The Sing-song Girls of Shanghai, a novel of manners and romance written in late-nineteenth-century China, captures the golden age of Shanghai’s high-class brothels. While the novel’s main setting is the expensive sing-song houses in the International Settlement and its chief protagonists are the courtesans and their clients, the networks of relationships it portrays give the reader a richly textured picture of Shanghai’s Chinese society. Featuring some 120 characters, it is a huge and finely woven tapestry depicting a world that was at once alluring and dangerous, narrated by someone familiar with its workings.
Aside from the interest its subject matter may generate, The Sing-song Girls of Shanghai also stands as a landmark of literary innovation. First, it is an unsentimental and realistic account whose narrative method and structure differ significantly from traditional zhanghui, or episodic, fiction and that ends with a distinctly modern touch. Second, the dialogue of the whole novel is written in the Wu dialect,1 the main language used in the brothels, making it the most substantial literary work in nonstandard Chinese and a major experiment in evoking authentic voices in fiction. In terms of familiarity, on the other hand, the author clearly states that he derived his narrative technique from the Qing dynasty novel Ruilin waishi (The Scholars). He also borrowed from another masterpiece of Qing dynasty fiction, Honglou meng (The Story of the Stone), both in characterization and setting, and from Tang dynasty stories about courtesans and their often faithless lovers. Exemplifying the evolution of Chinese fiction, these properties obviously make for an interesting work.
Despite its attraction for a small number of aficionados, however, The Sing-song Girls of Shanghai never achieved great popularity. The fact that the dialogue is not easily accessible has been seen as a major obstacle, but an even more important barrier must have been the rapid decline of the brothel world in the twentieth century, when the glamorous and alluring soon became outdated and seedy. The sing-song house, with its special rules and rituals, was a familiar scene to the author Han Bangqing and his contemporary readers, but to twentieth-century readers used to an ever-quickening pace of life, it is part of an alien and distant past. Though readers of the novel in translation are shielded from the problem of dialect, they are at least as unfamiliar with the novel’s world. I hope that my essay “The World of the Shanghai Courtesans” will bridge some of the knowledge gaps.
About This Translation
One the most famous aficionados of The Sing-song Girls of Shanghai was the Chinese writer Eileen Chang (1920–1995). Her belief that it deserved a wider audience made her begin work on a Mandarin version as well as an English translation some three decades ago. Though her standard Mandarin version of the novel has been available to readers for over two decades (Taipei: Crown, 1981), except for two chapters, her English translation was never published.2 After Chang’s death, her literary executors, Stephen and Mae Soong, donated her manuscripts to the University of Southern California. Discovered among the various papers in this collection were different versions of The Sing-song Girls of Shanghai in translation. Inspection revealed that these were rough drafts in which personal and place names showed no consistency, romanization followed no particular system, and occasionally a paragraph or a page was missing. The biggest problems, however, were those of language, cadence, and style. Chang seemed to have opted for an almost word-for-word approach, with the translation often following Chinese—rather than English—syntax. The translation of wordplay, idioms, and literary clichés also proved to be a major hurdle.
When Columbia University Press first approached me in 2001 about making Chang’s draft publishable, I was more than a little reluctant. What finally persuaded me to undertake this task was my role as a translation historian. Collaborative translation has been a significant part of China’s cultural tradition for some two thousand years, starting with sutra translations in the mid-second century. Since one of the duties I envisaged for myself as editor of Renditions was that of advocating teamwork, it seemed that I should put into practice what I preached. Thus began twenty-two months of extensive revision. In terms of translation approach, use of language and prose rhythm, as well as choice of names for the characters, roughly 60 percent of this published text represents my work rather than Chang’s. Her assessment of the strengths and weaknesses of this novel cannot be faulted, however, so the structural changes made in this translation—in the form of deletion and minor rewriting—are mostly based on choices made by Eileen Chang.
The major deletions center on occasions when some of the brothel clients flaunt their literary talent.3 The most important reason for the decision to omit these is that refined poetic composition was not Han Bangqing’s strong suit, and so the pieces attributed to the “talented young men” in the novel fall far short of the standards of those with true talent.4 Other than such literary scenes, the only major deletions occurs at the beginning and the end of the novel, as explained in Eileen Chang’s preface.
Chang’s approach to the translation of personal names is one that may arouse objection: she decided against transliteration in favor of semantic translation of all personal names. While this approach makes perfect sense in the case of the prostitutes—whose names were given to them by their owners when they entered the profession and were chosen to convey various aspects of feminine allure—it may be more problematic in relation to other characters. The primary objection is that such names may cause each character to be perceived as a representative of a certain human trait, suggesting a kind of allegory similar to The Pilgrim’s Progress. The obvious and simple solution is to adopt a bifurcated approach (as David Hawkes did in The Story of the Stone), with semantic translation for names of prostitutes and straightforward romanization for other characters. After careful consideration, however, I decided that Chang’s decision was backed by sound reason.5 Chinese personal names in general carry a much more obvious semantic and cultural load than English ones: to put it in the simplest way, they reflect the background, good wishes, and ambitions of the parents or name givers. In the case of fiction, they are of course one of the easiest means of characterization. Indeed, the majority of names in this novel serve more than the simple purpose of identification; they also tell us something about the background or personality of the characters concerned. Unlike what occurs in The Pilgrim’s Progress, however, names are not used uniformly in this novel. Cases such as Simplicity Zhou and Juvenity Zhang are direct and obvious: they are young and unsophisticated men whose personalities and lack of experience are central to their roles in the novel. In cases such as Prosperity Luo and Dragon Ma,6 their names mainly indicate their circumstances in life, while those of Devotion Yin and Second Bai Gao are indicative of certain personality traits. Then there are the somewhat ironic cases, notably Constance (who despite her apparent good nature is not constant) and Benevolence (who is good to his friends but fails to look out for his kin). Finally, some names characterize through literary reference: Green Phoenix bears more than a passing resemblance to another “phoenix”—Wang Xifeng in Honglou meng—and Water Blossom reminds us of Lin Daiyu, who is represented by the water lily in the same novel. As long as readers bear in mind that this is a Chinese novel, where the webs of relationships—cultural, semantic, and social—are different from English ones, the translation of personal names should be a bonus rather than a distraction. After all, besides the reasons stated above, translated names—unlike many romanized according to pinyin—are pronounceable and easier to remember.
In translating place and street names I have relied primarily on late-nineteenth- and early-twentieth-century English maps of Shanghai. Names of small lanes and alleyways—the sites of the brothels—are not recorded in the English maps I consulted, however. With reference to these, my decision was again to translate rather than transliterate, for the same reasons as those pertaining to personal names.
It is the practice of most academic translations to provide footnotes on all sorts of minor details in the text. Readers will find that, with very few exceptions, in this translation footnotes are only given where they make a direct contribution to the understanding of characterization and plot. The reason is simple: The Sing-song Girls of Shanghai is a novel, so this translation is meant to function first and foremost as a novel. The advantage of providing peripheral information has to be weighed against the aesthetic disruption footnotes entail. Scholars with a good appreciation of translation discourse know that footnotes are a cumbersome means of conveying factual information and are not conducive to literary appreciation. A far better way of giving essential information on social and cultural background is to do so comprehensively in a separate text—an approach similar to providing a new visitor with a city guide and a street map for orientation. I would therefore urge readers of this novel to read the essay “The World of the Shanghai Courtesans” to get their social and cultural bearings. The street map of old Shanghai is provided for the same reason. Readers who have visited the city—or who care to check a contemporary map—will find that they can easily follow the characters through its streets and alleyways.
Like most episodic fiction written by men educated in the old style, this novel contains occasional passing references to older works. In terms of frequency, however, mention of popular opera ranks the highest. Since most such references are just an indirect reflection on the popular culture of the day, footnotes are only provided in cases where the information contributes to an overall understanding of events and characters. There are, of course, various sourcebooks on Beijing opera and Kun opera that detail their storylines and sources, and the interested reader may want to consult these. With reference to literary allusions, annotation is only used when the primary meaning would be obscure if the source of the allusion were not revealed. Where the allusion serves only a secondary or peripheral function in terms of message communication, the preferred approach here is manipulation of style and register rather than annotation. After all, allusions are used at all levels of discourse in all literary cultures without distinctive markers (such as footnotes) tagged to them. Imagine the tears of boredom that would drown us all if every time someone said “lend me your ears,” we had to endure an aside stating: “Julius Caesar, act 3, scene 2, line 79” or the unnecessary complication that would ensue if every “good-bye” were followed by the tag “a corrupt form of ‘God be with you.’”
Acknowledgments
After twenty-two months of revision, my impression of The Sing-song Girls of Shanghai changed substantially. With the realization that this is a novel meant for repeated reading, what began as a duty eventually became something of a pleasure. This transformation would not have been possible without the following people: Professor Joseph Lau of Lingnan University and Professor David Wang of Columbia University, who cornered me into accepting the project; Professor Yuan Jin of Shanghai University, who first showed me traces of old Shanghai in 1997 and who later obtained for me various sourcebooks on the late Qing and early Republican eras; Mrs. Mae Soong, who assured me that my extensive rewriting of Chang’s draft would be a favor rather than a sacrilege; my research assistants, Kaman Chan, Audrey Heijns, and Maggie Leung who undertook the hefty task of proofreading three revisions of the translation; and my secretary, Alena Chow, who helped me with word processing and compiling the “Cast of Major Characters.” The person to whom I owe the greatest debt is my husband, David Pollard, who is my most reliable source of information on matters related to realia as well as language and whose appreciation of this novel contributed significantly to my own reevaluation of it.
Editions of Haishang hua Used
Haishang hua liezhuan. 1926. Reprint, Taipei: Tianyi chubanshe, 1974.
Haishang hua. Ed. Wang Yuanfang. Shanghai: Yadong tushuguan, 1935.
Haishang hua. Ed. Eileen Chang. Taipei: Crown, 1983.
Haishang hua. Ed. Jiang Hanchun. Taipei: Sanmin chubanshe, 1998.
Notes
1. Eileen Chang noted that the only exception is one line spoken by Lai the Turtle, probably to show that Lai is a northerner.
2. Chapters 1 and 2 of her translation were published in Renditions, nos. 17 and 18 (1982). It seems that Chang did look for a publisher for the full novel. The manuscript, however, never reached the stage where it was publishable.
3. As a result of such deletion, the titles of chapters 10 and 33 have been rewritten.
4. Liu Fu, who was vehement in his condemnation of the introduction of the Rustic Retreat group of characters into this novel, was of the same opinion. See Liu Fu, “Du Haishang hua liezhuan” (On reading Haishang hua liezhuan), in Haishang hua, ed. Wang Yuanfang (Shanghai: Yadong tushuguan, 1935), pp. 21–24.
5. Many of the personal and place names that Chang used proved problematic and have been replaced.
6. Dragon Ma’s personal name in Chinese—Longchi, or dragon pond—suggests that he is a man of great talent (dragon) trapped in a small space (pond), that is, someone with unfulfilled ambition.