, which is said to have exerted a great impact on the Chinese concept of literary creation as well as that of the cosmological origin as a whole. For an account of the Chinese view of the creative process, see Thomé H. Fang, “The World and the Individual in Chinese Metaphysics.” 10. Karl S. Y. Kao observes that “the Six Dynasties attitude towards the zhiguai revolved essentially around the question of facticity: the phenomena recorded were in general accepted as real.” See his Classical Chinese Tales of the Supernatural and the Fantastic, 21. 11. To verify the relative sketchiness of the narrative in wenyan, one can compare many of the narrative segments in wenyan works with those vernacular narratives that are based on their wenyan precursors. For instance, the legendary story of Emperor Shun, told in Shiji in a most skeletal brief, was later expanded into a detailed account in the “Shunzi zhixiao bianwen.” Sun Kaidi also mentions that a brief anecdote written in wenyan by the Ming writer Xu Hao—an anecdote only 124 characters long about the Jin (265–420) scholar Chen Shou—was later fleshed out by Feng Menglong into a vernacular story running over nine thousand characters in length (Sujiang, shuohua yu baihua xiaoshuo, 8). What Lu Xun says about some residual wenyan epithets probably serves as testimony to the impotence of the literary language in presenting physical details: Suppose a meticulous reader invites me over and hands me a pencil and a piece of paper, asking: “In your writings you often use such words as lengceng and chanyan to describe mountains, but what on earth do those mountains look like?” I would sweat with shame and wish to find a hole in the ground to hide myself, for I have no idea myself what those words really mean. They are epithets taken over from books of the old times, and nobody has ever clarified their meaning. Once you take them seriously, it becomes very

242

Notes to Pages 184–195 embarrassing. (See Lu Xun, “Rensheng shizi hutu shi,” in Lu Xun lun wenxue yu yishu, 2: 837)

12. The Nature of Narrative, 84. 13. This phrase appears in the chapter of “Yiwenzhi” in the Hanshu, juan 30, and Liu Xie cites it in the Wenxin Diaolong. See Fan Wenxin diaolong zhu, 2: 283. 14. Liaozhai zhiyi, 32. 15. Jaroslav Prusek, “The Beginning of Chinese Popular Literature,” Archiv Orientální, 36 (1968): 115. 16. This refers to the lines in Li Shangyin, “Jiaoer shi”: Huoxiao Zhang Fei hu, huoxiao Deng Ai chi. The poet expresses his delight in seeing his five-year-old son making fun of one of the father’s guests who stuttered like Deng Ai. It suggests that by the late Tang period the storyteller’s dramatic rendering of the Sanguo story cycles had become so popular that even a five-year-old was acquainted with the general’s peculiar way of speaking. 17. Hu Shiying, Huaben xiaoshuo gailun, 375. 18. Zhang Dai, Taoan meng yi (Shanghai: Shanghai Yuandong Chubanshe, 1995), 147. 19. Hu Shiying, Huaben xiaoshuo gailun, 372. 20. Børdahl, “Narrative Voice in Yangzhou Storytelling,” CHINOPERL Papers 18 (1995): 4. 21. Shuihu zhuan huiping ben, 296–297. 22. “How to Read The Fifth Book of Genius,” 134; “Du Diwu caizishu fa,” 17. 23. Jin Shengtan’s interlinear comment in chapter 14 of the Guanhuatang edition, in Shuihu zhuan huiping ben, 275. 24. Jin Shengtan’s interlinear comment in chapter 25 of the Guanhuatang edition, in Shuihu zhuan huiping ben, 1: 488. 25. “How to Read The Fifth Book of Genius,” 135; “Du Diwu caizishu fa,” 17. 26. This is an interlinear comment in the “Xiezi” (Prologue) of the Guanhuatang edition. See Shuihu zhuan huiping ben, 49. 27. Shuihu zhuan huiping ben, 696–697. 28. “Xu san” (The Third Preface), Shuihu zhuan huiping ben, 9. 29. Liu Xinzhong, Jin Shengtan de xiaoshuo lilun, 20–22. 30. Ibid., 21. 31. Chapter comment on chapter 55 of the Guanhuatang edition, Shuihu zhuan huiping ben, 2: 1018. 32. Sun Kaidi, Sujiang, shuohua yu baihua xiaoshuo, 12. 33. Andrew H. Plaks, “Full-length Hsiao-shuo and the Western Novel: A Generic Reappraisal,” 175. 34. Ibid., 168. 35. Ian Watt, The Rise of the Novel, 11. 36. Ibid., 32. 37. Ibid., 27, 29. 38. Ibid., 101.

Notes to Pages 195–196 39. Ibid., 194. 40. Plato, The Republic, Desmond Lee, trans. (Penguin, 1974), 150. 41. Ibid., 150. 42. Gérard Genette, Narrative Discourse: An Essay in Method, 164. 43. M. M. Bakhtin, “Discourse of the Novel,” in his Dialogic Imagination, 335. 44. Ibid., 336.

243

Glossary Note: Personal names and titles of publications given in the Selected Bibliography are not listed here. A Ying ¸^ an Õ ba ‚ bagu wen K— bagua Kˆ Bagua zhen Kˆ} Bai Juyi ’~ˆ “Bai Niangzi yong zhen Leifengta” ’Ql√Ìpp Bai Pu ’Î Bai Renfu ’Øj Bai Xiuying ’q^ baiguan ˚x baihua ’‹ Baiju ’s “Baiju Chang zhuiyu” ’sıÛy Baitu ji ’flO Baiyue ting ÙÎF bajiufen KE¿ “Bangmang wenxue yu bangxian wenxue” ∞£Â«P∞~« Bao Longtu ]sœ “Bao Longtu anduan wai wupen zhuan” ]sœ◊_nQ÷« “Bao Longtu Chenzhou tiaomi ji” ]sœØ{˝ÃO “Bao Longtu duan baihu jing zhuan” ]sœ_’ÍΫ

“Bao Longtu duan Cao Guojiu gong’an zhuan” ]sœ_‰Í§ Ω◊« “Bao Longtu duan Zhao huangqin Sun Wenyi gong’an zhuan” ]sœ_ Ø”À]ˆΩ◊« Bao Shizhi ]ÕÓ “Bao Shizhi chushen zhuan” ]ÕÓ X≠« Bao Xi ]Î Bao Zhao j” Bao Zheng ]@ Baojian ji _CO baojuan _˜ Baozi heshang zi huansu \lM| ¤ŸU bei O Beijing Damingfu _?jW≤ beiqu _± benshi ª∆ bian s bian K bian ‹ Bianjing X? bianjiu KN bianjiu candeng xia shaole KN›OU NF Bianliang XÁ

245

246

Glossary

bianwen ‹Â bijing ¶∫ bili zhi ci ¿Z߸ binbai ´’ Bing Yang Xiong f®Ø Bingma dujian L®£? bishang Liangshan GWÁs Biyan lu —•˝ bu fu chaoting £t¬? bu li wenzi £flÂr buduoshi £h… butong wenmo £q• buzai huaxia £b‹U cai ~ Cai Bojie Pipa ji ≤Bn\]O Cai Jing ≤? Cai Yidian ≤M cairen ~H caixue ~« Caiyuanzi ÊÈl canben ›ª Cang Jie ‹e Cangzhou ‹{ “Cao Boming cuo kan zang ji” ‰B˙ ˘…BO Cao Cao ‰fi Cao Zheng ‰ø Chai Jin „i chang qunr ¯»‡ changnuodao ¤’D changzhuan ¤» Chao Gai –\ chaoting fu wo ¬?t⁄ chaoyutiwen WyÈ “Chen Kechang Duanyang xianhua” Øi`›ßP∆ Chen Shou Øÿ “Chen Xunjian Meilin shiqi ji” صÀ ˆ≠¢dO Chen Yujiao ØP•

cheng ” Cheng Hao {V Cheng Yi {[ Chenghua ®∆ chi ˘ chiren shuomeng ˆH°⁄ chong R˛ƒ “Chongkan Songben Xuanhe yishi ba” ´Z∫ª≈MÚ∆[ chongke qi shu ´Ë‰— chu ruowei BY∞ chu ruowei chenshuo BY∞Ø° Chu Zhao-wang shuzhe xiachuan °L˝®ÃUÓ chuangzuo lian –@Ï chuanqi «_ Chunqiu KÓ chunqiu bifa KÓßk chuzhi shi Bm° ci ¸ Cihai „¸ cihua ¸‹ “Cui Daizhao shengsi yuanjia” Z›@ Õ∫fia “Cui Yanei baiyao zhaoyao” Z≈∫ ’c¤Ø Cui Yingying Zaa cun ¯ “Cuo ren shi” ˘{r “Da Deng Shiyang” ™H¤ß Da Tang Sanzang fashi qujing ji j T√kv˙gO “Da xu” j« da yi kan shi ¥@›… dadao ™D Dafeng j◊ daguanren jxH dai π˛a˛N Dai Zong πv dandan ¶¶˛ÊÊ

Glossary

Danfu 7˜ Dangchu zhiyou sanfen fannao . . . you tianle wufen fannao ÌÏu≥T¿ –o...SKF≠¿–o danjian ˝£ dao you wei liao DSºF daode Ïo Daode jing Dwg daoshi DO Daxue zhijie j«Ω— de o˛∫ “Deng Daleian yu mei shu” njp§ Pf— Dengzhou n{ desheng lishi touhui w”Q´Y^ di ∫˛≥ Di Qing fC diaose J‚ diben ≥ª Ding Zhenghua Bøÿ dique ∫T disha xing aŸP Dong Ba ≥Q Dong Chao ≥W Dong Ping ≥≠ Dong Qichang ≥‰˜ “Dong Yong yu xian ji” ≥√JPO Dongchuang shifa F°∆o Dongpingfu F≠≤ Dongpo zhilin FY”L Dongshan }s Dongwu nongzhu ke FdÀ]» Dong-Xi Han tongsu yanyi FË~qU tq Dongyuan FÏ Du Qian ˘d˛˘E duande ›∫ Dunhuang ∞◊ Duode, Ganniang bianjiu shoule h∫A zQKN¨F

247

Er pai GÁ erda G¥ ezi jiduo _r•h Facong ko fafen zhushu o´¤— faji biantai oÒ‹ı Fan Lihua ‘˘· fanben cª Fang La Ëæ fanqie œ¡ faxie buping on£≠ Fei Huzi ÚÍl “Feicui xuan” BAa “Feng Boyu fengyue xiangsi” æB… ∑Τ‰ Fenglao geban, zaihuo qiansheng ^“ qÒAAMen Fengpei ◊K “Fengyue ruixian ting” ∑ÎÁPF “Fengyue xiangsi” ∑Τ‰ Feng yun hui ∑≥| “Foyin Shi si tiao Qinniang” ÚLv |’^Q fu · “Fu Lu Shou san xing du shi” ÷Sÿ TP◊@ ganbang ÏŒ gang ˲ı˛˚ Ganniang raoshu zQ«§ Ganniang raoshu zege zQ«§h” Ganwen age, ni jiao shenme ±›¸ÙA As∞Ú Ganwen guanren gaoxing daming ±› xH™mjW Gao Ming ™˙ Gao Qiu ™fi Gao Ru ™ß Gao Wenxiu ™Âq Gao You ™§ Gaotangzhou ™{

248

Glossary

Gejiang douzhi jøʺ Gengzhi Zhang Qian ti sha qi «Ωid ¿˛d gewu Ê´ gong tu qi laosao buping @R‰cã≠ gong’an Ω◊ gongming \W Gongsun Sheng Ω]” goulan ƒÊ Gu Dasao UjA “Gu fen” t´ Guan Dawang san zhuo hong yi guai ˆj˝TªıÁ« Guan Hanqing ˆ~Î Guan Suo ˆ¡ Guan Yu ˆ– Guan Yunshi e≥¤ Guan Zhang shuang fu Xishu meng ˆi ˘uËæ⁄ Guangwu ˙Z GuanhuatangeÿÛ guanxi dahan ˆËj~ guaren ËH gui ≠ Gui er ji Q’∞ Gujin xiaoshuo jµp° Guo Wei ¢¬ Guo Xun ¢‘ Guo Ying ¢^ Guochao Yinglie zhuan ͬ^P« guojiu Íh˛Í§ Guoyu Íy gututu © guwen j Guzhang juechen ™x¥– guzici ™l¸ Han Fei ˙D “Han Li Guang shihao Fei Jiangjun” ~ıs@π∏Nx Han Yu ˙U

“Hanjiang Wang Ling bian” ~N˝ Æ‹ Hanshi ~v Hanshu ~— he X He Jiu-shu ÛE˚ Heixuanfeng fujing ¬¤∑t Heixuanfeng qiao duan’an ¬¤∑Ï_◊ Heixuanfeng qiao jiaoxue ¬¤∑Ï–« Heixuanfeng zhang yi shucai ¬¤∑Mq ®] “Hetong wenzi ji” XPÂrO Hongzhi ∞v “Hongzi” Li Er ırıG Hou Han shu ·~— “Hou Meng zhuan” JX« hu G Hu Xu J÷ Hu Zhong Jæ Hu Zhongbin JÚl Hua Guan Suo ·ˆ¡ “Hua Guan Suo chushen zhuan” ·ˆ¡X≠« “Hua Heshang” ·M| Hua Rong ·a huaben ‹ª huaben xiaoshuo ‹ªp° “Huadengjiao Liannü cheng fo ji” ·O‚¨k®ÚO Huai Lin hL Huainanzi anl Huang Chao ¿_ Huang ming cong xin lu ”˙qH˝ Huang Pilie ¿AP Huang Song shi chao gang yao ”∫Q¬ ın Huangdi ¿“ Huanghua yu ¿·n huashi gang ·¤ı huashuo ‹°

Glossary

Huhai sanren Ú¸≤H Huizong ≤v hulala øff Hunyuan yiqi zhen V∏@} Huo Si N| Huoxiao Zhang Fei hu, huoxiao Deng Ai chi Œ∫i∏GAŒ∫H„Y Huyan Chuo IµÔ Huyan Zhuo Iµ` ji Y “Ji Yafan jinman chanhuo” p„f˜¡ £◊ Ji Yun ˆ˚ Jia Yi Î≥ jiafu µg Jiafu zhi qing, ai he dai yan µgß°A RÛ›• Jiajing ≈t Jiajing canben ≈t›ª Jian’an ÿw jianben ≤ª jiang N jiang danzhao shoucang yu yushujia nei N¶@¨√Ûs—X∫ Jiang Daqi ±jπ “Jiang Shuzhen wenjing yuanyang hui” ±QuFVpm| “Jiang Xingge chonghui zhenzhu shan” ±≥Ù´|√]m jianghu haohan øÚn~ Jianghu sanren øÚ≤H jiang ji NŒ jiang jing øg jiang jing wen øg Jiangnan øn jiangshi øv “Jiantie heshang” ≤´M| jiao –˛Ê jiaoben }ª “Jiaoer shi” ∫‡÷

249

jiaofang yuebu –{÷° jican kejin, yezhu xiaoxing »\˜ºA ]ÌÂÊ “Jie zhong de shu xuji” T§o—ÚO jiedushi `◊œ jietan xiang yu ÛÕ—y “Jiezhier ji” Ÿ¸‡O Jiguge Vj’ jing w˛∫ Jingben Zhong yi zhuan ?ªæq« Jingchai ji ¶O Jing-Kang baishi td˚v “Jinming Chi Wu Qing feng Aiai” ˜˙¿dM{RR jinshi ih jinxi ⁄§ jinxin xiukou Afl∏f Jishi yu Œ…B jiu N jiu bi you ben ¬≤≥ª jiu ci jinzhan yu zhi NÁ˜¯Pß jiu di Na jiu liuzai gongzhong guole yiye Ndb c§LF@] jiu mai xiege jiu qu shaoxiang NR«” shNª Jiu ni buqu, wo ye yao qu NA£hA ⁄]nh jiu zuo Ny jiubian NK Jiugong bagua zhen EcKˆ} Jiu-san Shuhui ET—| jiushi NO Jiutian Xuannü zhen E—»k} Jizhou ¨{ juanshou ˜∫ jueju ¥y jufa yk juren |H Kaifeng }?

250

Glossary

kan niangzi zuo shenghuo ›QlµÕ° kan niangzi zuo shenghuo zege ›Ql µÕ°h” “Kan pixue dan zheng Erlangshen” …÷uÊ“G¶´ Kang Jinzhi diß kanshi ›… Keqin J‘ Kong Sanzhuan ’T« “Kong Shufang shuangyu shanzhui zhuan” ’Q⁄˘Ω∞Y« Kong you yishen benshi ≈≥@≠ª∆ Kong you yishen benshi, buyu mingzhu ≈≥@≠ª∆A£J˙D “Kuaizui Li Cuilian” ÷LıA¨ “Kuiguan Yao Bian diao Zhuge” ‹ˆ ¿ÀQ—Ø Kun [ Kunming pingshu ¯˙˚— kunqu ¯± Lang Ying ¶Î “Lanqiao ji” ≈ÙO Lao Huihui donglou jiaofo —^^F” sÚ laolang —¶ laoniang —Q laoshen —≠ Laozi zhu —l` lei L Lei Heng pÓ Li Gu ıT Li Heng ıÓ Li Jingde §tw Li Jinyi ıiq Li Jun ıT Li Keyong ıJŒ Li Kui ıf Li Li ıfl “Li Ling bianwen” ıÆ‹Â Li Mengyang ı⁄ß Li Mi ıK

Li Sanniang ıTQ Li sao ˜Ã Li Shangyin ı”Ù Li Shishi ıvv Li Shizhong ı…§ Li Si ıµ Li Song ıC Li Taibai bian Yelang ı”’S]¶ Li Wan ıU Li Wenwei ı´ “Li Yuan Wujiang jiu zhushe” ı∏ døœ∂D Li Zhi ı≤ Liang shi pinghua Áv≠‹ Lianggong jiu jian ÁΩEœ Liangshanbo Ásy Lianhuan ji sÙp Lianhuan jian sÙœ liao F Liaoerwa d‡⁄ Lin Chong LR Lin Chong baojian ji LR_CO Lin Chong ye shang Liangshanbo LR ]WÁsy Lin’an {w Ling baodao F_M lingguai F« Linzi li zhide Wang Si de huishu, ba zai xian qian kan LlÿBo˝| ∫^—A‚b§e› Lisao ˜Ã Liu Bang Bπ Liu Bei B∆ Liu Biao san qiuji BÌTDp Liu Ji BÚ Liu Jingting hqF “Liu Qiqing shijiu Wanjianglou ji” hœÎ÷s±ø”O Liu Tang B Liu Wenzheng BÂF Liu Xibi Jinchai ji B∆≤˜¶O

Glossary

Liu Yiqing Bqy Liu Yong h√ Liu Zongyuan hv∏ Liushi jia xiaoshuo ªQap° Lü Buwei f£≥ Lu Da |F “Lü Dongbin feijian zhan Huang Long” f}´∏CŸ¿s Lu Guimeng ∞tX Lu Jiuyuan ∞EW Lu Junyi cTq Lü lan f˝ Lu Zhishen |º` luefanfa §«k lun ◊ Lun heng ◊≈ Lunyu ◊y Luo Cheng π® “Luoyang san guai ji” •ßT«O lüshi fl÷ Lüshi Chunqiu fÛKÓ lüyan sushuo [EU° Ma Lian ®G Ma Zhiyuan ®P∑ Maiwangguan flÊ] Mang Zhang Fei da’nao xiangfu yuan ıi∏jx¤≤| Mao shi zheng yi Ú÷øq Mao Zonggang Úv^ Masuda Mataru W–A mei C mei jieguo S≤G “Mei xing zhengchun” ˆˆßK meiren xiangcao ¸HªÛ Meixiang ˆª men Ã˛˘˛V Meng Huo sÚ Mengzhou s{ Mengzi sl “Mian” ¯ Mihun zhen gÓ}

251

Ming you bucheng, gong you bujiu WS £®A\S£N Mo Houguang ˆ·˙ moshen zhi bi ±´ßß moyangr “À‡ Mu Guiying p¤^ Mu Heng pÓ Mu Hong p∞ na ≥ nai @˛` Nai An @g Nanci xulu n¸‘˝ Nanjing Jiankangfu n?ÿd≤ nanqu n± nanxi n∏ “Nao Fanlou duoqing Zhou Shengxian” x‘”h°P”P nasi ∫r neifuben ∫≤ª Ni bian xiachang lai ti yihui AKUı” @^ Ni fuzi liangge jiang qu zuo panchan A˜l‚”NhµLÒ Ni jiang shao qu qu, zhi de yishao ANc h˙Auo@c Ni zendi hao cun AÁan¯ nihuaben ¿‹ª nin z Nog oltae ònhae —^jŒ— . . . pai shifen . . . P…¿ . . . pai shihou . . . P…‘ pai zuoci ∆y∏ Pak t’ongsa ònhae µq∆Œ— Pan Jinlian Ô˜¨ Panguo Shuoren LL”H “Panguo Shuoren Xixiang dingben de “jicheng’ he ‘chuangzao’ ” LL ”HmË[wªn∫ß~”®M ß–y® Pei Ruhai pp¸ peng ∑

252

Glossary

peng da wan ∑jJ pian wen c pingdian ˚I pinghua ≠‹˛˚‹ pinglin ben ˚Lª pingshu ˚— Pipa ji \]O “Piping Shuihu zhuan shuyu” Â˚ Ùq«zy podao ÎM po’nai r@ Qi ci zhazha buya, guaigui bujing ‰¸ Óæ£ÆA«fi£g Qi xiu lei gao C◊˛Z qi _ qia Í qiadai Í› qiahao Ín Qian Æ Qian qi zi eCl “Qian Sheren tishi Yanzi Lou” ˙ŸH D÷Pl” Qian Xiyan ˙∆• Qian Xuantong ˙»P Qianlong Æ© “Qiantang meng” ˙Ì⁄ qieshuo B° Qin Ming ≥˙ Qin Qiong ≥£ Qin Shihuang ≥l” qindongxin À?fl qing Î Qingfengzhen M∑Ì qingke –» “Qingmian Shou” C±~ Qinzong ‹v qiyan shi C•÷ qu ± qu ˙

qu bai xiangsheng ±’¤Õ Qu Yuan }Ï Qu zao ±¶ quanbin ˛´ que o quedai o› quehao on queshuo o° Qujiangchi ±ø¿ r‡ “Ren Xiaozi liexing wei shen” Ùµl P?∞´ “Rensheng shizi hutu shi” HÕ—r k\l Renzong Øv “Renzong renmu zhuan” Øv{¿« Rongyutang ePÛ Ruan Jin øi Ruan Yuan ø∏ ruhua J‹ ruowei Y∞ ruowei chenshuo Y∞Ø° sajia sa San chao bei meng huibian T¬_˘ |s san diao bajiaoshan T’›º∞ san fen T¿ san gu maolu TUTf san qi Zhou Yu TPÏ san rang Xuzhou T˝}{ “San xianshen Bao Longtu duanyuan” T{≠]sœ_fi sanda T¥ Sanguo zaju Tͯ@ Sanguo zhi TÍ” Sanguo zhi sushuo TÍ”U° Sanguo zhizhuan TÍ”« sanqu ≤± Sanxia wuyi TL≠q

Glossary

Sao ti ÃÈ sha Ÿ˛˛ Sha gou ji ˛ØO Sha gou quan fuxu ˛ØU“B Sha Xi ˛§ shangguzi |jl˛|©l shang jiang WN shangwuzi |al˛|a¤ “Shangyuan shiwu ye kandeng zhuan” W∏Q≠]›O« Shanhaiju Zhuren s¸~DH Shanzhao ΩL Shen bao ”¯ Shen Chucheng HÏ® Shen Defu Hw≈ Shen Guoyuan HÍ∏ “Shen Xiaoguan yi niao hai qi ming” Hpx@æ`CR Shen Yue H˘ shengchengang Õ∞ı sheng yuan Õ˚ shenwen qubi `屧 shenxian ´P shi ÷ shi h Shi En I¶ “Shi Hongzhao long hu junchen hui” v∞FsÍg⁄| Shi Hui If Shi Jin vi Shi Jingtang ¤qÌ Shi Nai’an di ben, Luo Guanzhong bian ci I@g∫ªAπe§s∏ “Shi Nai’an shiji zhi xin shangque” I@g∆Òßs”e “Shi Nai’an yu Zhang Shicheng” I@gPih¤ “Shi Nai’an zai Baiju Chang” I@g b’sı

253

Shi Tingzuo I?ı Shi Xiu ¤q Shi yan zhi ÷•” Shi Yanduan I¤› Shidetang @wÛ shifen Q¿ Shiji vO Shijing ÷g “Shilang fuma zhuan” ¤¶t®« Shiliju congshu hß~O— Shimo san xi Tang Sanzang r]T∏ T√ Shiquge ¤Î’ Shishi jiabu pu IÛa–Ø Shishuo xinyu @°sy Shitong san er xiaoshuo xing vŒ≤ ”p°≥ “Shitou Sun Li” ¤Y]fl “Shiwu guan xiyan cheng qiaohuo” Q≠e∏•®©◊ shizan ÷Ÿ shoupa ‚»˛‚¨ Shu er buzuo z”£@ shuai xi shuihu vËÙq shuan Ͳ¨ Shuangfengtang ˘pÛ Shuang jian Xiaoqing wenda ˘•pÎ ›™ shuazi Íl shuhui —| shuhui xiansheng —|˝Õ Shuihu zaju Ùq¯@ Shuihu zhuan cihua Ùq«¸‹ Shuiyue Guanyin ÙÎ[µ “Shunzi zhixiao bianwen” œl‹µ‹Â shuo ° “Shuo nan” °¯ Shuo qilai, zhuang nide huangzi °_”AÀA∫El

254

Glossary

Shuo qilai, zhuang niniang de huangzi °_”AÀAQ∫¿l Shuo Tang ° shuo yan wei jue °•º¥ shuo yan wei liao °•ºF shuo you wei jue °Sº¥ shuo you wei liao °SºF˛°—ºF Shuo Yue quanzhuan °®˛« shuobai °’ shuochang cihua °¤¸‹ shuohua °‹ shuohuade °‹∫ shuohuaren °‹H shuoshu °— shuoshude °—∫ si da qishu |j_— “Si sheng jiao Fan Zhang ji shu” ∫Õ Êdi˚¡ Sima Qian q®E Sima Zhao fuduo shoushantai q®L _‹¸Ix siwufen |≠¿ siyan shi |•÷ Song Gongming nao yuanxiao zaju ∫Ω˙x∏d¯@ song | Song Jiang ∫ø “Song Sigong da’nao ‘Jinhun’ Zhang” ∫|ΩjxTÓi songgu |j “Su Changgong Zhang Tailiu zhuan” ¨¯Ωπ?h« Su Dongpo ¨FY sui E Sui shi yiwen ¶vÚ Sui Tang ¶ Sun Bin ]¡ Sun Erniang ]GQ Sun Quan ]v

Sun Wukong ]©≈ Sun Xin ]s Sun Yan ]¢ Suo Chao ¡W “Suxiang Ting Zhang Hao yu Yingying” JªFiEJ aa Tai’anzhou ıw{ Taihangshan ”Ês Taiyi huntianxiang zhen ”AV—H} Taiyi sancai zhen ”AT~} Taizong ”v Taizu ”™ Tan Yu ZJ tanci u¸ Tang Sanzang xitian qujing T√Ë— ˙g Tang Xiaoshan ps Tao Yuanjing ÁΩ∫ taozhen ≥u “Tian Dao” —D Tiandu Waichen —£~⁄ tiangang xing —\P Tianshu —— tianxia diyi guodu —Uƒ@Í£ Tianxia zhi wen xin shao er li er duo —U ßÂfl÷”Ω’h Tianyige —@’ Tiao Shuang jian ’˘• timu Dÿ Ting taozhen jinshi cunren •≥u…O ¯H tongbao P]˛PM Tong jian changbian jishi benmo q≥¯s ˆ∆ªΩ Tong jian gangmu q≥ıÿ tu fl “Wan Xiuniang chou bao Shantingr” UqQ≥¯sF‡

Glossary

Wang Bi ˝] Wang Chong ˝R Wang Daokun LD¯ Wang Daosheng ˝DÕ Wang Lun ˝¤ Wang Mang ˝ı Wang Qi ˝¶ Wang Shaotang ˝÷Û Wang Shizhen ˝@s Wang Tao ˝¸ Wang Xiong ˝Ø Wang Ye ˝Á Wang Zhaojun ˝Lg “Wang Zhaojun bianwen” ˝Lg‹Â wangnian jiao —~Ê Wanli U˙ Wanli yehuo bian U˙•Ús washe ÀŸ wazi Àl weizhi . . ., qieting xiahui fenjie ºæKA B•U^¿— Weizhou Ù{ wen  wenhua chengdu Â∆{◊ Wenhui bao Â◊¯ “Wenjing yuanyang hui” FVpm| wenren hua ÂHe wenren xiaoshuo ÂHp° wenxue « wenyan • wenzhang Âπ wo ⁄ wo laoye ⁄—› Wo que bai shen niao ⁄oÙ∆æ Womende benshi you bushi buru bieren ⁄Ã∫ª∆S£O£pOH wu © wu a Wu Changling d˜÷

255

Wu Congxian dq˝ Wu Song ZQ “Wu Xingzhe” ZÊà Wu Yong dŒ Wudai Liang shi pinghua ≠NÁv≠‹ wudi a∫ Wuding Zw “Wujie Chanshi si Honglian ji” ≠Ÿ Ivpı¨O Wulin ZL wuna a∫ wushi aO Wutaishan ≠xs Wuxi L¸ wuyan shi ≠•÷ wuzi a¤ wuzuguan L¨[ “Xi ci” t„ Xi xia ∏ Xiang! Ji xiang! ≥I•≥I Xiang Yu µ– “Xianggong ershiwu nian” ∏Ω GQ≠~ Xianqian zhiyou wufen fannao . . . gen tianshang shifen fannao ˝eu≥ ≠¿–oKÛKWQ¿–o xianxiege I«” xianyu ~y Xiao Dexiang Ωw‘ “Xiao furen jinqian zeng nianshao” p“H˜˙ÿ~÷ xiao jiaor p}‡ Xiaochuang ziji p°¤l Xiaojing µg Xiaojing zhijie µgΩ— xiaoke pi xiaoshuo p° xiezi §l “Xihu san ta ji” ËÚTO

256

Glossary

Ximen Qing ˢy Xin Qiji ØÛe Xin’an sw Xinbian Liu Zhiyuan huanxiang baitu ji ssBæ∑Ÿm’flO Xinghua ≥∆ xingshengzi Œnr Xinkan quanxiang pinghua Qian-Han shu xuji sZ˛¤≠‹e~— Ú∞ Xinkan quanxiang pinghua Wuwang fa Zhou shu sZ˛¤≠‹Z˝ÔÙ— Xinkan quanxiang pinghua Yue Yi tu Qi qi guo Chunqiu houji sZ˛¤≠‹ ÷›œÙCÍKÓ·∞ Xinkan quanxiang Qin bing liu guo pinghua sZ˛¤≥√ªÍ≠‹ “Xinqiaoshi Han Wu mai chunqing” sÙ´˙≠ÊK° Xiong Longfeng µsp Xiongfei guan Ø∏] Xiongnu I£ Xiongzhong you mei wenxue ›§SS « xiu  Xiu! Xiu! Xiu! III xiwen ∏ Xixiang dingben Ë[wª Xiyou zaju ˯@ Xu Fenpeng }ƒP Xu Heng \≈ Xu Mengxin }⁄Ò Xu wenxian tongkao ÚÂmq“ Xu Zhongyou }Ú— Xu Zichang 礘 Xuande ≈w Xue Ba £Q Xue Chao £W Xue Dingshan £Bs Xue Dingshan zhengxi £Bs∫Ë

xue qu zhiyu dhPy Xunzi ˚l Yan Da Pj “Yan Pingzhong er tao sha san shi” À≠ÚGÁ˛Th Yan Poxi FC§ Yan Qing PC Yan Qing bo yu PC’Ω Yan Shun P∂ yan you wei liao •SºF yan zhi wen •ß yanei ≈∫ yanfen ⁄ª “Yang Jiaoai sizhan Jing Ke” œ§s ∫‘_ Yang Jingxian ®∫ Yang Minglang ®˙„ “Yang Siwen Yanshan feng guren” ®‰≈Ps{GH “Yang Wen ‘Lanluhu’ zhuan” ®≈ dÙÍ« Yang Xianzhi ®„ß Yang Xiong ®Ø Yang Ye ®~ Yang Zhi ®” Yang Zhongliang ®Ú} Yang Zongbao ®v_ Yangguxian ߶§ Yang jia jiang ®aN Yangzhou pinghua ≠{˚‹ Yanhua meng œ·⁄ Yanjingshi ji „g«∞ yanyi tq yaoshu ØN ye ] Ye Zhou ≠fi yeshi •v Yeshiyuan ]OÈ yi y yi N

Glossary

“Yi ku gui lai daoren chu guai” @]≠ ÓDH£« Yi! zhengshi xIøO yida @¥ yifen @¿ yijin huanxiang ÁAŸm yijiu @N Yin Chang ®` “Ying Ning” ¶Á Yinglie zhuan ^P« Yingxiong pu ^Ø– “Yingying zhuan” aa« yinshou fi∫ Yinshu ±— yinyuan shengfa ]tÕk “Yinzhi jishan” ±cnΩ yiqi q “Yiwenzhi” ¿Â” Yizhang Qing @VC Yongjia shuhui √≈—| Yongle dadian √÷j You jia nan ben, you guo nan tou ≥a ¯bA≥ͯΠyou jieguo ≥≤G youcai ≥~ youfenjiao ≥¿–˛≥¿Ê yu y Yu Xiangdou EHÊ “Yu Zhongju tishi yu shanghuang” \Ú|D÷JW” Yuan bao yuan Zhaoshi guer fi¯fiØÛ t‡ “Yuan Dao” ÏD Yuan Hongdao KªD Yuan Wuya KLP yuanben |ª Yuandai xiaoshuo ∏Np° yuandu zhushu Ër¤— yuanyang tui pmL Yuanzhen shuhui ∏s—|

257

Yue Fei ®∏ Yue Fei zhuan ®∏« yuefu ÷≤ “Yueming heshang du Liu Cui” Î˙ M|◊hA yuequ ÷± yuhuan bu …ÙB yulu y˝ Yunchengxian P∞§ yutiwen yÈ Yuxijian s∏? Yuzhangcheng ren yue liang tuanyuan ›π∞H΂ŒÍ “Yuzhangcheng Shuangjian gan Su Qing” ›π∞˘•∞¨Î “Za shuo” ¯° zaban ¯Í Zaishengsi liyuan zhili b”x˘Èmfl zaju ¯@ zan •˛ı zao º˛Í “Zaojiaolin dawang jiaxing” m§L j˝≤Œ zawang ¯Ù zege h”˛hñ zengcha ben W°ª Zengtoushi øY´ “Zengxiang Qiantang meng” W¤ ˙Ì⁄ zewei h∞ “Zhachuan Xiao Chen bian Bawang” ≠tΩ`SQ˝ “Zhang Sheng cailuandeng zhuan” iÕm}O« “Zhang Zifang mudao ji” il– }DO Zhang Cen i¬ Zhang Duanyi i›q “Zhang Gulao zhong gua qu Wennü” ij—ÿ?˘Âk

258

Glossary

Zhang Heng iÓ Zhang Junrui igÁ Zhang Qian id Zhang Qing iC˛iM Zhang Shangde i|w Zhang Shicheng ih¤ “Zhang Shunmei dengxiao de linü” iœ¸OdoRk “Zhang Shuye zhuan” i˚]« “Zhang Wengui zhuan” iÂQ« Zhang Zai iA Zhang Zhupo iÀY Zhang Zugang i™˚ zhangfa πk zhanghui xiaoshuo π^p° Zhangyi shucai Mq®] Zhangzhe qu dao duyu fashi ¯Ã˙M ◊Pkv Zhangzong πv Zhao Bosheng chasi yu Renzong” ØB…˘vJØv Zhao Delin ØwÔ Zhao Hong’en Øx¶ Zhao Kuangyin ØJN zhe à zhe zhengshi oøO Zhedanr Wu Song da hu È·‡ZQ ¥Í zhen ” zhen e zhencai shixue u~Í« Zheng Dehui Gw˜ “Zheng Jieshi ligong shenbi gong” G`œfl\´u} zhenge u” zhengfanfa ø«k zhengse yueren ø‚÷H zhengshi øv zhengshi øO zhengxiege ß«”

zheng yao øn zhengzhi ø»˛øΩ zhenren uH zhi ß zhiguai ”« zhijiao Ω–˛u– Zhikong zhuren ciqu wu jieguo er u£ DHπhL≤G’ Zhiqi hesuo jinxi? ‹≠Û“⁄§ Zhishuo dage shi xiangxiaren u°jÙ OmUH Zhishuo Daxue yaolue Ω°j«n§ zhiwei u∞˛≠∞˛Ó∞ Zhizheng ‹ø Zhizhi ‹v Zhizhi xinkan quanxiang pinghua Sanguo zhi ‹vsZ˛¤≠‹ TÍ” Zhongren zhide ba shiban yiqi kangqi ≥Huo‚¤O@Ù™_ “Zhong yi Shuihu quanzhuan fafan” æq Ùq˛«oZ Zhong yi Shuihu zhuan æqÙq« “Zhong yi Shuihu zhuan xu” æq Ùq«‘ Zhong yong zhijie §eΩ— Zhou Bangyan Pπ¤ Zhou Lianggong PGu Zhou Yu PÏ Zhu Bajie fiKŸ Zhu Guiying ∂¤^ Zhu Shikai ∂hÕ Zhu Wen ∂≈ Zhu Xi ∂Q Zhu Youdun ∂≥W Zhu Yuanzhang ∂∏˝ zhuan ‡ zhuang nide wangzi ÀA∫Êl Zhuangzi ¯l Zhuangzi jijie ¯l∞—

Glossary

zhuban pindiao —Î~’ zhufu ¨I˛ÒJ zhugongdiao —c’ Zhujia zhuang ¨a¯ zi r ziban cihua ¤h¸‹ zifa rk zimei lf˛nf ziwei l∞

259

Zizhi tong jian Ívq≥ zou µ zuannong °À Zuo Qiuming ™C˙ Zuo zhuan zhushu ™«`® Zuo zhuan ™« zuodi §a zuoshi jishi, youshi jiyan ™vO∆Akv O•

Selected Bibliography Premodern Sources in Chinese Chanzong yulu jiyao Ivy˝Ën. Ed. Shanghai Guji Chubanshe. Shanghai: Shanghai Guji Chubanshe, 1992. Chu Renhuo uHÚ. Sui Tang yanyi ¶tq. Ed. Chen Shuliang Ø—}. Guiyang: Guizhou Renmin Chubanshe, 1993. Da Tang Sanzang qujing shihua jT√˙g÷‹. In Liu Jian and Jiang Shaoyu, eds., Jindai Hanyu yufa ziliao huibian (Songdai juan), 235–259. Da Yuan shengzheng guochao dianzhang j∏tFͬÂπ. Reprint. 2 vols. Taipei: Wenhai Chubanshe, 1964. Dong Jieyuan ≥—∏. Tang Xianzu ping Xixiang ji zhugongdiao ˆ„™˚Ë[O —c’. Lanzhou: Gansu Renmin Chubanshe, 1982. ———. Ming Jiajing ben Dong Jieyuan Xixiang ji ˙≈tª≥—∏Ë[O. Reprint. 2 vols. Shanghai: Zhonghua Shuju, 1963. Dong jing menghua lu wai sizhong F?⁄ÿ˝~|ÿ. Shanghai: Gudian Wenxue Chubanshe, 1985. Du Shanfu ˘Ω“. “Zhuangjia bushi goulan” ¯a£—ƒÊ. In Zhongguo wenxue xinshang quanji §Í«Y‡˛∞. 46 vols. Taipei: Zhuangyan Chubanshe, 1981–1984. 25: 8–12. Dunhuang bianwen ji ∞◊‹Â∞. Eds. Wang Zhongmin ˝´¡ et al. Beijing: Renmin Wenxue Chubanshe, 1956. Feng Menglong æ⁄s. Jingshi tong yan µ@q•. 3 vols. Photoreprint ed. in Guben xiaoshuo jicheng jªp°∞®. ———. Xingshi heng yan Ù@Ì•. 2 vols. Beijing: Renmin Wenxue Chubanshe, 1987. ———. Yushi ming yan Î@˙•. Shanghai: Shanghai Guji Chubanshe, 1992. Gao Ru ™ß. Baichuan shuzhi ?t—”. Shanghai: Gudian Wenxue Chubanshe, 1957. Gong Shengyu «t÷. “Song Jiang sanshiliu zan” ∫øTQªŸ. In Zhou Mi PK, Guixin zashi xuji —د—Ú∞. Congshu jicheng O—∞® ed., 276–287. Guben xiaoshuo jicheng jªp°∞®. Ed. Guben xiaoshuo jicheng bianji weihuanhui j ªp°∞®sËe˚|. 1–5 series. Shanghai: Shanghai Guji Chubanshe, 1990–.

261

262

Selected Bibliography

Guben xiqu congkan jª∏±OZ. Ed. Guben xiqu congkan biankan weiyuanhui jª ∏±OZsZe˚|. Shanghai: Shangwu Yinshuguan, 1954. Gudai hanyu jN~y. Eds. Guo Xiliang ¢¸}et al. 3 vols. Tianjin: Tianjin Jiaoyu, 1979. Hong Pian x·. Qingping shantang huaben M≠sÛ‹ª. Shanghai: Shanghai Guji Chubanshe, 1992. Hu Yinglin J≥Ô. Shaoshi shanfang bicong ÷«s–ßO. Beijing: Zhonghua Shuju, 1959. Huanmen zidi cuo lishen ∆˘lÃ˘fl≠. In Xiao Sun tu deng sanzhong p]O•Tÿ. Guben xiqu congkan. Jia Zhongming ÎÚ˙. Lugui bu xubian ˝≠ØÚs. In Zhongguo gudian xiqu lunzhu jicheng, 2: 277–300. Jiaoding Yuankan zaju sanshi zhong ’q∏Z¯@TQÿ. Ed. Zheng Qian G?. Taipei: Shijie Shuju, 1962. Jin Shengtan ˜tƒ. “Du Diwu caizishu fa” ™ƒ≠~l—k. In Shuihu zhuan huiping ben, 1: 15–22. ———. “Shuihu zhuan xu yi” Ùq««@. In Shuihu zhuan huiping ben, 1: 1–6. ———. “Shuihu zhuan xu er” Ùq««G. In Shuihu zhuan huiping ben, 1: 6–8. ———. “Shuihu zhuan xu san” Ùq««T. In Shuihu zhuan huiping ben, 1: 8–11. Kong Yingda ’oF, annotator, Zhouyi zheng yi Pˆøq. Sibu beiyao |°∆n ed. Lanling Xiaoxiaosheng ıÆ∫∫Õ (pseudonym). Jin ping mei cihua ˜~ˆ¸‹. Taipei: Zengnizhi Wenhua Shiye Youxian Gongsi, 1980. Li Dou ıÊ. Yangzhou huafang lu ≠{eÈ˝. Beijing: Zhonghua Shuju, 1960. Li Kaixian ı}˝. Ci xue ¸¶. In Zhongguo gudian xiqu lunzhu jicheng, 3: 257–418. Li Ruzhen ıº√. Jing hua yuan Ë·t. Beijing: Renmin Wenxue Chubanshe, 1955. Li Yu ıÆ. Xianqing ouji ~°∏H. In Ming Qing zaji ˙M¯O. Shanghai: Shanghai Wenyi Chubanshe, 1992. Li Zhi ıÙ. Fen shu, Xu Fen shu I—AÚI—. Beijing: Zhonghua Shuju, 1975. Li Zhuowu piping Zhong yi Shuihu zhuan ıÙ^Â˚æqÙq«. 5 vols. Photoreprint ed. in Guben xiaoshuo jicheng. Ling Mengchu ‚¤Ï. Tanqu zazha ”±¯æ. In Zhongguo gudian xiqu lunzhu jicheng, 4: 253–261. ———. Erke Pai’an jingqi GËÁ◊Â_. Shanghai: Shanghai Guji Chubanshe, 1992. ———. Pai’an jingqi Á◊Â_. Shanghai: Shanghai Guji Chubanshe, 1992. Liu Xie BE. Wenxin diaolong zhu ÂflJs`. Annotator Fan Wenlan dÂi. 2 vols. Hong Kong: Shangwu Yinshuguan, 1986. Liu Zhiyuan zhugongdiao Bæ∑—c’. In Zhugongdiao liangzhong —c’‚ÿ. Eds. Ling Jingshan ‚∫∫ and Xie Boyang ¬Bß. Ji’nan: Qi Lu Shushe, 1988.

Selected Bibliography

263

Lu Ji ∞˜. Wen fu yizhu ·∂`. Annotator Zhang Huaijin ih‘. Beijing: Beijing Chubanshe, 1984. Luo Guanzhong πe§. Sanguo yanyi TÍtq. Beijing: Renmin Wenxue Chubanshe, 1972. Luo Ye πM. Xinbian Zuiweng tanlu ssKŒÕ˝. Tokyo: Bun kyû dò, 1941. Meng Yuanlao s∏—. Dong jing menghua lu F?⁄ÿ˝. In Dong jing menghua lu wai sizhong, 1–88. Ming Chenghua shuochang cihua congkan ˙®∆°¤¸‹OZ. Ed. Yang Jialuo ≠ad. Taipei: Dingwen, 1979. Naide Weng @oŒ (pseudonym). Ducheng jisheng £∞ˆ”. In Dong jing menghua lu wai sizhong, 89–110. Nogoltae ònhae —^jŒ—. In Nogoltae, Pak t’ongsa ònhae —^jµq∆Œ—. Photoreprint ed. Seoul, 1973. Pak t’ongsa ònhae µq∆Œ—. In Nog oltae, Pak t’ongsa ònhae. Pu Songling ZQ÷. Liaozhai zhiyi ·Ù”ß. Hangzhou: Zhejiang Guji Chubanshe, 1993. Qian Cai ˙m. Shuo Yue quanzhuan °®˛«. 4 vols. Photoreprint ed. in Guben xiaoshuo jicheng. Qiguo Chunqiu pinghua houji CÍKÓ≠‹·∞. In Song Yuan pinghua ji, 483–560. Qin bing liuguo pinghua ≥√ªÍ≠‹. In Song Yuan pinghua ji, 561–664. Rulin waishi huijiao huiping ben ßL~v|’|˚ª. Ed. Li Hanqiu ı~Ó. Shanghai: Shanghai Guji Chubanshe, 1984. San Sui ping yao zhuan TE≠Ø«. 20 hui. Photoreprint ed. in Guben xiaoshuo jicheng. Sanguozhi pinghua TÍ”≠‹. In Song Yuan pinghua ji, 739–882. Shemotashi qucong ¯Ø¶«±O. Ed. Wu Mei dˆ. Series 1–2. Shanghai: Hanfenlou, 1927. Shi Nai’an I@g and Luo Guanzhong πe§. Rong yutang ben Shuihu zhuan ePÛªÙq«. 2 vols. Shanghai: Shanghai Guji Chubanshe, 1988. ———. Shuihu quanzhuan Ùq˛«. Shanghai: Shanghai Renmin Chubanshe, 1975. Shuihu xiqu ji Ùq∏±∞. Eds. Fu Xihua «§ÿ and Du Yingtao ˘o≥. 2 vols. Shanghai: Gudian Wenxue Chubanshe, 1957. Shuihu zhizhuan pinglin Ùq”«˚L. 3 vols. Photoreprint ed. in Guben xiaoshuo jicheng. Shuihu zhuan huiping ben Ùq«|˚ª Eds. Chen Xizhong Øf¡ et al. 2 vols. Beijing: Beijing Daxue Chubanshe, 1981. Sima Qian q®E. “Bao Ren An shu” ¯Ùw—. In Sima Qian q®E. Hong Kong: Shanghai Shuju, 1975. 282–296. Song Yuan difang zhi sanshiqi zhong: Xianchun Lin’an zhi ∫∏aË”TQCÿRwE {w”. Taipei: Guotai Wenhua Shiye Youxian Gongsi, n. d. Song Yuan pinghua ji ∫∏≠‹∞. Ed. Ding Xigen B¸⁄. 2 vols. Shanghai: Shanghai Guji Chubanshe, 1990.

264

Selected Bibliography

Song Yuan si da xiwen duben ∫∏|j∏™ª. Ed. Yu Weimin \∞¡. Nanjing: Jiangsu Guji Chubanshe, 1988. Tao Zongyi ≥vˆ. Chuogeng lu ˘—˝. Shanghai: Fuying Shuju, 1885. Tian Rucheng –º®. Xihu youlan zhiyu ËÚ½”l. Taipei: Shijie Shuju, 1963. Tuo Tuo ÊÊ et al. Song shi ∫v. 40 vols. Beijing: Zhonghua Shuju, 1977. Wang Bocheng ˝B®. Tianbao yishi zhugongdiao —_Ú∆—c’. In Zhugongdiao liangzhong. Eds. Ling Jingshan and Xie Boyang. Ji’nan: Qi Lu Shushe, 1988. Wang Jide ˝kw. Qu lü ±fl. In Zhongguo gudian xiqu lunzhu jicheng, 4: 43–191. Wang Shifu ˝Íj. Xinkan qimiao quanxiang zhushi Xixiang ji sZ_Æ˛¤`¿ Ë[O. 2 vols. In Guben xiqu congkan. ———. Xixiang ji Ë[O. Ed. Wang Jisi ˝u‰. 2nd ed. Shanghai: Shanghai Guji Chubanshe, 1978. Wang Zhuo ˝`. Biji manzhi —˚©”. In Zhongguo gudian xiqu lunzhu jicheng, 1: 91–152. Wu Cheng’en d”¶. Xiyou ji (Shidetang ben) ËCO]@wÛª^. Photoreprint ed. in Guben xiaoshuo jicheng. ———. Xiyou ji ËCO. Beijing: Renmin Wenxue Chubanshe, 1973. Wu Jingzi dqÍ. Rulin waishi ßL~v. Beijing: Renmin Wenxue Chubanshe, 1978 Wu Zimu d¤™. Meng liang lu ⁄Á˝. In Dong jing menghua lu wai sizhong, 129–328. Wudaishi pinghua ≠Nv≠‹. In Song Yuan pinghua ji, 15–251. Wuwang fa Zhou pinghua Z˝ÔÙ≠‹. In Song Yuan pinghua ji, 399–482. Xia Tingzhi Lx¤. Qinglou ji C”∞. In Zhongguo gudian xiqu lunzhu jicheng, 2: 17–84. Xiao Sun tu p]O. In Xiao Sun tu deng sanzhong p]O•Tÿ. Guben xiqu congkan. Xihu Laoren ËÚ—H (pseudonym). Xihu Laoren Fansheng lu ËÚ—Hc”˝. In Dong jing menghua lu wai sizhong, 111–128. Xu Shen \V. Shuowen jiezi °Â—r. Sibu beiyao ed. Xuande xieben Jinchai ji ≈wgª˜¶O. Guangzhou: Guangdong Renmin Chubanshe, 1985. Xuanhe yishi ≈MÚ∆. In Xuanhe yishi, Zengcha Tian Hu Wang Qing Zhong yi Shuihu quanzhuan ≈MÚ∆AW°–Í˝yæqÙq˛«. Photoreprint ed. in Guben xiaoshuo jicheng. Yongle dadian ben Xue Rengui zheng Liao shilue √÷jª£ØQ∫Ò∆§. In Ming Chenghua shuochang cihua congkan, 209–288. Yuan Zhen ∏“.“Huizhen ji” |uO. In Xixiang ji Ë[O. 2nd ed. Shanghai: Shanghai Guji Chubanshe, 1978. 199–204. Yuandai baihua bei ∏N’‹O. Ed. Feng Chengjun æ”v. Shanghai: Shangwu Chubanshe, 1931.

Selected Bibliography

265

Zang Jinshu N?˚ (Maoxun ∑`), ed. Yuanqu xuan ∏±Ô. 4 vols. Beijing: Zhonghua Shuju, 1958. Zeng Zuyin ø™Æ et al., eds. Zhongguo lidai xiaoshuo xuba xuanzhu §Í˙Np° «[Ô`. Wuhan: Changjiang Wenyi Chubanshe, 1982. Zengcha Tian Hu Wang Qing Zhong yi Shuihu quanzhuan W°–Í˝yæqÙq˛«. In Xuanhe yishi, Zengcha Tian Hu Wang Qing Zhong yi Shuihu quanzhuan ≈MÚ∆AW°–Í˝yæqÙq˛«. Photoreprint ed. in Guben xiaoshuo jicheng. Zhang Dai iß. Tao’an meng yi ≥g⁄–. Shanghai: Shanghai Yuandong Chubanshe, 1995. Zhang Tingyu i?… et al. Ming shi ˙v. 7 vols. Taipei: Guofang Yanjiuyuan, 1962. Zhang Xie zhuang yuan iÛ¨∏. In Xiao Sun tu deng sanzhong p]O•Tÿ. Guben xiqu congkan. Zhao Lingzhi ØOœ. “Cui Yingying shangdiao dielianhua guzici” Zaa”’∫?· ™l¸. In Xixiang ji Ë[O. 2nd ed. Shanghai: Shanghai Guji Chubanshe, 1978. 205–208. Zhong Sicheng ¡‡®. Lu gui bu ˝≠Ø. In Zhongguo gudian xiqu lunzhu jicheng, 2: 85–274. Zhongguo gudian xiqu lunzhu jicheng §ÍjÂ∏±◊¤∞®. Ed. Zhongguo Xiqu Yanjiuyuan. 10 vols. Beijing: Zhongguo Xiju Chubanshe, 1959–1960. Zhou Deqing PwM. Zhongyuan yinyun §Ïµ˝. In Zhongguo gudian xiqu lunzhu jicheng, 1: 169–285. Zhou Lianggong PGu (1612–1672). Yinshuwu shuying L—Œ—v. Shanghai, 1957. Zhou Mi PK. Wulin jiushi ZL¬∆. In Dong jing menghua lu wai sizhong, 329–526. Zhu Quan ∂v. Taihe zheng yinpu ”Møµ–. In Zhongguo gudian xiqu lunzhu jicheng, 3: 1–231. Zhu Shenglin —tF. Da Tang Qinwang cihua j≥˝¸‹. Beijing: Wenxue Guji Kanxingshe, 1955. Zhuzi jicheng —l∞®. 8 Vols. Beijing: Zhonghua Shuju, 1954. Zhuzi yulei ∂ly˛. Ed. Wang Yunwu ˝≥≠. Taipei: Taiwan Shangwu Yinshuguan, 1968.

Modern Sources in Chinese and Japanese Cao Guangshun ‰s∂. Jindai Hanyu zhuci ÒN~yU¸. Beijing: Yuwen Chubanshe, 1995. ———. “Shishuo ‘jiu’ he ‘kuai’ zai Songdai de shiyong ji youguan de duandai wenti” ’°ßN®Mß÷®b∫N∫œŒŒ_N›D. Zhongguo yuwen §Íy 1987. 4: 288–294. Chen Liao ØÒ. “Qu wei cun zhen, Shi Nai’an zhi mi ke jie” h∞suAI@gߺ i—. Nantong shizhuan xuebao nqvM«¯. 1992. 1: 8–13.

266

Selected Bibliography

Chen Liming Ø˙˙. “Liu Xibi Jinchai ji de faxian jiqi gaimao” mB∆≤˜¶On∫ o{Œ‰ß™. In Xuande xieben Jinchai ji ≈wgª˜¶O. 141–145. Chen Maotong ØZP. Zhongguo lidai xuanguan zhidu §Í˙NÔxÓ◊. Shanghai: Huadong Shifan Daxue Chubanshe, 1994. Chen Ruheng غ≈. Shuoshu shihua °—v‹. Beijing: Zuojia Chubanshe, 1958. ———. Songdai shuoshu shi ∫N°—v. Shanghai: Shanghai Wenyi Chubanshe, 1979. Deng Deyou Hwˆ. “Tan Xuanhe yishi yu ‘Song Jiang sanshiliu zan’ de shidai xianhou wenti” Õm≈MÚ∆nPm∫øTQªŸn∫…N˝·›D. Ming Qing xiaoshuo yanjiu ˙Mp°„s 1986. 3: 415–420. Deng Siyu H‡Í. Zhongguo kaoshi zhidu shi §Í“’Ó◊v. Taipei: Taiwan Xuesheng Shuju, 1967. Du Jiwen ˘~ and Wei Daoru QDß. Zhongguo Chanzong tongshi §ÍIvqv. Nanjing: Jiangsu Guji Chubanshe, 1993. Fang Zhengyao ËøÛ. Zhongguo xiaoshuo piping shilue §Íp°Â˚v§. Beijing: Zhongguo Shehui Kexue Chubanshe, 1990. Fu Xihua ≈§ÿ. Yuandai zaju quanmu ∏N¯@˛ÿ. Beijing: Zuojia Chubanshe, 1957. Gu Tinglong U?s and Shen Jin Hz. “Guanyu xin faxian de Jingben Zhong yi zhuan canye” ˆ_so{∫m?ªæq«n›∂. In Shuihu pinglun ji Ùq ˚◊∞. Shanghai: Shanghai Renmin Chubanshe, 1976. 105–106. Gu Xuejie U«e. Yuan-Ming zaju ∏˙¯@. Shanghai: Shanghai Guji Chubanshe, 1979. Reprint. Taipei: Guowen Tiandi, 1980. Guo Shaoyu ¢–∑. Zhongguo wenxue piping shi §Í«Â˚v. Shanghai: Shanghai Guji Chubanshe, 1979. Guo Zhenyi ¢e@. Zhongguo xiaoshuo shi §Íp°v. Hong Kong: Taixing Shuju, 1961. Han Rulin ˙ßL. Yuanchao shi ∏¬v. Beijing: Renmin Chubanshe, 1986. Han Sheng Hn. “Xixiang ji gujin banben mulu jiyao” Ë[Ojµ©ªÿ˝Ën. In his Xixiang ji xinlun. 161–205. ———, ed. Xixiang ji xinlun Ë[Os◊. Beijing: Zhongguo Xiju Chubanshe, 1992. He Manzi Û°l. Lun Jin Shengtan pigai Shuihu zhuan ◊˜tƒÂÔÙq«. Shanghai, 1954. He Xin Ûs. Shuihu yanjiu Ùq„s. Shanghai, 1985. Hou Shaowen ‘–Â. Tang-Song kaoshi zhidu shi ∫“’Ó◊v. Taipei, 1973. Hou Zhongyi ‘æq and Wang Rumei ˝ºˆ, eds. Jin ping mei ziliao huibian ˜~ˆÍ∆◊s. Beijing: Beijing Daxue Chubanshe, 1985 Hu Shi JA. Hu Shi gudian wenxue yanjiu lunji JAj«„s◊∞. Shanghai: Shanghai Guji Chubanshe, 1988. ———. Zhongguo zhanghui xiaoshuo kaozheng §Íπ^p°““. Taipei: Yunfeng Shuju, 1976.

Selected Bibliography

267

Hu Shiying Jh˜. Huaben xiaoshuo gailun ‹ªp°ß◊. 2 vols. Beijing: Zhonghua Shuju, 1980. Hu Zhu’an JÀw. “Song Yuan baihua zuopinzhong de yuqi zhuci” ∫∏’‹@~§ ∫yU¸. Zhongguo yuwen §Íy 1958. 6: 270–274. Huang Lin ¿M. “Ming-mo Yuan-chu ren Shi Nai’an ji ‘Shi Nai’an diben’ ” ˙Ω ∏ÏHI@gŒ©I@g∫ª™. Fudan xuebao _π«¯ 1982.5. ———. “Yizhong zhide zhumu de Shuihu guben” @ÿ»o`ÿ∫mÙqnjª. Fudan xuebao _π«¯ 1980.4: 86–89. Huang Lin and Han Tongwen ˙PÂ, eds. Zhongguo lidai xiaoshuo lunzhu xuan §Í˙Np°◊¤Ô. 2 vols. Nanchang: Jiangxi Renmin Chubanshe, 1982–1985. Huang Ticheng ¿‘®. “Shi Nai’an jiguan kaobian” I@gye“Î. Yangzhou shiyuan xuebao ≠{v|«¯ 1992. 1: 135–139. Huang Weiliang ¿˚Á Zhongguo wenxue zongheng tan §Í«aÓÕ. Taipei: Dongda Tushu Gongsi, 1988. Jiang Xingyu ±P‘. Xixiang ji kaozheng Ë[O““. Shanghai: Shanghai Guji Chubanshe, 1988. Kong Lingjing ’O“. Zhongguo xiaoshuo shiliao §Íp°v∆. Beijing: Zhonghua Shuju, 1957. Kòsaka Jun’ichi ª¡∂@. Baihua yuhui yanjiu ’‹y◊„s. Trans. Jiang Lansheng ø≈Õ and Bai Weiguo ’˚Í. Beijing: Zhonghua Shuju, 1997. ———. Shuihu cihui yanjiu (xuci bufen) Ùq¸◊„s]͸°¿^. Trans. Ueda Hitoshi ”–°. Beijing: Wenjin Chubanshe, 1992. Li Weishi ı∂Í. “Shuihu zhuan chengshu yu Yuan-mo Ming-chu zhi shuo buneng chengli” mÙq«n®—_∏Ω˙Ïß°£‡®fl. Shehui kexue zhanxian ¿|Ï«‘u. 1993. 6: 224–227. ———. “Cong Du Jin de Shuihu renwu quantu kan Shuihu zhuan de chengshu niandai” q˘ª∫mÙqH´˛œn›mÙq«n∫®—~N. Shehui kexue zhanxian ¿|Ï«‘u 1991. 3: 274–276. Lin Geng L∞. “Cong Shuihu xi kan Shuihu zhuan” qÙq∏›mÙq«n. In Guoxue yanjiu (di yi juan) Í«„s]ƒ@˜^. Ed. Beijing Daxue Zhongguo Chuantong Wenhua Yanjiu Zhongxin _?j«§Í«ŒÂ∆„s §fl. Beijing: Beijing Daxue Chubanshe, 1993. 209–233. Liu Cunren hsØ (Liu Ts’un-yan). “Luo Guanzhong jiangshi xiaoshuo zhi zhenwei xingzhi” πe§øvp°ßu∞?Ë. In Zhongguo gudai xiaoshuo yanjiu: Taiwan Xianggang lunwen xuanji §ÍjNp°„sROWª‰◊ ÔË. Ed. Liu Shide B@w. Shanghai: Shanghai Guji Chubanshe, 1983. 74–172. Liu Dajie Bj«. Zhongguo wenxue fada shi §Í«oFv. Beijing: Zhonghua Shuju, 1967. Liu Dong BV and Huang Qingjiang ¿Mø “Shi Nai’an yu Shuihu zhuan” I@g PÙq« Wenyi bao ¿¯ 1952, 21.

268

Selected Bibliography

Liu Jian BÌ et al. Jindai Hanyu xuci yanjiu ÒN~y͸„s. Beijing: Yuwen Chubanshe, 1992. Liu Jian and Jiang Shaoyu ±–M, eds. Jindai Hanyu yufa ziliao huibian ÒN~yyk Í∆◊s. Beijing: Shangwu Yinshuguan, 1992. Liu Shide B@w. “Lun Jingben Zhong yi zhuan de shidai, xingzhi he diwei” ◊m?ª æq«n∫…NA?ËMaÏ. In Xiaoshuo xiqu yanjiu (IV) p°∏±„s ]ƒ|∞^. Ed. Guoli Qinghua Daxue Renwen Shehui Xueyuan Zhongwenxi ÍflMÿj«H¿|«|§Ât. Taipei: 1993. ———, ed. Zhongguo gudai xiaoshuo yanjiu: Taiwan Xianggang lunwen xuanji §ÍjN p°„sROWª‰◊ÂÔË. Shanghai: Shanghai Guji Chubanshe, 1983. Liu Xiaopeng BƒP. “Yongle dadian sanben xiwen yu wu da nanxi de jiegou bijiao” √÷jÂTª∏ÂP≠jn∏∫≤cÒ˚. Wenxue pinglun «˚◊ (Taipei) 1980. 3: 63–134. Liu Xinzhong BY§. Jin Shengtan de xiaoshuo lilun ˜tƒ∫p°z◊. Shijiazhuang: Hebei Renmin Chubanshe, 1986. Lu Decai |w~. Zhongguo gudai xiaoshuo yishu lun §ÍjNp°¿N◊. Tianjin: Nankai Daxue Chubanshe, 1987. Lu Xun |≥. Lu Xun lun wenxue yu yishu |≥◊«P¿N. Beijing: Renmin Wenxue Chubanshe, 1980. ———. Zhongguo xiaoshuo shilue §Íp°v§. In Lu Xun quanji |≥˛∞. 16 vols. Beijing: Renmin Wenxue Chubanshe, 1981. 9: 1–297. Lü Shuxiang fQ. “Ba zi yongfa de yanjiu” ‚rŒk∫„s. In his Hanyu yufa lunwenji. 125–144. ———. Hanyu yufa lunwen ji ~yyk◊Â∞. Beijing: Kexue Chubanshe, 1955. ———. Lü Shuxiang yuwen lunji fQyÂ◊∞ Beijing: Shangwu Yinshuguan, 1983. ———. “Shuo men” °Ã. In his Hanyu yufa lunwen ji. 145–168 ———. “Wenyan he baihua” •M’‹. In his Lü Shuxiang yuwen lunji. 55–76. ———. Xiandai Hanyu babai ci {N~yK?¸. Beijing: Shangwu Yinshuguan, 1980. Lun Zhongguo gudian xiaoshuo de yishu: Taiwan Xianggang lunzhu xuanji ◊§Íj p°∫¿NROWª‰◊¤ÔË. Eds. Ning Zongyi Áv@ and Lu Decai |w~. Tianjin: Nankai Daxue Chubanshe, 1984. Luo Ergang π∏ı. “Cong Luo Guanzhong San Sui ping yao zhuan kan Shuihu zhuan zhuzhe he yuanben” qπe§mTE≠Ø«n›mÙq«n¤ÃM Ϫ. In his Shuihu zhuan yuanben he zhuzhe yanjiu, 58–84. ———. Shuihu zhuan yuanben he zhuzhe yanjiu Ùq«ÏªM¤Ã„s. Nanjing: Jiangsu Guji Chubanshe, 1991. Ma Nancun ®n6. Yanshan yehua Ps]‹. Beijing: Beijing Chubanshe, 1993. Ma Tiji ®·e, ed. Shuihu shulu Ùq—˝. Shanghai: Shanghai Guji Chubanshe, 1986.

Selected Bibliography

269

———, ed. Shuihu ziliao huibian ÙqÍ∆◊s. Beijing: Zhonghua Shuju, 1980. Ma Youyuan ®ÆÆ (Y. W. Ma). “Cong zhao’an bufen kan Shuihu zhuan de chengshu guocheng” q¤w°¿›Ùq«∫®—L{. In his Shuihu lunheng, 141–175. ———. “Huyu yanjiu jianben Shuihu yijianshu” IS„s≤ªÙqN£—. In his Shuihu lunheng, 29–53. ———. Shuihu lunheng Ùq◊≈. Taipei: Lianjing Chuban Shiye Gongsi, 1992. ———. “Xiancun zuizao de jianben Shuihu zhuan” {sÃ≠∫≤ªÙq«. In his Shuihu lunheng, 55–96. Mao Dun Tfi. Mao Dun gudian wenxue lunwen ji Tfij«◊Â∞. Shanghai: Shanghai Guji Chubanshe, 1983. Masuda Mataru W–A. “ ‘Wahon’ to iu koto ni tsuite” b{∆¢§±∆…¬¢ƒ. Jimbun kenkyû H„s 16.5 (1965): 22–33. Mei Tsu-lin ˆ™Ô. “Cong yuyanshi kan jiben Yuan zaju binbai de xiezuo shiqi” qy•v›Xª∏¯@´’∫g@…¡. Yuyanxue luncong y•«◊O 13 (1984): 111–153. Min Ze ”A. Zhongguo wenxue lilun piping shi §Í«z◊Â˚v. Beijing: Renmin Wenxue Chubanshe, 1981. Nie Gannu øÊ∏. “Shuihu wu lun” Ùq≠◊. In his Zhongguo gudian xiaoshuo lunji §ÍjÂp°◊∞. Shanghai: Shanghai Guji Chubanshe, 1981. 9–204. Okamoto Sae ™{≥¶. Shindai kinsho no kenkyû ¥„÷ëçÜ. Tokyo: Tokyo University Tòyò Bunka Kenkyûjo, 1996. Òta Tatsuo ”–∞“. Chûgokugo rekishi bumpò §Íy˙vÂk. Tokyo, 1958. Ouyang Jian ⁄ß∑. “Shuihu zhuan jianben fanben dishan guocheng xinzheng” mÙq«n≤ªcªºQL{s“. In his Ming-Qing xiaoshuo xinkao ˙Mp°s“. Beijing: Zhongguo Wenlian Chuban Gongsi, 1992. 43–80. ———. “Jingben Zhong yi zhuan canye pingjia shangdui” m?ªæq«n›∂˚˘ ”I. In his Ming-Qing xiaoshuo xinkao ˙Mp°s“. 81–99. ———. “Ping yao zhuan: Shuihu zhuan de zimei pian” m≠Ø«nRmÙq«n∫ nfg. In his Ming Qing xiaoshuo caizheng ˙Mp°ˆø. Taipei: Guanya Wenhua, 1992. Ouyang Jian and Xiao Xiangkai Ω_. Shuihu xinyi Ùqs≥. Chongqing: Chongqing Chubanshe, 1983. Qian Xuantong ˙»P. “Rulin waishi xin xu” ßL~vs‘. In Rulin waishi huijiao huiping ben ßL~v|’|˚ª. 2: 785–786. Rong Zhaozu eF™. Mingdai sixiang shi ˙N‰Qv. Taipei, 1962. Shang Yanliu ”l‹. Qingdai keju kaoshi shulu MNÏ|“’z˝. Beijing: Sanlian Shudian, 1958. Shen Jianshi H›h. Zhongguo kaoshi zhidu shi §Í“’Ó◊v. Taipei: Taiwan Shangwu Yinshuguan, 1971.

270

Selected Bibliography

Shi Nan’an yanjiu I@g„s. Ed. Jiangsu Shehui Kexueyuan Wenxue Yanjiusuo ø¨¿|Ï«|«„s“ Nanjing: Jiangshu Guji Chubanshe, 1984. Shijing jinzhu jinyi ÷gµ`µ∂. Ed. Wang Yunwu ˝≥≠ and annotator Ma Chiyin ®˘’. Taipei: Taiwan Shangwu Yinshuguan, 1970 Shiroki Naoya ’ÏΩ]. “Kaku Butei hon shikò” ¢Zwªp“. In Hiroshima daigaku bungakubu kiyò 27.1 (1968): 140–159. ———. “Ko-shi, Shin-shi kyòcho kanyo shinhakkenteki Kyòbon chûgiden zanya hihan” ⁄Åæŧò÷óV≠ªIû{â``c≈·ª Tohogaku 66 (June 1983): 76–88. Shuihu yanjiu lunwen ji Ùq„s◊Â∞. Beijing: Zuojia Chubanshe, 1957. Shuihu zhengming ÙqßÔ. Ed. Hubei Shehui Kexueyuan Wenxue Yanjiusuo. 4 vols. Wuhan: Changjiang Wenyi Chubanshe, 1982–1985. Si Su ‰¨. “Shuoshu youwu jiaoben” °—≥L}ª. Quyi ¿ 52 (1962): 44–45. Song Yuan Ming Hanyu yanjiu ∫∏˙~y„s. Ed. Cheng Xiangqing {M. Ji’nan: Shandong Jiaoyu Chubanshe, 1992. Sun Kaidi ]¢ƒ. Cangzhou ji …{∞. 2 vols. Beijing: Zhonghua Shuju, 1965. ———. “Shuihu zhuan jiuben kao” Ùq«¬ª“. In his Cangzhou ji …{∞, 1: 121–143. ———. Sujiang, shuohua yu baihua xiaoshuo UøB°‹P’‹p° Beijing: Zuojia Chubanshe, 1956. ———. Zhongguo tongsu xiaoshuo shumu §ÍqUp°—ÿ. 1933. Reprint. Beijing: Zuojia Chubanshe, 1957. Sun Shuyu ]zt. Shuihu zhuan de laili, xintai he yishu Ùq«∫”˙AflAM¿N. Taipei: Shibao Wenhua Chuban Shiye Youxian Gongsi, 1983. Tan Daxian ”F˝. Zhongguo pingshu (pinghua) yanjiu §Í˚—]˚‹^„s. Hong Kong: Shangwu Yinshuguan, 1982. Tan Zhengbi ”øz. Huaben yu guju ‹ªPj@. Shanghai: Gudian Wenxue Chubanshe, 1957. ———. San yan Liang pai ziliao T•‚ÁÍ∆. Shanghai: Shanghai Guji Chubanshe, 1980. Wang Daocheng ˝D®. Keju shihua Ï|v‹. Beijing: Zhonghua Shuju, 1988. Wang Guowei ˝Í˚. Song Yuan xiqu shi ∫∏∏±v. Beijing: Dongfang Chubanshe, 1996. Wang Li ˝O. Hanyu shigao ~yvZ. 3 vols. Beijing: Kexue Chubanshe, 1958. Wang Liqi ˝Qπ. “Shuihu yu nongmin gemin” ÙqPA¡≤R. Shuihu yanjiu lunwen ji Ùq„s◊Â∞. Beijing: Zuojia Chubanshe, 1957. 61–76. ———. “Shuihu quanzhuan shi zenyang bei zuanxiu de” Ùq˛«OÁÀQ°◊∫. Wenxue pinglun «˚◊ 1982. 2: 86–101. Wang Xianpei ˝˝Æ and Zhou Weimin P∂¡. Ming-Qing xiaoshuo lilun piping shi ˙Mp°z◊Â˚v. Guangzhou: Huacheng Chubanshe, 1988. Wang Xiaojia ˝Âa. Shuihu xi kaolun Ùq∏“◊. Ji’nan: Ji’nan Chubanshe, 1989.

Selected Bibliography

271

Wang Zhouming ˝w˙ and Wu Runting ZÌ@. Zhongguo wenhua jinghua wenku: Zaju yu chuanqi §ÍÂ∆ÎÿÂwR¯@P«_. Ji’nan: Shandong Wenyi Chubanshe, 1992. Wong Yongjian ˝√∑. “Cong Ming-chu de Shuihu-xi kan Shuihu zhuan zuben de chengshu niandai” q˙Ï∫Ùq∏›mÙq«n™ª∫®—~N. Shuihu zhengming 3: 233–241. Wu Mei dˆ. Gu qu zhu tan U±ØÕ. In his Wu Mei xiqu lunwen ji dˆ∏± ◊Â∞. Beijing: Zhongguo Xiqu Chubanshe, 1985. Wu Tongrui dPÁ et al. Zhongguo suwenxue gailun §ÍU«ß◊. Beijing: Beijing Daxue Chubanshe, 1997. Wu Zuxiang d’. “Guanyu Sanguo yanyi” ˆ_TÍtq. In his Shuo bai ji, 37–38. ———. Shuo bai ji °˚O. Beijing: Beijing Daxue Chubanshe, 1987. Xia Zhiqing L”M (C. T. Hsia). “Suiwen yishi chongkan xu” ¶vÚ´Z«. In Zhongguo gudian xiaoshuo lunji §ÍjÂp°◊∞. 2 vols. Taipei: Yushi Wenhua Gongsi, 1975. 2: 119–144. Xie Bixia ¬—¯. Shuihu xiqu ershizhong yanjiu Ùq∏±GQÿ„s. Taipei: Guoli Taiwan Daxue Chuban Weiyuanhui, 1981. Xu Jinbang \˜]. Zhongguo xiqu wenxue shi §Í∏±Â«v. Beijing: Zhongguo Wenxue Chubanshe, 1994. Xu Shuofang }“Ë. “Cong Song Jiang qiyi dao Shuihu zhuan chengshu” q∫ø _qÏmÙq«n®—. In his Lun Tang Xianzu ji qita ◊ˆ„™Œ‰L. Shanghai: Shanghai Guji Chubanshe, 1983. 179–200. ———. “Jin ping mei de xieding zhe shi Li Kaixian” ˜~ˆ∫gwÃOı}˝. In his Lun Tang Xianzu ji qita, 133–147. Yan Dunyi Y∞ˆ Shuihu zhuan de yanbian Ùq«∫t‹. Beijing: Zuojia Chubanshe, 1957. ———. “Xiyou ji he gudai xiqu de guanxi” ËÂOMjN∏±∫ˆt. In his Yuan Ming Qing xiqu lunji, 134–140. ———. Yuan Ming Qing xiqu lunji ∏˙M∏±◊∞. Zhengzhou: Zhongzhou Shuhuashe, 1982. Yang Tiange ®—‡. “Shuo ‘wu’ ” °ßa®. Zhongguo yuwen §Íy 1980. 5: 363–367. Yang Yi ®q. “Shuihu zhuan de xushi shenli” mÙq«n∫‘∆´z. Qilu xuekan N |«Z 1994. 1: 16–28. Ye Dejun ≠w°. Song-Yuan-Ming jiangchang wenxue ∫∏˙ø¤Â«. Shanghai: Zhonghua Shuju, 1957. Ye Lang ≠‘. Zhongguo xiaoshuo meixue §Íp°¸«. Beijing: Beijing Daxue Chubanshe, 1982. Ye Qingbing ≠y± (and Shao Hong Úı), ed. Zhongguo wenxue piping ziliao huibian: Ming dai §Í«Â˚Í∆◊sR˙N. Taipei: Chengwen, 1979. Yi Yongwen Ï√Â. “Zai lun Shuihu zhuan shi fanying shimin jieceng liyi de zuopin”

272

Selected Bibliography

A◊mÙq«nOœM´¡•hQq∫@~. Hebei daxue xuebao e_j«« ¯ 1980. 4: 65–75. You Guoen ÂͶ et al. Zhongguo wenxue shi §Í«v. Hong Kong: Sanlian Shudian, 1986. Yu Jiaxi E≈¸. Shuihu zhuan sanshiliu ren kaoshi Ùq«TQªH“Í. Beijing: Zuojia Chubanshe, 1955. Yuan Zhenyu K_t and Liu Mingjin B˙µ. Mingdai wenxue piping shi ˙N« Â˚v. Shanghai: Shanghai Guji Chubanshe, 1991. Zhang Guoguang iÍ˙. “Lu Xun de Shi Nai’an wei fanben shuihu zuozhe zhi tuoming shuo wuke zhiyi” |≥∫I@g∞cªÙq@Ãß´W°Li Ë√. Shuihu zhengming 1: 49–64. ———. “1982 nian de ‘Shi Nai’an re’ cong he er lai” @EKG~∫ßI@gˆ® qÛ””. Jianghan luntan ø~◊¬. 1983.3. ———. “Shuihu zuben tankao” Ùq™ª¥“. Jianghan luntan ø~◊¬. 1982. 1: 39–44. ———. Shuihu yu Jin Shengtan yanjiu ÙqP˜tƒ„s. Zhengzhou: Zhongzhou Shuhuashe, 1981. Zhang Peiheng πˆÌ. “Guanyu Shuihu de Guo Xun ben yu Yuan Wuya ben” ˆ_ Ùq∫¢‘ªPKLPª. In his Xianyi ji m√∞. Changsha: Yuelu Shushe, 1993. 222–240. ———. “Guanyu xiancun de suowei ‘Song huaben’ ” ˆ_{s∫“◊ß∫‹ª®. Shanghai daxue xuebao W¸j««¯ 1996. 1: 15–30. ———. “Shi Nai’an muzhi bianwei ji qita” mI@g””nÎ∞Œ‰L. In his Xianyi ji, 179–202. ———. “Shi Yanduan shifou Shi Nai’an” I¤›O_I@g. In his Xianyi ji, 203–221. Zhang Xiaohu iSÍ. “Shi Nai’an wei jianghu yiren shuo” I@g∞øÚ¿H°. In Zhonghua wenshi luncong §ÿÂv◊O 1982. 4: 89–100. Zhang Yiming π@Ô. Jin ping mei cihua he Mingdai kouyu cihui yufa yanjiu ˜~ˆ ¸‹M˙Nfy¸◊yk„s. Shanghai: Shanghai Guji Chubanshe, 1997. Zhang Yuechao iÎW. “Zhongxi wenlun jige fangmian de chubu bijiao yanjiu” §Ë Â◊X”˱∫ÏBÒ˚„s. In Bijiao wenxue lunwen ji Ò˚«◊Â∞. Tianjin: Nankai Daxue Chubanshe, 1984. Zhao Cong Øo. Zhongguo si da xiaoshuo zhi yanjiu §Í|jp°ß„s. Hong Kong: Yulian Chubanshe, 1964. Zhao Jingshen Ø∫`. “Tan Ming Chenghua kanben Shuochang cihua” Õ˙®∆Z ª°¤¸‹. Wenwu ´ 11 (1972): 19–22. ———. Yuan Ming nanxi kaolue ∏˙n∏“§. Beijing: Renmin Wenxue Chubanshe, 1990. Zheng Gongdun GΩfi. Shuihu zhuan lunwen ji Ùq«◊Â∞. 2 vols. Yinchuan: Ningxia Renmin Chubanshe, 1983.

Selected Bibliography

273

Zheng Zhenduo G∂M. “Shuihu zhuan de yanhua” Ùq«∫t∆. In his Zhongguo wenxue yanjiu xinbian §Í«„sss. 101–157. ———. “Song-Jin-Yuan zhugongdiao kao” ∫˜∏—c’“. In his Zhongguo wenxue yanjiu xinbian, 843–970. ———. “Xixiang ji de benlai mianmu shi zenyangde” Ë[O∫ª”±ÿOÁÀ∫. In his Zhongguo wenxue yanjiu xinbian, 596–612. ———. Zhongguo suwenxue shi §ÍU«v. Beijing: Dongfang, 1996. ———. Zhongguo wenxue yanjiu xinbian §Í«„sss. Taipei: Minglun, 1971. Zhongguo tongsu xiaoshuo zongmu tiyao §ÍqUp°`ÿ£n. Ed. Jiangsu Sheng Shehuikexueyuan Ming Qing Xiaoshuo Yanjiu Zhongxin ø¨Ÿ¿|Ï« |˙Mp°„s§fl. Beijing: Zhongguo Wenlian Chuban Gongsi, 1990. Zhongguo wenxue shi §Í«v. Eds. Zhang Peiheng πˆÌ and Luo Yuming d…˙. 3 vols. Shanghai: Fudan Daxue Chubanshe, 1996. Zhongguo wenxue shi §Í«v. Ed. Zhongguo Kexueyuan Wenxue Yanjiusuo §Í Ï«|«„s“. Beijing, 1962. Zhongguo wenxue shi §Í«v. Ed. Beijing Daxue Zhongwenxi _?j«§Ât. 4 vols. Beijing: Renmin Wenxue Chubanshe, 1959. Zhou Zhaoxin P¸s. “ ‘Huaben’ shiyi” ß‹ª®¿q. Guoxue yanjiu (di er juan) Í«„s]ƒG˜^. Ed. Beijing Daxue Zhongguo Chuantong Wenhua Yanjiu Zhongxin _?j«§Í«ŒÂ∆„s§fl. Beijing: Beijing Daxue Chubanshe, 1994. 195–209. Zhu Dexi ∂w≥. “Laoqida yanjie, Pu tongshi yanjie shuhou” m—^jŒ—nmÎq∆ Œ—n—·. Beijing daxue xuebao _?j««¯ 1958. 2: 69–75. Zhu Dongrun ∂FÌ Zhongguo wenxue piping shi dagang §Í«Â˚vjı Shanghai: Gudian Wenxue Chubanshe, 1957. Zhu Qing «M and Li Yonggu ı√˛. “Shuihu zhuan zuben ji ‘Guo Wu dingben’ wenti xinyi” mÙq«n™ªŒß¢Zwª®›Ds≥. Wenxue yichan «ڣ 1997. 5: 81–92. Zhu Yixuan ∂@», ed. Ming Qing xiaoshuo ziliao xuanbian ˙Mp°Í∆Ôs. 2 vols. Ji’nan: Qi Lu Shushe, 1989. Zhu Yixuan and Liu Yuchen B∂’, eds. Shuihu zhuan ziliao huibian Ùq«Í∆ ◊s. Tianjin: Baihua Wenyi Chubanshe, 1981.

Sources in Western Languages Abrams, M. H. The Mirror and the Lamp: Romantic Theory and the Critical Tradition. New York: Norton, 1958. Addison, Joseph. The Spectator. 8 vols. London: J.M. Dent, 1897. Auerbach, Erich. Mimesis. Trans. Willard R. Trask. Princeton: Princeton University Press, 1953. Augustinus, Aurelius (St. Augustine). On Christian Doctrine. Trans. D. Robertson Jr. New York: Liberal Arts, 1958.

274

Selected Bibliography

Bakhtin, M. M. The Dialogic Imagination. Trans. Caryl Emerson and Michael Holquist. Austin: University of Texas Press, 1981. Bender, Mark. “Key to Performance in Kunming Storytelling.” CHINOPERL Papers 18 (1996): 21–37. Bishop, John Lyman. “Some Limitations of Chinese Fiction.” In Studies in Chinese Literature, ed. J. L. Bishop. Cambridge, MA: Harvard University Press, 1965. Bloom, Harold. The Anxiety of Influence. New York: Oxford University Press, 1973. Børdahl, Vibeke, ed. The Eternal Storyteller: Oral Literature in Modern China. Richmond, UK: Curson, 1999. ———. “A ‘Poetics’ of Chinese Storytelling: The System of Terms Used among Yangzhou Storytellers.” In Vibeke Børdahl, ed., The Eternal Storyteller: Oral Literature in Modern China, 223–252. ———. “Narrative Voices in Yangzhou Storytelling.” CHINOPERL Papers 18 (1995): 1–31. Buck, Pearl S., trans. All Men Are Brothers. New York: John Day, 1937. Carlitz, Katherine. The Rhetoric of Chin p’ing mei. Bloomington: Indiana University Press, 1986. Carothers, John Coplin. “Culture, Psychiatry, and the Written Word.” Psychiatry 22 (1959): 307–320. Chadwick, H. M., and N. K. Chadwick. The Growth of Literature. 3 vols. Cambridge: Cambridge University Press, 1988. Cheung, Samuel H. N. “Structural Cyclicity in Shuihu zhuan: From Self to Sworn Brotherhood.” CHINOPERL Papers 15 (1990): 1–16. Chou, Chih-p’ing. Yuan Hung-tao and the Kung-an School. Cambridge: Cambridge University Press, 1988. Culler, Jonathan. Structuralist Poetics. London: Routledge, 1975. de Bary, W. T. “Chinese Despotism and the Confucian Ideal: A SeventeenthCentury View.” In John Fairbank, ed., Chinese Thought and Institutions, 163–203. DeFrancis, John. The Chinese Language: Fact and Fantasy. Honolulu: University of Hawai‘i Press, 1984. ———. Visible Speech: The Diverse Oneness of Writing Systems. Honolulu: University of Hawai‘i Press, 1989. Dreyer, Edward L. Early Ming China: A Political History. Stanford: Stanford University Press, 1981. Dundes, Alan. “From Etic to Emic Units in the Structural Study of Folktales.” Journal of American Folklore 75 (1965): 95–105. ———. “Metafolklore and Oral Literary Criticism.” The Monist 50 (1966): 506–516. ———. “Structural Typology in North American Indian Folklore.” In Alan Dundes, ed., The Study of Folklore. Englewood: Prentice-Hall, 1965. 206–215.

Selected Bibliography

275

Eliot, T. S. “Tradition and the Individual Talent.” In his Sacred Wood. London: Methuen, 1920. 47–59. Elman, Benjamin A. “Changes in Confucian Civil Service Examinations from the Ming to the Ch’ing Dynasty.” In Benjamin A. Elman and Alexander Woodside, eds., Education and Society in Late Imperial China, 1600–1900. Berkeley: University of California Press, 1994. Eoyang, Eugene. “Oral Narrative in the pien and pien-wen.” Archiv Orientální 46 (1978): 233–252. ———. “A Taste for Apricots: Approaches to Chinese Fiction.” In Andrew H. Plaks, ed., Chinese Narrative: Critical and Theoretical Essays. 53–72. ———. The Transparent Eye: Reflections on Translation, Chinese Literature, and Comparative Poetics. Honolulu: University of Hawai‘i Press, 1993. ———. Word of Mouth: Oral Storytelling in the Pien-wen. Ann Arbor: University Microfilms, 1972. Eoyang, Eugene and Lin Yao-fu, eds. Translating Chinese Literature. Bloomington: Indiana University Press, 1995. Fairbank, John, ed. Chinese Thought and Institutions. Chicago: University of Chicago Press, 1957. Fang, Thomé H. “The World and the Individual in Chinese Metaphysics.” In Charles A. Moore, ed., The Chinese Mind: Essentials of Chinese Philosophy and Culture. Honolulu: University of Hawai‘i Press, 1967. 128–251. Feng Yu-lan. A History of Chinese Philosophy. 2 vols. Trans. Derk Bodde. Princeton: Princeton University Press, 1953. Fenik, Bernard. Typical Battle Scenes in the Iliad. Wiesbaden: F. Steiner, 1968. Finnegan, Ruth. Literacy and Orality: Studies in the Technolog y of Communication. New York: Blackwell, 1988. Foley, John Miles, ed. Oral-Formulaic Theory: A Folklore Casebook. New York: Garland, 1990. ———, ed. Oral Tradition in Literature. Columbia: University of Missouri Press, 1986. Frye, Northrop. Anatomy of Criticism: Four Essays. Princeton: Princeton University Press, 1957. Gardner, Daniel K. “Modes of Thinking and Modes of Discourse in the Sung: Some Thoughts on the Yü-lu (Recorded Conversations) Texts.” The Journal of Asian Studies 50:3 (1991): 574–603. Genette, Gérard. Narrative Discourse: An Essay in Method. Trans. Jane E. Lewin. Ithaca: Cornell University Press, 1980. ———. Narrative Discourse Revisited. Trans. Jane E. Lewin. Ithaca: Cornell University Press, 1988. Goody, Jack, and Ian Watt. “The Consequence of Literacy.” In Jack Goody, ed., Literacy in Traditional Societies. Cambridge: Cambridge University Press, 1968. 27–68.

276

Selected Bibliography

Gray, Bennison. “Repetition in Oral Literature.” Journal of American Folklore 84 (1971): 289–303. Hanan, Patrick. The Chinese Short Story. Cambridge, MA: Harvard University Press, 1973. ———. The Chinese Vernacular Story. Cambridge, MA: Harvard University Press, 1981. ———. “The Composition of the P’ing-yao chuan.” Harvard Journal of Asiatic Studies 31 (1971): 201–219. ———. “The Early Chinese Short Story: A Critical Theory in Outline.” Harvard Journal of Asiatic Studies 27 (1987): 168–207. ———. “The Nature of Ling Meng-chu’s Fiction.” In Andrew Plaks, ed., Chinese Narrative: Critical and Theoretical Essays. 85–114. ———. “Sources of the Chin P’ing Mei.” Asia Major n.s. 10 (1963): 23–67. ———. “The Text of the Chin P’ing Mei.” Asia Major n.s. 9 (1962): 1–57. Hannas, William. Asia’s Orthographic Dilemma. Honolulu: University of Hawai‘i Press, 1997. Hansen, William F. The Conference Sequence: Patterned Narration and Narrative Inconsistency in the Odyssey. Berkeley: University of California Press, 1972. Havelock, Eric Alfred. Literate Revolution in Greece and Its Cultural Consequences. Princeton: Princeton University Press, 1982. ———. The Muse Learns to Write: Reflections on Orality and Literacy from Antiquity to the Present. New Haven: Yale University Press, 1986. ———. Origins of Western Literacy. The Ontario Institute for Studies in Education, 1976. ———. Preface to Plato. New York: Atheneum, 1965. Hawkes, David. “Reflections on Some Yuan Tsa-chü.” Asia Major 16 (1971): 69–81. Hayden, George A. “A Skeptical Note on the Early History of Shui-hu chuan.” Monumenta Serica 32 (1976): 374–399. Hegel, Robert E. Reading Illustrated Fiction in Late Imperial China. Stanford: Stanford University Press, 1998. ———. “Distinguishing Levels of Audiences for Ming-Ch’ing Vernacular Literature: A Case Study.” In Popular Culture in Late Imperial China, eds. David Johnson et al. 112–142. ———. The Novel in Seventeenth-Century China. New York: Columbia University Press, 1981. ———. “Traditional Chinese Fiction—The State of the Field.” The Journal of Asian Studies 53:2 (1994): 394–426. Hennessey, William O. “Classical Sources and Vernacular Resources in Xuanhe yishi: The Presence of Priority and the Priority of Presence.” Chinese Literature: Essays, Articles, Reviews 6 (1984): 33–52.

Selected Bibliography

277

Ho, Ping-ti. The Ladder of Success in Imperial China. New York: Columbia University Press, 1976. Homer. The Iliad. Trans. Richmond Lattimore. Chicago: University of Chicago Press, 1951. ———. The Odyssey. Trans. Robert Fitzgerald. Garden City, NY: Doubleday, 1963. Howe, Michael J. A. Introduction to Human Memory: A Psychological Approach. New York: Harper, 1970. Hrdlickova, V. “The Professional Training of Chinese Storytellers and the Storytellers’ Guilds.” Archiv Orientální 33 (1965): 225–248. ———. “Some Questions Connected with Tun-huang pien-wen.” Archiv Orientální 30 (1962): 211–230. Hsia, C. T. The Classic Chinese Novel: A Critical Introduction. New York: Columbia University Press, 1968. Huang, Martin Weidong. “Dehistoricization and Intertextualization: The Anxiety of Precedents in the Evolution of the Traditional Chinese Novel.” Chinese Literature: Essays, Articles, Reviews 12 (1990): 45–68. ———. Literati and Self-Re/Presentation: Autobiographical Sensibility in the EighteenthCentury Chinese Novel. Stanford: Stanford University Press, 1995. Ichisada, Miyazaki. China’s Examination Hell: The Civil Service Examination of Imperial China. Trans. Conrad Schirokauer. New Haven: Yale University Press, 1976. Idema, Wilt L. Chinese Vernacular Fiction: The Formative Period. Leiden: Brill, 1974. ———. “Data of the Chu-kung-tiao.” T’oung Pao 79 (1993): 69–112. ———. The Dramatic Oeuvre of Chu Yu-tun [1379–1439]. Leiden: Brill, 1985. ———. “Zhu Youdun’s Dramatic Prefaces and Traditional Fiction.” Ming Studies 10 (1980): 17–22. Irwin, Richard G. The Evolution of a Chinese Novel: Shui-hu chuan. Cambridge, MA: Harvard University Press, 1953. ———. “Water Margin Revisited.” T’oung Pao 48:4–5 (1960): 393–415. Iser, Wolfgang. The Implied Reader: Patterns of Communication in Prose Fiction from Bunyan to Beckett. Baltimore: Johns Hopkins University Press, 1974. Jameson, Fredric. “Metacommentary.” Reprinted in Contemporary Literary Criticism. Ed. Robert Con Davis. New York: Longman, 1986. 112–123. Johnson, David, et al., eds. Popular Culture in Late Imperial China. Berkeley: University of California Press, 1985. Johnson, Linda Cooke, ed. Cities of Jiangnan in Late Imperial China. Albany, NY: State University of New York Press, 1993. Kao, Karl S. Y. “Aspects of Derivation in Chinese Narrative.” Chinese Literature: Essays, Articles, Reviews 7 (1985): 1–36. ———, ed. Classical Chinese Tales of the Supernatural and the Fantastic: Selections from the Third to the Tenth Century. Bloomington: Indiana University Press, 1985.

278

Selected Bibliography

Karlgren, Bernhard. The Chinese Language: An Essay on Its Nature and History. New York: Ronald Press, 1949. Kracke, E. A., Jr. “Region, Family, and Individual in the Chinese Examination System.” In John Fairbank, ed., Chinese Thought and Institutions. Chicago: University of Chicago Press, 1957. 251–268. Lévy, André. “About the Chinese Storyteller’s Change of Name.” In Vibeke Børdahl, ed., The Eternal Storyteller: Oral Literature in Modern China. 33–39. Li, Peter. “Narrative Patterns in San-kuo and Shui-hu.” In Andrew H. Plaks, ed., Chinese Narrative: Critical and Theoretical Essays. 73–84. Liu Hsieh. The Literary Mind and the Carving of Dragons. Trans. Vincent Yu-chung Shih. New York: Columbia University Press, 1959. Liu, James J. Y. Chinese Theories of Literature. Bloomington: Indiana University Press, 1966. ———. Language-Paradox-Poetics: A Chinese Perspective. Princeton: Princeton University Press, 1988. Liu, Wu-chi. An Introduction to Chinese Literature. Bloomington: Indiana University Press, 1966. Lord, Albert B. “Perspectives on Recent Work on the Oral Traditional Formula [excerpted].” In John Miles Foley, ed., Oral Formulaic Theory: A Folklore Casebook. New York: Garland, 1990. 379–405. ———. The Singer of Tales. Cambridge, MA: Harvard University Press, 1960. Lu, Sheldon Hsiao-peng. From Historicity to Fictionality: The Chinese Poetics of Narrative. Stanford: Stanford University Press, 1994. Lu Xun. A Brief History of Chinese Fiction. Trans. Yang Hsien-yi and Gladys Yang. Beijing: Foreign Languages Press, 1976. Lukacs, Georg. The Theory of the Novel: A Historico-Philosophical Essay on the Forms of Great Epic Literature. Trans. Anna Bostock. Cambridge, MA: Massachusetts Institute of Technology Press, 1971. Ma, Laurence J. C. Commercial Development and Urban Change in Sung China (960–1279). Ann Arbor: Department of Geography, University of Michigan Press, 1971. Ma, Tai-loi. “Novels Prohibited in the Literary Inquisition of Emperor Ch’ien-lung, 1722–1788.” In Winston L. Y. Yang and Curtis P. Adkins, eds., Critical Essays on Chinese Fiction, 201–212. Ma, Y. W., and Joseph S. M. Lau, eds. Traditional Chinese Stories: Themes and Variations. New York: Columbia University Press, 1978. Mair, Victor H. “Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages.” The Journal of Asian Studies 53:3 (1994): 707–751. ———. “Lay Students and the Making of Written Vernacular Narrative: An Inventory of Tun-huang Manuscripts.” CHINOPERL Papers 10 (1981): 5–96.

Selected Bibliography

279

———. Painting and Performance: Chinese Picture Recitation and Its Indian Genesis. Honolulu: University of Hawai‘i Press, 1988. ———. T’ang Transformation Texts. Cambridge, MA: Harvard University Council on East Asian Studies, 1989. ———. Tun-huang Popular Narratives. Cambridge: Cambridge University Press, 1983. Marmé, Michael. “Heaven on Earth: The Rise of Suzhou, 1127–1550.” In Linda Cooke Johnson, ed., Cities of Jiangnan in Late Imperial China, 17–45. McLaren, Anne E. Chinese Popular Culture and Ming Chantefables. Leiden: Brill, 1998. McMahon, Keith. Causality and Containment in Seventeenth-Century Chinese Fiction. Leiden and New York: E. J. Brill, 1988. Mencius. The Works of Mencius. Trans. James Legge. Hong Kong, 1961. Mote, F. W. “China’s Past in the Study of China Today—Some Comments on the Recent Work of Richard Solomon.” Journal of Asian Studies 32:1 (November 1972): 106–120. Olrik, Axel. “Epic Laws of Folk Narrative.” In Alan Dundes, ed., The Study of Folklore. Englewood: Prentice-Hall, 1965. 129–141. Ong, Walter J. Interfaces of the Word: Studies in the Evolution of Consciousness and Culture. Ithaca: Cornell University Press, 1977. ———. Orality and Literacy: The Technologizing of the Word. London: Methuen, 1982. ———. Presence of the Word: Some Prolegomena for Cultural and Religious History. Minneapolis: University of Minnesota Press, 1981. ———. “Text as Interpretation: Mark and After.” In John Miles Foley ed., Oral Tradition in Literature, 147–169. Origen. Origen: An Exhortation to Martyrdom Prayer. Trans. Rowan A. Greer. New York: Paulist Press, 1979. Owen, Stephen. Readings in Chinese Literary Thought. Cambridge, MA: Council on East Asian Studies, Harvard University, 1992. ———Traditional Chinese Poetry and Poetics: Omen of the World. Madison: University of Wisconsin Press, 1985. Parry, Milman. The Making of Homeric Verse. Ed. Adam Parry. Oxford: Clarendon, 1971. Plaks, Andrew H. The Four Masterworks of the Ming Novel. Princeton: Princeton University Press, 1987. ———. “Full-length Hsiao-shuo and the Western Novel: A Generic Reappraisal.” In William Tay et al., ed., China and the West: Comparative Literature Studies, 163–176. ———. “Shui-hu chuan and the Sixteenth-Century Novel Form: An Interpretive Analysis.” Chinese Literature: Essays, Articles, Reviews 2:1 (1980): 3–54. ———. “Towards a Critical Theory of Chinese Narrative.” In Andrew Plaks, ed., Chinese Narrative: Critical and Theoretical Essays, 309–352.

280

Selected Bibliography

———, ed. Chinese Narrative: Critical and Theoretical Essays. Princeton: Princeton University Press, 1977. Plato. The Republic. Trans. Desmond Lee. 2nd ed. Penguin, 1974. Polo, Marco. The Travels. Penguin, 1958. Porphyry. The Cave of the Nymphs. Trans. Thomas Taylor. Grand Rapids, MI: Phanes Press, 1991. Porter, Deborah. “Setting the Tone: Aesthetic Implications of Linguistic Patterns in the Opening Section of Shui-hu chuan.” Chinese Literature: Essays, Articles, Reviews 14 (1992): 51–75. ———. “Toward an Aesthetic of Chinese Vernacular Fiction: Style and the Colloquial Medium of Shui-hu chuan.” T’oung Pao 79 (1993): 114–153. Poulet, Georges. “Phenomenology of Reading.” Reprinted in Contemporary Literary Criticism. Ed. Robert Con Davis. New York: Longman, 1986. 350–362. Propp, Vladimir. Morpholog y of the Folktale. Trans. Laurence Scott. 2nd ed. Austin: University of Texas Press, 1968. Prusek, Jaroslav. “The Beginning of Popular Chinese Literature: Urban Centers— The Cradle of Popular Fiction.” Archiv Orientalni 36 (1968): 67–121. ———. Chinese History and Literature. Dordrecht: D. Reidel, 1970. ———. The Origins and the Authors of the Hua-pen. Prague: Academia, 1967. Rawski, Evelyn Sakakida. Education and Popular Literacy in Ch’ing China. Ann Arbor: University of Michigan Press, 1979. Richardson, Samuel. Correspondence. 6 vols. London, 1804. Ricoeur, Paul. “What Is a Text?” In Mythic-Symbolic Language and Philosophical Anthropolog y. Ed. David Rasmussen. Hague: Martinus Nijhoff, 1971. Rolston, David L. “Point of View in the Writing of Traditional Chinese Fiction Critics.” Chinese Literature: Essays, Articles, Reviews 15 (1993): 113–142. ———. “Sources of Traditional Chinese Fiction Criticism.” In David L. Rolston, ed., How to Read the Chinese Novel, 3–34. ———. Traditional Chinese Fiction and Fiction Commentary. Stanford: Stanford University Press, 1997. ———, ed. How to Read the Chinese Novel. Princeton: Princeton University Press, 1990. Roy, David T. “The Fifteenth-Century Shuochang cihua as Examples of Written Formulaic Composition.” CHINOPERL Papers 10 (1981): 97–128. ———, trans. The Plum in the Golden Vase, or Chin P’ing Mei. Vol. 1. Princeton: Princeton University Press, 1993. Said, Edward. The World, the Text, and the Critic. Cambridge, MA: Harvard University Press, 1983. Santangelo, Paolo. “Urban Society in Later Imperial Suzhou.” Trans. Adam Victor. In Linda Cooke Johnson, ed., Cities of Jiangnan in Late Imperial China, 81–116.

Selected Bibliography

281

Scholes, Robert, and Robert Kellogg. Nature of Narrative. New York: Oxford University Press, 1966. Shapiro, Sidney, trans. Outlaws of the Marsh. 2 vols. Bloomington: Indiana University Press and Foreign Language Press, 1981. Shih, Chung-wen. The Golden Age of Chinese Drama: Yüan Tsa-chü. Princeton: Princeton University Press, 1976. Solomon, Richard. Mao’s Revolution and the Chinese Political Culture. Berkeley: University of California Press, 1971. Stahl, Sandra, K. Dolby. Literary Folkloristics and the Personal Narrative. Bloomington: Indiana University Press, 1989. Sveinsson, Einar Ol. Njáls Saga: A Literary Masterpiece. London: University of Nebraska Press, 1971. Tay, William, et al., ed. China and the West: Comparative Literature Studies. Hong Kong: Chinese University Press, 1980. Taylor, Insup. Writing and Literacy in Chinese, Korean and Japanese. Philadelphia: John Benjamins, 1995. T’ien, Ju-k’ang. Male Anxiety and Female Chastity: A Comparative Study of Chinese Ethical Values in Ming-Ch’ing Times. New York, Leiden: Brill, 1988. Tulving, Endel. Elements of Episodic Memory. Oxford: Clarendon, 1983. Wang, Ching-Hsien. The Bell and the Drum: Shih Ching as Formulaic Poetry in an Oral Tradition. Berkeley: University of California Press, 1974. Wang, John C. Y., trans. “How to Read The Fifth Book of Genius.” Annotated by David L. Rolston. In David L. Rolston, ed., How to Read the Chinese Novel. 131–145. Ward, Barbara. “Readers and Audiences: An Exploration of the Spread of Traditional Chinese Culture.” In Text and Context: The Social Anthropolog y of Tradition, ed. Ravindra Jain. Philadelphia: Institute for the Study of Human Issues, 1977. 181–203. Watt, Ian. The Rise of the Novel. Berkeley: University of California Press, 1957. Reprint 1971. West, Stephen. “Translation as Research: Is There an Audience in the House?” In Translating Chinese Literature, eds. Eugene Eoyang and Lin Yao-fu. Bloomington: Indiana University Press, 1995. 131–155. West, Stephen, and Wilt L. Idema, trans. The Moon and the Zither: The Story of the Western Wing. Berkeley: University of California Press, 1991. Wilhelm, Hellmut. “The Scholar’s Frustration: Notes on a Type of Fu.” In John K. Fairbank, ed., Chinese Thoughts and Institutions. Chicago: University of Chicago Press, 1957. Williams, Raymond. Writing in Society. New York: Schocken Books, 1983. Wu, Yenna. “Outlaws’ Dreams of Power and Position in Shuihu zhuan.” Chinese Literature: Essays, Articles, Reviews 18 (1996): 45–67.

282

Selected Bibliography

Yang, Winston L. Y., and Curtis P. Adkins, eds. Critical Essays on Chinese Fiction. Hong Kong: Chinese University Press, 1980. Yu, Anthony C. “History, Fiction, and Reading Chinese Narrative.” Chinese Literature: Essays, Articles, Reviews 10 (1988): 1–19. ———, trans. The Journey to the West. 4 vol. Chicago: University of Chicago Press, 1977–1984. Zhang, Longxi. The Tao and the Logos: Literary Hermeneutics, East and West. Durham: Duke University Press, 1992. Zhao, Henry Y. H. The Uneasy Narrator: Chinese Fiction from the Traditional to the Modern. New York: Oxford University Press, 1995.

Index Abrams, M. H., 215n. 65 audience, 8, 99, 112, 144; disdain for country people, 155–156; enchantment with storytelling, 152; familiar with tradition, 41, 42; impact on storymaking, 8, 152–153, 154 –161, 236n. 58; objectified in text, 181; relationship with storyteller, 8, 49, 65, 152, 154; response of, 145–148. See also storyteller; storytelling bagua (Eight Trigrams), 12, 201n. 9, 216n. 73 bagu wen (Eight-Legged Essay), 53, 167–168, 176, 203n. 29, 216n. 73, 238nn. 97, 98. See also civil service examinations Baichuan shuzhi, 3 baihua (Chinese written vernacular), 2, 3, 4, 11, 17, 27, 34 –35, 93; as different language mixes, 21; in pinghua, 29–30; mixtures with wenyan, 17; in the Yuan period, 20. See also binbai; vernacular prose Bai Juyi (772–846), 17, 34 “Bai Niangzi yong zhen Leifengta,” 165 Bai Pu (1226–?), 19 Baitu ji, 27–28 Baiyue ting, 27–28 Bakhtin, M. M., 196 baojuan, 67, 227n. 40 “Bao Longtu Chenzhou tiaomi ji,” 47, 96, 214n. 51 “Bao Longtu duan baihu jing zhuan,” 45 “Bao Longtu duan Cao Guojiu gong’an zhuan,” 47, 96 “Bao Longtu duan Zhao huangqin Sun Wenyi gong’an zhuan,” 96 “Bao Shizhi chushen zhuan,” 47, 93, 214n. 50

Bao Xi, 12 Baozi heshang zi huansu, 41, 132, 207n. 62, 230n. 88 Bender, Mark, 82, 154 bianwen, 21–22, 32, 34, 45, 67, 148, 204nn. 42, 45, 227n. 40; homophonic substitutes in, 112; prosimetric form of, 22; relationship to spoken presentation, 21–22 binbai, 24, 25, 26; amplified in Ming-edition zaju, 26; in early and late xiwen texts compared, 27–28; scarcity in Yuanedition saju, 33, 82, 206n. 58; in Shuihu zaju, 26 –27; in Xixiang dingben, 26; in Xixiang ji zaju, 25–26; in Yuan-edition and Ming-edition zaju compared, 26; in zaju performance, 26; in Zhu Youdun’s plays, 27. See also baihua; vernacular prose Bing Yang Xiong, 132 Boccsccio, Giovanni, 2 Børdahl, Vibeke, 82, 154, 187 Buck, Pearl, 94 Buddhism: influence on fanqie, 17; transcribing foreign names and terms in Chinese characters, 18; translations of sutras, 17–18, 203n. 32. See also Chan Buddhism bu li wenzi (no establishment of written words), 18 cairen, 164 –165, 167, 169, 171, 172. See also laolang; shuhui; shuhui xiansheng Cang Jie, 12, 201n. 8 Canterbury Tales, 2, 151 Cervantes, 146 Chai Jin, 71, 85, 87, 173 Chan Buddhism, 18; distrust of writing, 18. See also Buddhism; yulu changzhuan, 50, 67

283

284

Index

chantefables, 50, 67 Chao Gai, 38, 39, 46, 91, 211n. 19 chaoyutiwen, 10–11 Chaucer, Geoffrey, 2 Chen Ruheng, 214n. 55, 215n. 56 “Chen Xunjian meilin shiqi ji,” 43 Chinese characters: ancient Chinese perception of, 11–13; divergence from speech, 10–11; exalted over speech, 11; legendary inventions of, 12–13, 201nn. 8, 9; phonetic denotation of, 4 –5; pictographic origins of, 12, 14; presumed signifying capacity of, 14; in registering the r sound, 24 –25; relationship to speech, 4, 13; representing patterned forms, 11–13; as schematizations of aesthetic patterns, 16; semanticphonetic composites, 4, 200n. 12; terminology for, 4; in transcribing Buddhist terms and names, 18. See also script; writing Chinese dialects: dialectal heterogeneity, 21; northern dialect as basis for a lingua franca, 21, 24, 35, 203n. 41; southern dialects, 46 – 47 chuanqi (dramatic genre), 29, 208n. 71 chuanqi (narrative genre), 182–183 Chuo geng lu, 151, 235n. 45 cihua, 44 – 45; as another term for huaben, 212n. 36; formal features of, 45– 46, 213n. 43, 214n. 51; recurrent patterns in, 93–94, 223n. 61; shuochang cihua, 45– 47, 93, 213nn. 41, 43, 214n. 50. See also cihua cluster cihua cluster, 47– 48, 49; Judge Bao, 47– 48, 93–94, 96 –97, 223n. 61; Shuihu, 48. See also cihua civil service examinations, 162–164, 167– 168, 173, 177, 236nn. 63, 65, 237n. 71, 238n. 97; competitiveness of, 163; failures in, 164; quota in, 163; suspended in the Yuan period, 19, 163. See also bagu wen; cairen; laolang; men of letters; shuhui; shuhui xianshang Clarissa, 56 Confucius, 13–14, 15, 16, 174, 203n. 36 Conrad, Joseph, 181 “Cui Yanei baiyao zhaoyao,” 115 “Cui Yingying Shangdiao dielianhua ci,” 22 Dai Zong, 73, 86 Dame Wang, 46, 190–191

Da Tang Qin Wang cihua, 44 – 45, 224n. 3 Da Tang Sanzang qujing shihua, 29, 208n. 73, 229n. 75, 231n. 93, 232n. 99 “Da xu” (Greater Preface), 16 Daxue zhijie, 20 Decameron, 2, 151 Defoe, Daniel, 195 DeFrancis, John, 4, 10, 148 dishaxing (Stars of Earthly Fiends), 38 Dong Chao, 69, 71, 93 Dong Jieyuan (fl. 1200), 23, 127, 205n. 51, 235n. 45 Dongjing menghua lu, 29, 30, 208n. 78 Dong Qichang (1555–1636), 2 Dongwu nongzhu ke, 3 Don Quixote, 147 Ducheng jisheng, 31, 208n. 78, 209nn. 81, 82, 234nn. 24, 25 Dundes, Alan, 79, 147 Du Shanfu (1202–1283), 155 Elman, Benjamin A., 236nn. 68, 69, 237n. 71 Eoyang, Eugene, 56, 146, 148, 152 Er pai, 130 expressive theory of Chinese poetics, 15–16 Fang La, 37, 39, 173, 210n. 14 fanqie, 17 Feng Menglong (1574 –1646), 28, 45, 51, 65, 149, 152, 212n. 36 Fielding, Henry, 57 Finnegan, Ruth, 7 Frye, Northrop, 182, 241n. 5 Gao Ru, 3, 102, 105, 226n. 23 Gao Wenxiu, 40, 132 Gejiang douzhi, 43 Genette, Gérard, 181, 196 Gengzhi Zhang Qian ti sha qi, 44 genres in prose, 16, 202n. 28 Gil Blas, 56 gongming, 170–174, 238n. 100, 239n. 107 Gong Shengyu (1222–1304), 38, 132, 154, 209n. 8, 211n. 15. See also “Songjiang sanshiliu zan” Guan Dawang san zhuo hongyi guai, 43 Guan Hanqing (fl. 1279), 19, 35, 205n. 55 Guan Zhang shuang fu Xishu meng, 206n. 58 Guan Yunshi (1286 –1324), 20 Gujin xiaoshuo, 45, 128, 129

Index Guo Xun (1475–1542), 104, 105, 106, 205n. 55, 214n. 55 guzici, 22, 25, 67 Hanan, Patrick, 30, 31, 32, 45, 64, 68, 93, 98, 101, 115, 117, 118, 119, 125–126, 127, 137, 199nn. 1, 6, 208nn. 73, 76, 77, 81, 218n. 100, 220n. 16, 228n. 65, 229n. 74, 241n. 5 Hangzhou (Lin’an), 150, 234nn. 24, 25; centers for public entertainment, 151 Hannas, Williams, 4, 13 Hansen, William, 68–69 Havelock, Eric, 1, 13, 99, 181, 201n. 10 Hayden, George A., 41, 109, 124, 207n. 62, 227n. 42 Hegel, Robert E., 213n. 41, 217n. 80, 236n. 58 Heixuanfeng fujing, 132 Heixuanfeng qiao duan’an, 132, 211n. 21 Heixuanfeng qiao jiaoxue, 132, 211n. 21 Heixuanfeng shuang xiangong, 42, 44 Heixuanfeng zhangyi shucai, 27, 41, 96 Hennessey, William O., 40 He Xin, 107, 227n. 36 historiography, 37, 38, 59, 93, 181–183 Homer, 54, 78, 195–196. See also Homeric epics Homeric epics, 1, 56, 61, 62; censured morally, 57; “echo principle” in, 99; formulary diction in, 66 –67; spirit of vengeance in, 56 –57; “themes” in, 66 – 67. See also Homer; Western narrative tradition homophonic substitutes, 108–109; in bianwen, 112; in Liu Xibi Jinchai ji, 114; in Shuihu zhuan, 111–113, 116; in shuochang cihua, 112; of stylistic features, 120–123, 142, 229nn. 73, 74, 75; in vernacular stories, 116; in zaju, 112. See also linguistic features; stylistic features Honglou meng, 197 “Hongzi” Li Er, 40, 132, 164 Ho, Ping-ti, 236nn. 65, 68, 69 Hrdlickova, Vena, 82 Hsia, C. T., 64, 108, 161, 217nn. 82, 85, 88 huaben, 30, 59, 60, 113, 204n. 42; “dual identity” of, 31; in oral-literary reciprocation, 60, 110; not necessarily storyteller’s script, 31, 60, 218nn. 100, 101; synonymous with cihua, 212n. 36. See also storytelling; vernacular stories

285

“Hua Guan Suo chushen zhuan,” 94 “Hua Heshang,” 38 Huai Lin, 110, 111, 216nn. 68, 71 Huanghua yu, 132 Huang Lin, 225n. 19, 226n. 26, 228n. 43 Huan laomo, 44, 212n. 24 Huanmen zidi cuo lishen, 27–28, 126, 164, 207n. 66, 226n. 30 Hu Shi, 31, 54, 107, 176 Hu Shiying, 165 Hu Xu (1655–1736), 148 Hu Yinglin (1551–1602), 50, 102, 105, 179, 211n. 16, 215n. 60 Idema, Wilt L., 27, 30, 36, 56, 59, 64, 205n. 56, 207n. 62, 218n. 100 Iliad, 57, 58, 60, 78, 195–196. See also Homeric epics Irwin, Richard, 47, 107, 166, 210n. 10, 227n. 36 Iser, Wolfgang, 146 Jameson, Fredric, 54 Jiang Daqi, 3, 102 jiangshi, 29, 30, 31, 39, 49, 67, 151, 208n. 72, 214n. 55, 215n. 57 “Jiang Xingge chonghui zhenzhu shan,” 45 “Jiantie heshang,” 32, 93, 165 jiaoben, 82 Jia Zhongming (1343–?), 40, 102, 164, 224n. 3 Jingchai ji, 27–28, 207n. 66 Jingshi tongyan, 93, 127, 128, 129 Jin ping mei, 2, 3, 197; chapter-ending formulas in, 124; Jin ping mei cihua, 3, 45, 124; linguistic features in, 129–130, 226n. 30; recurrent patterns in, 92, 222n. 45. See also vernacular fiction Jin Shengtan (1608–1661), 9, 51–54, 61, 102, 105, 153, 179, 218n. 102, 238n. 91; “direct repetition” and “partial repetition,” 91; on gewu, 192; hermeneutic strategy of, 54; on linguistic stratification, 188; on qindongxin, 192–193; on recurrent narrative patterns, 90–92, 98–99; on venting indignation, 176 –178; on writing moment of Shuihu zhuan, 53 Ji Yun (1724 –1805), 2 Kang Jinzhi, 40 “Kan pixue danzheng Erlangshen,” 115

286

Index

Kao, Karl S. Y., 241n. 10 Karlgren, Bernhard, 10 Kellogg, Robert, 184, 240n. 1 Kòsaka Jun’ichi, 127, 229nn. 71, 73, 74, 230n. 83 language: Chinese distrust of, 14; inadequacy of, 14 –15; seen by Chan Buddhists, 18. See also speech; writing Lang Ying (1487–?), 102, 105 Lao Huihui donglou jiaofo, 43 laolang, 165, 167, 171, 237n. 87. See also cairen; men of letters; shuhui; shuhui xiansheng Laozi, 14 Latin, 2 Liaison Dangereuses, 56 Lianggong jiu jian, 29, 208n. 73 Liangshan bandits, 8–9, 168, 239n. 105; compared to failed examination candidates, 169–174; “forced to go up Liangshan Mountain,” 168; one hundred and eight chieftains of, 109, 124, 206n. 62, 227n. 41, 239n. 106; ranking of, 42, 134, 173, 239nn. 105, 106; seventy-two minor chieftains of, 38, 41, 109, 133, 212n. 23; thirty-six major chieftains of, 38, 39, 42, 109, 133, 212n. 23; in Zhu Youdun’s plays, 109, 124, 206n. 62. See also Shuihu complex; Shuihu zhuan Liang shi pinghua, 95 Lianhuan ji, 43 Li Dou (fl. 1797), 151 Li Kaixian (1502–1568), 206n. 62, 230n. 88 Li Kui, 73, 74, 75, 86, 91, 96, 132, 133, 142, 153, 168, 188, 189, 190 Li Kui fujing, 41 Lin Chong, 38, 69–70, 80, 86, 87, 91, 93, 95, 125, 132, 142, 156, 168, 170, 171, 173, 176, 230n. 88 Ling Mengchu (1580–1644), 65, 130, 205n. 56, 206 n. 62, 208n. 71 linguistic apathy, 191–192 linguistic empathy, 191–192, 196; for later writers, 193 linguistic features, 7; in early vernacular literature, 129–130, 136 –137, 226n. 30, 230n. 82, 231n. 93, 232nn. 94, 97, 99; frequency ratios; 135–141; shared by Shuihu zhuan and vernacular stories, 125–126, 130–134; in Shuihu zhuan,

106, 131, 135–136, 140–141, 229n. 71, 230nn. 82, 87, 231n. 92, 232nn. 95, 98; in vernacular stories, 125, 126 – 129, 137–139; in Xixiang ji zaju and Yuan-edition zaju compared, 26; in Xixiang ji zhugongdiao, 23–24; in Xixiangji zhugongdiao and Xixiang ji zaju compared, 24 –25. See also stylistic features linguistic mimesis, 194 –196, 197; as “mimesis of words,” 196 Li, Peter, 68, 220n. 12 Li Shangyin (813–858), 187, 242n. 16 literacy, 6, 7; level of, 8; rate of, 148 literary vernacular, 8, 180, 186, 187, 197 Liu Jingting (1587–ca. 1670), 187 Liu, James J. Y., 215n. 65 Liu Shide, 225n. 21 Liushi jia xiaoshuo, 3, 30 Liu Ts’un-yan, 107, 224n. 3, 227n. 36 Liu Xibi Jinchai ji, 113–114, 126, 228n. 54 Liu Xie (fl. 500), 12, 13, 15, 16, 201n. 8, 202n. 28, 242n. 13 Liu Zhiyuan zhugongdiao, 22 Li Wenwei, 40 Li Yu (1611–1689), 33 Li Zhi (1527–1602), 51, 52, 110, 147, 175, 177, 179, 216n. 71 Lord, Albert B., 6, 66, 78, 79, 80, 213n. 43, 219n. 5 Lu gui bu, 40, 43, 164, 205nn. 51, 55, 209n. 91, 235n. 45; Tianyige edition of, 43, 164 Lu gui bu xubian, 40, 43, 102, 209n. 91, 224n. 3 Lu Ji (261–303), 15, 16, 202n. 28 Lu Junyi, 52, 69, 70, 80, 91, 93, 173 Lukacs, Georg, 182 Lunyu, 10, 152, 203n. 36, 241n. 9 Luo Ergang, 222n. 47 Luo Guanzhong, 51, 52, 101–102, 223n. 57, 224nn. 3, 5; possibly failed examination candidate, 166 Luo Ye, 37, 208n. 78 Lu Sheldon Hsiao-peng, 182 Lü Shuxiang, 10–11, 226n. 30, 230n. 83, 231n. 93, 232n. 97 Lu Xun (1881–1936), 51, 103, 106, 107, 241n. 11 Lu Zhishen (Lu Da), 38, 39, 46, 69, 71, 73, 74, 75, 86, 89, 95, 96, 125, 142, 156, 172, 188, 189

Index Mair, Victor, 22, 112, 203n. 32, 204nn. 42, 46, 227n. 40 Maiwangguan Library, 27, 40, 41, 113, 227n. 41 Ma Nancun (Deng Tuo), 153 Mang Zhan Fei da’nao xiangfu yuan, 43 Mao Dun, 167 Mao Zonggang (1632–1709), 51 Masuda Mataru, 218n. 100 Ma, Y. W., 41, 109, 124, 206n. 61, 207n. 62, 212n. 27, 219n. 4, 226n. 23, 227nn. 35, 36, 42, 230nn. 85, 87 McLaren, Anne E., 213nn. 39, 41, 42, 223n. 54 Mei Tsu-lin, 135, 231n. 93, 232n. 94 memory: chunk, 83, 85; encoding and retrieval, 83–84; memory organization, 84; mnemonic devices, 83–90; procedural and prepositional memories, 82; selective, 83; semantic and episodic memories, 82–83. See also narrative sequence; oral transmission; recurrent thematic patterns; story making; storytelling Mencius, 162 Meng liang lu, 151, 164, 208n. 78, 209n. 6, 234n. 25, 237n. 75 Meng Yuanlao, 29, 288nn. 72, 78 Mengzi, 10 men of letters, 8, 161, 162, 179; amid storytellers, 164 –167; compared to Liangshan bandits, 169–174; role in textualization of Shuihu zhuan, 8, 167–174; “venting indignation,” 174 –178. See also cairen; civil service examinations; laolang; shuhui; shuhui xiansheng Ming Chenghua shuocheng cihua congkan, 27, 45, 213n. 40, 223n. 61, 62, 63 Ming Jiajing ben Dong Jieyuan Xixiang ji, 23 Moll Flanders, 56, 195 Mote, Frederick, 148 Muse, 1, 2 Naide Weng, 31, 208n. 78, 209n. 82, 226n. 26, 234nn. 24, 25 nanxi. See xiwen narratability, 180, 184 narrative sequence, 68–81; allosegments in, 80–81; of beating a bully, 73–75; closed sequence, 77, 80; of exileimprisonment, 69–71, 77, 80, 93, 220n. 16; of killing adulterous sister-

287

in-law, 75–77, 93, 220n. 17; open sequence, 78; recurrent on two levels, 78; tavern scenes, 87–90, 158–159; of tragicomic survival, 71–73, 80–81. See also memory; recurrent thematic patterns; story making Nie Gannu, 103, 107, 154, 227n. 36 nihuaben, 65, 219n. 4 Nog oltae ònhae, 20, 203n. 40, 226n. 30 Odyssey, 60, 61, 68. See also Homeric epics Ong, Walter, 6, 58, 145, 146, 180 orality, 1, 2, 5, 9, 21, 50, 59, 61, 101, 145, 179; allied with forces for literary change, 63, 179; “binary typology” with literacy, 7; Chinese-type of, 7; definition of, 6 –7; degree of, 66, 219n. 5; disparage of, 63; “heavy characters” in, 58; influence on wenyan, 17; interaction with writing, 1–2, 6, 7, 9, 20, 109–114, 142, 144, 171, 180, 197, 204n. 45; as locus for linguistic stratification, 186 –187; “primary orality,” 6, 58; “pure” type of, 6 –7; relationship to pinghua, 60; stimulation on vernacular prose, 34. See also memory; oral transmission; story making; storytelling oral-literary transmission, 7, 109–114, 116, 142. See also orality; storytelling; writing oral transmission: encoding and retrieval, 81, 83; retrieval cues, 83, 84. See also memory; orality; story making Origen, 61 Orlando furioso, 58 Ouyang Jian, 160, 222n. 47, 226n. 30, 227n. 38 Owen, Stephen, 16 Pak t’ongsa ònhae, 20, 203n. 40, 226n. 30 Panguo Shuoren, 26 Pan Jinlian, 46, 76, 77, 91, 191 Pan Qiaoyun, 56, 76, 77, 91, 116 Parry-Lord theory, 6 Parry, Milman, 6, 66, 80, 213n. 43, 220n. 10 picaresque novel, 56 Pike, Kenneth, 79 pinghua, 29, 31, 32, 39, 49, 59, 60, 113, 117, 204n. 46, 213n. 41, 214n. 55, 218n. 100, 227n. 40; dating of, 29, 208n. 73; linguistic features in, 126,

288

Index

pinghua (continued) 231n. 93; stylistic patterns in, 118; textual sources of, 30, 60; two written variants for the term, 214n. 55; vernacular portions of, 30, 60 Pinghua Sanguo zhi, 29 Pipa ji, 27–28, 207n. 65 Plaks, Andrew, 6, 59, 61, 87, 92, 98, 104 – 105, 107, 194, 199n. 5, 216nn. 67, 71, 219nn. 102, 103, 221nn. 35, 37, 222n. 45, 223n. 58, 226nn. 28, 29 Plato, 195–196 poetic forms, 16 Polo, Marco, 150 Pope, Alexander, 57 Porphyry, 61 Porter, Deborah, 223 Propp, Vladimir, 79, 99 Prusek, Jaroslav, 151, 186 –187 Pu Songling (1640–1715), 2, 184 Qian-Han shu xuji, 29 Qian Xuantong (1887–1939), 34 Qiguo chunqiu pinghua, 98 Qin bing liu guo pinghua, 29, 208n. 75 “Qingmian Shou,” 37–38 Qingping shantang huaben, 30, 43, 126, 127, 128, 129, 199n. 7, 204n. 48, 209n. 91 Qu Yuan (340?–278? b.c.), 16, 240n. 119 Rawski, Evelyn Sakakida, 148 recurrent thematic patterns: dissolution of force marriage, 95–96; Dong Chao and Xue Ba, 93; exile-imprisonment, 93, 115, 220n. 16, 222n. 47; in Jin ping mei, 92, 222n. 45; Judge Bao, 93–94; killing adulterous sister-in-law, 93, 115, 220n. 17; military contest, 97; parallels in Shuihu zhuan and other narratives, 93–98, 114 –118, 221n. 35, 222nn. 47, 50, 223n. 54; parallels in Shuihu zhuan and zaju, 93, 96; in Sanguo yanyi, 92; selling broadsword, 95; as shared base for story making, 99; in Shuihu zhuan, 6, 68–81, 85–90; in shuochang cihua, 93–94, 223n. 61; transsexual masquerade, 96, 223n. 58; warfare of deployments, 97–98; wedlock formation, 96 – 97; in Xiyou ji, 92–93. See also narrative sequence; story making “Ren xiaozi liexing wei shen,” 115, 116, 117

“Renzong renmu zhuan,” 48, 93, 165, 214n. 51 Richardson, Samuel, 57, 146, 195 Ricoeur, Paul, 145 Roderick Random, 56 Rolston, David, 176, 216nn. 67, 73, 217n. 80, 221n. 40, 222n. 50 Rongyutang commentator, 52, 53, 175, 216n. 71 Roy, David T., 213n. 43 Ruan brothers, 168, 169, 173, 176, 188 Rulin waishi, 167, 197 Sanguo yanyi, 2, 3, 36, 43, 102, 175, 217n. 85, 231n. 91; chapter-ending formulas, 124; duality of origin, 93; linguistic features in, 3, 129, 199n. 6, 230n. 82; recurrent patterns, 92, 221n. 35, 222n. 45; Sanguo zhi tongsu yanyi, 3, 102; sworn-brotherhood in, 94; ties to historiography, 37; warfare of deployments in, 98. See also vernacular fiction Sanguo zaju, 42– 43 Sanhu xiashan, 42 San Sui Pingyao zhuan, 68, 93, 98, 102, 126, 129 “San xianshen Bao Longtu duanyuan,” 117 Sanxia wuyi, 59 San yan, 127, 149 Scholes, Robert, 184, 240n. 1 script: alphabetic, 1, 4, 5; Chinese, 2, 4, 10, 17, 13; Greek, 1; Phoenician syllabary, 1; phonetic aspect of, 17; relationship with literature and the world, 13–14. See also Chinese characters; wen; writing Sha gou ji, 27–28, 207n. 66 Shaoshi shanfang bicong, 215nn. 60, 61 Shemotashi qucong, 27, 126, 130, 226n. 30 Shen bao, 103 Shen Defu (1578–1642), 3, 215n. 58 Shiji, 30, 53, 174 –175, 176, 241n. 11 Shi Jin, 75, 85–86, 173, 189 Shijing, 8, 16, 17; allegorical readings of, 62 Shi Nai’an, 52, 53, 101–104, 176, 255nn. 16, 17; possibly shuhui member, 166 Shishuo xinyu, 17, 203n. 31 “Shitou Sun Li,” 37 “Shiwu guan siyan cheng qiaohuo,” 117 Shi Xiu, 44, 75–77, 86, 116, 142 shuhui, 106, 144, 164 –165, 166, 172; in Shuihu zhuan, 166. See also cairen; laolang; men of letters; shuhui xiansheng

Index shuhui xiansheng, 144, 164 –165, 167, 169, 171; in Shuihu zhuan, 166. See also cairen; laolang; men of letters; shuhui Shuihu cihua, 44 – 49; cluster of, 48; Sun Kaidi’s hypothesis of, 42, 44, 46– 47, 212n. 34. See also cihua; cihua cluster; Shuihu complex; Shuihu zhuan Shuihu complex, 7, 9, 26, 36, 38, 39, 40, 41, 43, 44, 48, 63, 144, 149; generic multiplicity in, 49; prosification in, 49, 81, 119, 124; synthesis in, 49, 63, 81, 94, 109, 119, 124, 125. See also orality; Shuihu cihua; Shuihu stories; Shuihu zaju; Shuihu zhuan; storytelling Shuihu stories, 6, 37–38, 49, 117, 118. See also Shuihu complex; Shuihu zhuan Shuihu zaju, 26, 40– 44, 49, 206n. 61, 211n. 21; different from Shuihu zhuan in plot, 41, 211n. 19; parallels to Shuihu zhuan, 44, 93; stereotyped pattern in, 43– 44, 211nn. 19, 20. See also Shuihu complex; Shuihu zhuan; zaju Shuihu zhuan, 2, 5, 7, 8, 21, 31, 36, 37, 41, 44, 50, 172; banned, 51; belonging to mainstream culture, 9, 179–180; bifurcating attitudes toward, 53–55, 62– 63; championed by Jin Shengtan, 51–53; chapter ending formulas, 123– 124; compared with Jinghua yuan, 239n. 106; as culmination of vernacularization, 5, 34 –35, 66; disparaged, 55–56, 62, 217n. 85; dual nature of origin, 6, 63; individualized speech in, 52–53; ironic and allegorical readings of, 60, 218n. 102; kinship with vernacular stories, 101, 114 –119, 125; “limitations” of, 55; linguistic features of, 125–142, 199n. 6; literal meaning of the title, 8–9; moral condemnation of, 51; narrative patterns in, 6, 68–81, 85–92; oral provenance of, 5– 6, 62– 63, 64, 65– 66, 68, 179, 197, 217n. 85, 219n. 103; relationship to history, 37; representative of early vernacular literature, 5, 9; seen by Ming-Qing scholars, 50–54; as social protest and literary innovation, 8; stylistic features of, 118–124; sworn-brotherhood in, 94 –95; textual sophistication of, 61; urban settings in, 156 –159; “venting indignation,” 174 –178; verbal parallels in other works, 93, 115–117. See also

289

Shuihu cihua; Shuihu complex; Shuihu stories; Shuihu zaju; Shuihu zhuan, editions of; Shuihu zhuan, recensions of; Shuihu zhuan, textualization of; vernacular fiction Shuihu zhuan, editions of: Guanhuatang, 176, 192, 229n. 73, 238n. 91; Jiajing Fragments, 2–3, 104, 226n. 23; Jingben Zhongyi zhuan, 2, 104, 105, 106, 225n. 21, 226n. 28; Pinglin ben, 227n. 35; preWuding versions, 105–106, 107–109, 226n. 26, 228n. 43; Rongyutang, 46, 52, 53, 105, 106, 111, 112, 120, 125, 130, 155, 171–172, 175, 213n. 44, 216n. 71; Shiquge, 105, 215n. 58; Wuding, 104, 105, 106, 225n.22, 226n. 23; Xin’an, 215n. 58; Yingxiong pu, 175; Yuan Wuya, 105, 111, 187, 190; Zengcha ben, 227n. 35; Zhongyi Shuihu zhuan, 2, 3, 104. See also Shuihu zhuan; Shuihu zhuan, recensions of Shuihu zhuan, recensions of: fanben, 5, 6, 7, 32, 59, 65, 101, 103, 105, 106 –108, 109–110, 112–113, 119, 142, 144, 155, 179; jianben, 5, 59, 65, 106 –108; fanben and jianben compared, 106; issue of priority, 106–107, 227nn. 36, 38, 40. See also Shuihu zhuan; Shuihu zhuan, editions of Shuihu zhuan, textualization of, 5, 101, 104, 110, 144, 179, 197; before Zhu Youdun’s time, 126; as cumulative process, 5, 7, 66, 109, 119, 134, 142– 143; with development of vernacular prose, 34, 108; governed by cultural conditions, 114; homophonic substitutes in, 112; involvement of shuhui, 166; Li Kui-Yan Qing stories, 142; Lin Chong story, 132, 142; linguistic changes as context of, 118–143; literary ramifications of, 8, 180–181; Lu Zhishen story, 132–133, 142; parallel to maturation of vernacular prose, 32; role of men of letters, 8, 166–167, 167–174; roles of Shi Nai’an and Luo Guanzhong, 166 –167; Shi Xiu-Yang Xiong story, 130, 142; shuochang cihua, 45– 47; Song Jiang story, 142; synchronous with synthesis of Shuihu complex, 109; as vent of emotions, 172; Wu Song story, 130, 142; in Zhu Youdun’s time, 27, 133. See also orality; Shuihu cihua;

290

Index

Shuihu zhuan (continued) Shuihu complex; Shuihu stories; Shuihu zaju; Shuihu zhuan; writing shuohua, 30, 67 shuoshu, 49 Shuo Tang, 59 Sima Qian (fl. 100 b.c.), 53, 174, 177, 181, 182 Sima Zhao fuduo shoushantai, 43 Si Su, 82 Solomon, Richard, 148 Song Jiang, 39, 40, 41, 42, 46, 56, 69, 70, 71–73, 78, 80, 85, 86, 89, 91, 98, 132, 142, 156, 161, 170, 173–174, 175, 189, 211n. 19; as historical figure, 37, 154 “Song Jiang sanshiliu zan,” 38, 132, 154, 209n. 8; dating of, 210n. 15 “Song Sigong da’nao ‘Jinhun’ Zhang,” 115, 208n. 76 speech, 1–2, 5, 10–11, 20. See also orality speech presentation: in 18th-century English novel, 195; in Homeric epics, 195–196; in Shuihu zhuan, 187–191; in storytelling, 186 –187; in wenyan narrative, 184 –186 St. Augustine, 61 Stahl, Sandra Dolby, 147 story making: composition by themes in, 68; “echo principle” in, 100; the etic and emic in, 79–80; impact from urban audience, 154 –161; linkage incidents in, 85–90; link-chain structure, 87; oral mode of, 6, 65, 68; recurrent at two levels, 78; shared base for, 99; tavern scenes, 87–90; “two-fold quality” of, 99. See also memory; narrative sequence; recurrent patterns storyteller, 5, 8, 29, 124, 144, 234n. 38; “cliché” of, 55, 117; disdain for country people, 155–156; as dominant voice in narrative, 192; dramatizing character’s speech, 186 –187; in Hangzhou, 151; in Yangzhou, 151–152; linguistic empathy with, 193–194; and memory, 81–90; memory organization of, 84; mixed with men of letters, 164 –167; as narrative modulator, 193; nicknames of, 164, 165; objectified in text, 181; relationship with audience, 8, 49, 65, 67– 68, 145–146, 152, 157–158, 161; “storyteller’s manner,” 64, 65; training of, 82; use of promptbooks, 82,

164. See also audience; orality; storytelling storytelling, 108, 123–124, 144, 145, 197, 204n. 46, 209n. 6, 215n. 56; convention of, 92; formulary phrases in, 55, 64; in Hangzhou, 151, 214n. 55; as interface between writing and orality, 20; kinship to popular drama, 186 –187; in Kunming, 82, 154; in oral prose, 29, 49; relationship to pinghua and huaben, 60; relationship to rise of vernacular fiction, 50, 65, 93; in the Southern Song and early Yuan periods, 30–31, 151, 155, 204n. 46, 208n. 72; storytelling scenes in Shuihu zhuan, 153, 155–156, 214n. 55; storytelling scenes in Shuo Yue quanzhuan, 153–154, 156; in the Tang period, 22; textualization from, 29, 67; two categories of, 67; typology of, 30–31, 37–38, 117, 208n. 81, 209n. 6; in urban centers, 149–151; in Yangzhou, 82, 151–152, 154, 187, 215n. 56. See also audience; orality; storyteller stylistic features, 7, 101; in early vernacular literature, 299nn. 74, 75; homophonic substitutes of, 120–123; in Shuihu zhuan; 119–124, 299nn. 71, 72, 73, 74, 75; in Shuihu zhuan and vernacular stories compared, 118–120. See also linguistic features Su Dongpo (1037–1101), 152 Sui Tang yanyi, 95 Sun Kaidi, 42, 44, 46, 47, 107, 193–194, 212n. 34, 241n. 11 Sun Li, 37, 38, 39, 49 Suzhou, 150 sworn-brotherhood, 94 –95 Taihang Mountains, 154 –155, 223n. 54 Taihe zhengyin pu, 205n. 55 tanci, 67 Tang Sanzang xitian qujing, 43 taozhen, 50, 67, 156, 214n. 55 Tao Zongyi (fl. 1368), 151, 235n. 45 Tianbao yishi zhugongdiao, 22–23 Tiandu Waichen, 50, 104, 105, 215n. 58 tiangang xing (Stars of Heavenly Spirits), 38 Tian Rucheng (1503–?), 51, 102, 105, 214n. 55 T’ien, Ju-k’ang, 237n. 74 Tongjian gangmu, 30

Index vernacular fiction, 8, 9, 11, 18, 59, 93, 217n. 80, 218n. 102; compared with the Western novel, 55–56, 194 – 197, 217n. 82; disparaged, 55–56, 217nn. 81, 82, 83, 88; “episodic quality” of, 55; formal features of, 64, 65; “literati novel,” 59; “moral ambiguity” of, 55, 217n. 88; origins of, 9, 36, 50; as polyphonic field, 187; relationship to orality, 9, 59– 60, 65, 197, 217nn. 83, 85; seen by MingQing scholars, 50–54; “simulated context” in, 64 – 65. See also Jin ping mei; Sanguo yanyi; Shuihu zhuan; vernacular prose; vernacular stories; Xiyou ji vernacularity, 21, 32; degree of, 4, 25, 108 vernacularization, 2, 3, 4, 5, 8, 20, 25, 35, 66, 142, 172, 197; as arduous process, 5, 11, 34; and Chan yulu, 18; determined by general cultural conditions, 114; general course of, 21, 48; overlapping with Shuihu tradition, 142; in prose and in verse compared, 25, 32–33; reflected in Zhu Youdun’s plays, 27; slow and prolonged in prose, 33– 4, 108, 204n. 45; in the Yuan period, 20, 203n. 39. See also orality; vernacular prose; writing vernacular prose, 3, 4, 5, 7, 8, 21, 25, 34, 53, 107, 108, 110, 113, 114, 179, 193–194, 197; available for late fiction writers, 31, 65; despised, 33– 4; maturation of, 8, 28–29, 114, 108; maturation parallel to textualization of Shuihu zhuan, 32, 108; referential capacity of, 8; tentative writing of, 32, 60; writing toward, 7; in Xixiang ji zaju, 25–26; in Zhu Youdun’s plays, 27. See also baihua; binbai; vernacularization vernacular stories, 3, 7, 32, 142, 165, 199n. 7; affinities with Shuihu zhuan, 101, 114 – 118, 125; dating of, 30–31, 118–119, 208n. 77; linguistic features compared with Shuihu zhuan, 126 –131, 137–142; stylistic features compared with Shuihu zhuan, 118–124. See also huaben; vernacularization; vernacular prose Wang Guowei (1877–1927), 33, 205n. 55, 207n. 66 Wang Jide (?–1623), 33, 203n. 41

291

Wang Li, 135–136, 232n. 97 Wang Liqi, 225n. 19 Wang Qi (1565–1614), 102 Wang Shaotang, 67– 68, 215n. 57 Wang Shifu (fl. 1280), 24, 25, 26, 205n. 55, 206n. 58 Wanli yehuo bian, 3, 215n. 58 “Wan Xiuniang choubao Shantinger,” 115 Ward, Barbara E., 148, 149, 234n. 21 Watt, Ian, 194 –195; formal realism, 195 wen, 11–14, as cosmic and human patterns, 11–13; of the Dao, 12; polysemy of, 11–12, 13 Wen fu, 15, 16 “Wenjing yuanyang hui,” 127, 204n. 48 Wenxin diaolong, 11, 16, 201n. 8, 242n. 13. wenxue, 13 wenyan (classical Chinese), 2, 4, 5, 14, 53, 63, 93, 118, 148, 149, 179; divergent from speech, 10–11; efficacy of, 11; hybrid in Yuan period, 19–20; impact on literature, 11; incorporeal referential orientation of, 183, 241n. 11; infiltrated by vernacular elements, 11, 17; insulated from living orality, 16 –17; narrative in, 181–186, 192, 241n. 11; as pattern of words, 13; as perfect fit in historiography, 182; in pinghua, 29– 30; political privilege of, 167; as social dialect in vernacular fiction, 197; stylistic elasticity of, 16; temporary erosion in Yuan period, 19–20; in zhiguai and chuanqi, 182–183. See also baihua Wenyi bao, 103 Western narrative tradition: Christian Scriptures, 61– 62, 145; classical epics, 1, 56 –57, 58, 60, 61– 62, 66 –67, 99, 182, 195–196; folklore, 79, 84, 99, 147; formal realism, 195; generic diversity, 56; “heteroglossia,” 196; high mimetic mode, 182; Icelandic saga, 7, 200n. 16; “illustrative narrative,” 184; the novel, 55, 56, 146 –147, 181, 194 –195, 196, 197; picaresque novel, 56; speech presentation, 195–196; vernacularization, 2, 240n. 1. West, Stephen, 112, 205n. 56 Williams, Raymond, 145 writing, 1, 2, 9, 101, 240n. 1, 241n. 3; ancient Chinese notion of, 11–14, 20; approaching living voice in Shuihu

292

Index

writing (continued) zhuan, 53; inadequacy of, 15; interaction with orality, 1–2, 5, 6, 7, 9, 20, 109–114, 142, 144, 171, 180, 197; seen by Chan Buddhists, 18; shapes of, 1; as surrogate of speech, 13, 201n. 10. See also Chinese characters; orality; script Wudai shi pinghua, 29, 95, 126, 208n. 73 Wulin jiushi, 151, 164, 208n. 78, 234n. 25, 237n. 75 Wu Mei (1884 –1939), 33, 103 Wu Song, 38, 39, 44, 49, 52, 56, 69, 70, 71–72, 74, 75–77, 80, 85, 86, 91, 96, 115, 116, 125, 142, 157, 161, 187, 189 Wuwang fa Zhou shu, 29 “Wu Xingzhe,” 38, 132 Wu, Yenna, 240n. 107 Wu Zimu, 208n. 78, 234n. 25 Xian qing ou ji, 33 Xiaojing, 152 Xiaojing zhijie, 20, 231n. 93 xiaoshuo (fiction), 50 xiaoshuo (genre of storytelling), 30–31, 37, 49, 67, 117, 118, 228n. 65, 237n. 75 Xiao Sun tu, 27–28, 130, 164, 226n. 30 Xihu Laoren, 151, 208n. 78, 234nn. 24, 25 Xihu Laoren fansheng lu, 208n. 78, 234nn. 24, 25 Xihu youlan zhiyu, 102 Ximen Qing, 46, 91, 191 Xingshi hengyan, 128, 129, 237n. 87 xiwen, 27, 113; authorship of, 27, 207nn. 65, 66; Yongle dadian texts and late Ming texts compared, 27–28 Xixiang ji (zaju), 24 –26, 137, 205n. 56, 206n. 58, 226n. 30; authorship of, 205n. 55; compared with Xixiang ji zhugongdiao, 24 –25; compared with Yuan-edition zaju, 25–26, 206n. 58. See also Xixiangji zhugongdiao; zaju Xixiangji zhugongdiao, 23–25, 127, 205n. 53, 235n. 45; linguistic features in, 23– 24; linguistic features compared with Xixiang ji zaju, 24 –25. See also Xixiang ji; zhugongdiao Xiyou ji, 2, 3, 36, 43, 95, 231n. 91; chapterending formulas, 124; linguistic features in, 129, 230n. 82; recurrent patterns in, 92–93, 221n. 35, 222n. 45, 223n. 58; Shidetang edition of, 3;

sworn-brotherhood in, 94. See also vernacular fiction Xiyou zaju, 42– 43. See also zaju Xuanhe yishi, 6, 29, 39, 44, 49, 95, 110, 119, 155, 206n. 62, 208n. 73, 210nn. 10, 12, 14, 15, 231n. 90, 232n. 93; dating of, 210nn. 10, 15; differences from Shuihu zhuan, 39; like pinghua texts, 39– 40; Shuihu segment incorporated wholesale, 40; textual parallels of, 40. See also Shuihu complex; Shuihu zhuan Xue Ba, 69, 71, 93 “Xue Rengui zheng Liao shilue,” 97, 223n. 63, 231n. 93 Xu Heng (1209–1281), 20 Xunzi, 10 Xu Shuofang, 222n. 45, 224n. 5 Xu Wenxian tongkao, 102 Yan Dunyi, 31, 42, 103, 165, 166, 212n. 31, 224n. 9, 227n. 36, 238n. 94 “Yang Wen ‘Lanluhu’ zhuan,” 30, 115 Yang Xianzhi, 40, 132 Yang Xiong, 52, 76, 77, 86, 142 Yang Zhi, 38, 39, 49, 72, 89, 95, 97, 105, 155, 156, 168, 170, 171, 172, 173, 187, 188 Yangzhou huafang lu, 151–152 Yan Qing, 70, 71, 74, 115, 132, 142, 153, 173 Yan Qing puyu, 44 Yan Qing she yan, 211n. 21 Yeshiyuan Library, 103 Ye Zhou (?–1625), 51, 216nn. 68, 71 Yijing, 12–13, 14, 15, 16, 201n. 9 “Yingying zhuan,” 22 Yongle dadian, 28, 113, 126, 130, 164, 205n. 56, 207n. 66, 223n. 63, 226n. 30, 228n. 54 yu (speech utterance), 11, 14, 193. See also orality; speech yuanben, 24 Yuan Hongdao (1568–1610), 3, 149, 179 Yuankanben zaju sanshi zhong, 25. See also zaju Yuanqu xuan, 26, 43, 113, 226n. 30. See also zaju Yuan Wuya (fl. 1614), 105 Yuan Zhen (779–831), 22, 23 Yuan Zhongdao (1570–1623), 3 Yue Yi tu Qi qi guo chunqiu houji, 29 yulu, Chan Buddhist, 17, 19, 203n. 36, 231n.

Index 93; being esoteric, 18–19; dating of, 203n. 33; impact on vernacular fiction, 18; as jargon loaded, 34; limited readership, 19. See also Chan Buddhism yulu, Neo-Confucian, 19, 203n. 37, 231n. 93; dating of, 19; influence from Chan yulu, 203n. 36; language accessible, 19 Yushi mingyan, 126, 128, 129, 212n. 36, 237n. 87 yutiwen, 10–11 zaju, 20, 24, 25, 113, 132, 209n. 91; homophonic substitutes in, 112; Ming editions, 26; Yuan editions, 25, 108, 113, 126, 127, 206n. 57, 231n. 93; Yuan texts compared with Xixiang ji zaju, 25–26 Zang Maoxun (?–1621), 26, 33, 82, 113, 226n. 30, 228n. 53 “Zaojiao lin dawang jiaxing,” 93, 115 Zhang Guoguang, 225n. 21, 227n. 34 Zhang Peiheng, 208n. 73, 225n. 16, 229n. 70 Zhang Shangde, 3 Zhang Shuye, 37 “Zhang Wengui zhuan,” 96 Zhang Xie zhuangyuan, 27–28, 126, 130, 164, 205n. 50, 226n. 30 Zhang Zhupo (1670–1698), 51 Zhao, Henry, Y. H., 66, 200n. 17, 220n. 6 Zhao, Jingshen, 207n. 65, 213n. 41, 235n. 45 Zhao Lingzhi (fl. 1110), 22

293

Zhedanr Wu Song da hu, 132 Zheng bao’en, 44 Zheng Zhenduo (1898–1958), 2, 26, 44, 107, 205n. 53, 226n. 23 zhiguai, 182–183 Zhishuo Daxue yaolue, 20 Zhong Sicheng (fl. 1321), 40, 205nn. 51, 55, 235n. 45 Zhongyong zhijie, 20, 231n. 93 Zhongyuan yinyun, 24 Zhou Deqing (fl. 1324), 24 Zhou Lianggong (1612–1672), 103, 216n. 68 Zhou Mi, 151, 164, 208n. 78, 209n. 8, 234n. 25 Zhou shi pinghua, 95 Zhuangzi, 14 –15 Zhuangzi, 10 zhugongdiao, 22–23, 25, 29, 45, 67, 204n. 49, 205n. 50, 213n. 42, 235n. 45 Zhu Xi (1130–1200), 19, 35 Zhu Youdun (1379–1439), 27, 41, 96, 109, 124, 126, 130, 132, 133, 136, 165, 205n. 56, 206nn. 61, 62, 226n. 30, 230n. 88 Zhu Yuanzhang (Hongwu, r. 1368–1398), 27, 150, 236n. 63 Zhuzi yulei, 19, 231n. 93, 232n. 99 Zizhi tongjian, 30 Zuiweng tanlu, 30, 31, 37, 39, 117, 118, 119, 132, 208nn. 78, 81, 209n. 6

About the Author Liangyan Ge received his Ph.D. in Comparative Literature from Indiana University and is now an assistant professor of Chinese at University of Notre Dame. His other publications include articles in Chinese Literature: Essays, Articles, Reviews; Tamkang Review; The Comparatist; Comparative Civilizations Review; and Paragraph: A Journal of Modern Critical Theory.