"The Search for Evidence from China:

Qing Learning ²M ¾Ç and Kôshôgaku ¦Ò ÃÒ ¾Ç in Tokugawa Japan"

Benjamin A. Elman

Mellon Visiting Professor, Institute for Advanced Study, 1999-2001

Draft: NOT FOR CITATION

1. Introduction

In past research, I reconstructed the historical importance of "evidential research studies" (kaozhengxue ¦Ò ÃÒ ¾Ç, kôshôgaku in Japanese), a movement in classical studies which flourished in late imperial China.[1] This philological turn by seventeenth and eighteenth century Chinese literati-scholars (shidaifu ¤h ¤j ¤Ò) represented a new trend in modern Chinese intellectual history, a development akin to the role of philology in the emergence of legal, biblical, and historical fields of research in Renaissance and Enlightenment Europe.[2] I have described the intellectual community in which evidential research during the Qing ²Mdynasty (1644-1911) took shape and also delineated the epistemological transformation in scholarly discourse that ensued. Most members of this academic community lived in the urban centers of the Yangzi River delta and were bound together by associations and institutions for the propagation of a new form of classical studies. A consensus of ideas about how to find and verify such knowledge was the result. The institutional and intellectual context for the emergence of precise textual scholarship marked an initial stage in the "professionalization" of literati scholarship in late imperial China. Through the concentrated efforts of trained specialists who studied the literati canon made up primarily of the Five Classics (Wujing ¤ ¸g) and Four Books (Sishu ¥| ®Ñ), a semi-autonomous subsystem of elite Chinese society emerged and evolved within its own rubrics of status from 1700 to 1850. This community of literati-scholars were bound to the past by building on accumulated classical scholarship since the early empire of the Han º~ dynasties (206 B.C. - A.D. 220). They employed exacting empirical procedures of inquiry in their careful scrutiny of the Chinese classical heritage.[3]

Their appeal to philological methods made possible a better understanding of the scholarly intricacies of Han dynasty classical learning, known as Hanxue º~ ¾Ç(Han Learning), and its subsequent overthrow during the Song §º (960-1280), Yuan ¤¸ (1280-1368), and Ming ©ú (1368-1644) dynasties by a later form of classical learning known as Daoxue ¹D ¾Ç (Learning of the Way; i.e., what is today called "Neo-Confucianism"), which in Qing times was commonly referred to as Songxue §º ¾Ç (Song Learning).[4] A corrosive form of criticism emerged that by the late nineteenth century would exceed the intellectual boundaries that early Qing literati-scholars had taken for granted. In the process, the New Text (Jinwen ¤µ ¤å) school of Confucianism, which had its roots in Changzhou ±` ¦{ ©² prefectural traditions of learning associated with the distinguished Zhuang ²ø and Liu ¼B lineages there, eventually threatened the hegemony of even "Han Learning," which had reaffirmed an Old Text ¥j ¤å textual tradition that dated from the early empire.[5]

Search for authentic classical and historical texts in late imperial China exercized the critical minds of literati-scholars, as in Renaissance and Enlightenment Europe, not only by what scholars found, but also by the way it was found. Critical methods took on their own autonomy, and the result was the historicization of the official classical canon. Evidential scholars made verification a central concern for the emerging empirical theory of knowledge they advocated, namely "to search truth from facts" (shishi qiushi ¹ê¨Æ ¨D ¬O). This program placed proof and verification at the center of the organization and analysis of the classical tradition.

The seventeenth-century pioneering formation of evidential studies by Gu Yanwu ÅU ª¢ ªZ (1613-82) and Yan Roju ÀF Y Ú¡ (1636-1704) was continued in the eighteenth and nineteenth century philological research of Dai Zhen À¹ ¾_ (1723-77), Qian Daxin ¿ú ¤j ©ý(1728-1804), Duan Yucai ¬q ¥É µô (1735-1815), Wang Niansun ¤ý ©À ®] (1744-1832), Wang Yinzhi ¤ý ¤Þ ¤§ (1766-1834), Jiao Xun µJ ´` (1763-1820), and Ruan Yuan ¨¿ ¤¸ (1764-1849). Philological studies developed and evolved during the eighteenth and nineteenth centuries because their published works were parts of a dynamic classical research enterprise whose goals were not "scientific" or "objective" per se, but instead were tied to a new literati commitment to use the language of the ancient Classics as an impartial means to recapture the ideas and intentions of the sage-kings of antiquity. Even if they were scholarly iconoclasts in their own time, they still were firmly conservative in their social beliefs and commitments.[6]

Hamaguchi Fujiô's recent analysis of the exact steps forward in textual research made by evidential research scholars during the eighteenth and nineteenth centuries describes how philological studies developed and evolved during the Qing dynasty. As Hamaguchi shows, Qing dynasty evidential scholars such as Dai Zhen had in mind a systematic research agenda that built on textual studies to reconstruct the meaning (yiyin qiuyi ¥Hµ¨D¸q) of Chinese words. Later Wang Niansun, and his son Yinzhi, extended Tai's approach and attempted to use the "meanings" of Chinese words as a method to reconstruct the "intentions" of the sages, the farsighted authors of those words. Moreover, technical phonology when applied to the study of the history of the classical language reached unprecedented precision and exactness. To achieve this end, evidential scholars chose philological means, principally the application of phonology (guyinxue ¥j µ ¾Ç), paleography (wenzixue ¤å ¦r ¾Ç), and etymology (xunguxue °V µþ ¾Ç), to study the Classics.[7]

Such classical trends in Qing China spilled over to Yi Korea (1392-1910) and Tokugawa Japan (1600-1867). To some degree, the commercial and tribute exchanges of books and knowledge between China, Japan, and Korea in the seventeenth and eighteenth century marked the emergence, before the coming of the western powers, of an East Asian community of textual scholars who specialized in empirical research and philological studies of the Chinese Classics. Other papers in this conference volume by Osamu Ôba and Laura Hess make clear that the Chinese presence in the Nagasaki trade, after the Manchu conquest of China was secure in the 1680s, was considerable and that among the important commodities in that trade were the recent books published in China that Japanese scholars and shôguns desired and rare classical texts long since lost in China but still available in Japan which Chinese traders with scholarly interests sought.[8]

In the late eighteenth century, in particular, Japanese scholars interested in Chinese classical studies learned and adapted the evidential research techniques pioneered by Qing literati. Sometimes this transmission occurred through Korea's more frequent contact with the Qing court via tribute missions sent to Beijing, then also called Yanjing ¿P ¸g.[9] I will examine in the pages that follow the precise role of evidential research, known as kôshôgaku in eighteenth and nineteenth century Japan, in the writings one of the leaders of the "Eclectic School" of Tokugawa Confucianism.

2. Ôta Kinjo ¤Ó ¥Ð ÀA «° on Chinese Classicial Learning

Writing in the early part of the nineteenth century, Ôta Kinjo (1765-1825), a student of distinguished members of the "Eclectic School" (Setchû gakuha §é °J ¾Ç ¬£) of Chinese classical learning, such as Minakawa Kien ¬Ò ¤t ²N ¶é (1734-1807) and Yamamoto Hokuzan ¤s ¥»¥_ ¤s (1752-1812), was clearly aware of his role in the development of kôshôgaku in the Tokugawa era. Both Minakawa and Yamamoto, along with Inoue Kinga ¤« ¤W ª÷ ®o (1732-84), were considered by most Tokugawa scholars as the founders of the Eclectic School, and hence the transmitters of eighteenth century Qing classical studies to Japan.[10]

When Ôta Kinjo completed his remarkable study entitled Kyûkeitan ¤E ¸g ½Í (Discussion of the Nine Classics) in 1804 (published in 1815), he included a "Prolegomenon" ("Sôron" Á` ½×) in which he discussed the history of classical studies in China and Japan. In many ways, Ôta's account allows us to see how deeply kôshôgaku had already penetrated scholarly debate in late eighteenth and early nineteenth century Japan, despite the impact of the Kansei prohibition of heterodoxy (Kansei igaku kin ¼e ¬F ²§ ¾Ç ¸T), which since 1790 had forcefully placed the shogunate for the first time in direct, educational support of the Chinese literati tradition derived from Zhu Xi's ¦¶ ³ß (1130-1200) "Song Learning" and in direct opposition to the "eclectics," who were accused of threatening the doctrinal purity of Zhu Xi's "correct learning" (seigaku ¥¿ ¾Ç).[11]

At the outset of his study, Ôta noted that since antiquity there had been three major developments in Chinese classical learning. These he described as: 1) Kangaku º~ ¾Ç (Han Learning), i.e., the classical learning of scholars during the Han dynasties; 2) Sôgaku §º ¾Ç (Song Learning), i.e., the classical studies of Song dynasty literati-scholars; and 3) Shingaku ²M ¾Ç (Qing Learning), i.e., the classical scholarship of Qing dynasty literati who were roughly contemporaries of the Tokugawa era in Japan. According to Ôta, Han Learning scholars in early China had been expert in etymological research known as kunko °V µþ (xungu in Chinese, lit., "analysis of dictionary glosses"). Song followers of Zhu Xi had instead stressed moral philosophy known as giri ¸q ²z (yili in Chinese, lit., "meanings and principles"), whereas Qing Learning scholars were adept at evidential research, which Ôta referred to as kôshôgaku (lit., "search for evidence"). Although there had been minor variations in its intellectual trajectory, Han Learning, according to Ôta, had been dominant in China through the Tang ð dynasty (618-906). Song Learning had achieved classical dominance during the Song and Ming dynasties. In Ôta's view, Qing Learning had adapted elements of both Han and Song Learning, but its classical contributions lay in its precise empirical procedures of inquiry for studying the Classics. For Ôta, Qing Learning and its stress on evidential research was the dominant classical tradition in China during the seventeenth and eighteenth centuries.[12]

Ôta Kinjo then analyzed the strong and weak points of both Han and Song Learning in China. The tradition of classical scholia (chûso ª` ²¨) prepared by Han and Tang dynasty scholars were, according to Ôta, the starting point for classical exegesis: "Those who study the sacred Classics should begin at this point." Nevertheless, Ôta was forced to admit that Han classicists, however, important their research on the Classics, were not totally reliable. Too much of the ancient canon associated with Han scholars (Ju ¾§) had been falsified and forged after the fall of tha Han in A.D. 220 and before the Sui ¶¦ (581-618) and Tang reunification. A spirit of scepticism and questioning was needed, according to Ôta, to ferret out the true from the false: "One believes what is reliable but not what is unreliable. This is what is called superior learning."[13]

Ôta was very aware of the strengths of Song Learning, however. Principally, he stressed the official orthodoxy of imperial China since the Yuan and Ming dynasties, namely the Cheng-Zhu (Cheng Yi µ{ À[ , 1033-1107, and Zhu Xi) school of classical learning: "In their writings thay have approached the intentions of the sages, thereby going well beyond earlier literati-scholars. Therefore, to expound the sacred Classics and clarify the meaning of the Way (dô ¹D ), one should begin with them." At the same time, however, Ôta was quick to point out the errors of the Cheng-Zhu school: "The learning of Masters Cheng [Yi] and Zhu [Xi] has been contaminated with Buddhism and Daoism. This is their greatest flaw. The reason why I do not dare believe and follow them totally is because of such flaws."[14]

Ôta then documented the Buddhist notions that had penetrated Chinese classical learning during the Song and Ming dynasties. Buddhism and Cheng-Zhu Learning had been hopelessly mixed together. To remedy this defect, Ôta advocated an educational program of careful selectivity and purification: "The theories of Cheng [Yi] and Zhu [Xi] have lapsed into Buddhism and Daoism. This is the limitation in their scholarship. One must remove these flaws and build on their strengths. In this way the purity [of their teachings] can be restored."[15]

In a similar way, Ôta saw in Wang Yangming's ¤ý ¶§ ©ú (Shou-ren ¦u ¤¯, 1472-1529) Ming dynasty school of "mind" (shingaku ¤ß ¾Ç , lit., "learning of the heart and mind") the direct impact of Buddhist doctrines. Although Wang Yangming's actions were true to the classical heritage of Ju learning (Jugaku ¾§ ¾Ç), according to Ôta, his theories were very close to Chan (Zen ÁI) Buddhist teachings. Such impurities in Wang's school threatened to distort irreparably the purity of the classical tradition.[16]

Finally, Ôta took up the strengths and weaknesses of contemporary Qing Learning in China:

The sages perished two thousand years ago. The intentions they bequeathed to us survive only in words and phrases. Therefore, if we are not clear about the written graphs in each sentence, we will be unable to grasp the marvelous intentions of the sages. The study of written graphs and sentences through the use of evidential research is what Qing scholars are adept at.

Qing Learning, according to Ôta, supplied the linguistic means to recover the pristine doctrines of the sages of antiquity. For Ôta, Qing style kôshôgaku had developed a research methodology that enabled Qing literati-scholars to reconstruct the unadulterated truths of the classical era, before ancient classical learning had been sullied with Daoist and Buddhist doctrines. A bridge could be thrown across the era of Cheng-Zhu Song Learning, and the interrupted transmission of ancient wisdom could be recovered. Ôta noted that in comparison with their Song predecessors, the classical research of Qing scholars was vastly superior: "Getting one's hands on the work of a Ming scholar in 100 "chapters" (kan ¨÷, lit., "a rolled up volume") does not compare with even a single chapter from a Qing scholar." Clearly, Ôta was one of those who watched out for Qing editions that came to Japan as part of the Nagasaki trade with Zhejiang ®ý ¦¿ province in China, particularly via the commercially vital entrepot of Ningbo ¹ç ªi .[17]

It is interesting that in eighteenth century Qing China, classical scholars there routinely associated evidential research with their renewed interest in "Han Learning." They turned to Han classical studies because the latter were closer in time to the composition of the Classics and were thereby more likely to reveal the authentic meanings conveyed in the Classics. Qing classicists were in effect calling into question the dynasty's orthodox ideology, which Manchu rulers, following the lead of early Ming emperors, enshrined in dynastic schools, civil service examinations, and official rhetoric.[18]

Similarly, Ôta Kinjo in Tokugawa Japan was impugning the purity of the Cheng-Zhu orthodoxy that under the leadership of Matsudaira Sadanobu ªQ ¥ ©w «H (1758-1829), councilor of state from 1787-1793, was also declared as the educational orthodoxy of the Tokugawa shogunate during the Kansei era, 1789-1800.[19] At the same time, however, Ôta was balanced enough in his analysis of the differences between Song and Qing Learning to emphasize important areas of moral concern that the former had stressed but the latter had overlooked. As an "eclectic," Ôta was especially critical of Qing classicists for their fixation on philological issues, which had resulted in their missing the noble meaning of classical doctrines. In this regard, Ôta noted: "Meanings and principles [giri, i.e., "moral teachings"] are the root. Evidential studies are the branches." By overemphasizing the latter, Qing scholars has misrepresented the ethical vision articulated in the ancient Classics.[20]

In the early nineteenth century, many Qing dynasty literati were also caught up in the Han versus Song Learning debate, and some called for a comprehensive classical synthesis. Chen Genghuan ³¯ ©° ·Ø (1757-1820), a Fujian ºÖ «Ø literatus, proposed turning Han classicism and Song theory into complementary standards for policy questions on the civil examinations. Since the 1787-93 civil examination reforms, Qing examiners had frequently prepared policy questions (ce µ¦) reflecting either Han or Song Learning, but there had been little effort at synthesis. Similarly, Wang Tingzhen ¤ý §Ê ¬Ã (1757-1827), while serving as a Hanlin academician and examiner, called for 8-legged essays that would reflect both Cheng-Zhu learning and Han Learning.[21]

Others such as Chen Shouqi ³¯ ¹Ø ¸R (1771-1834), a Fujian scholar with Han Learning sympathies, when serving as one the two chief examiners for the 1807 Honan provincial examination nevertheless, made clear in the afterword for the official report that Han classical schools and Song Learning together were the foundations of classical models for governance.[22] Chang Haishan ±i ®ü ¬À (1782-1821), on the other hand, favored Song Learning, but he admitted in his writings on the civil examinations that the achievements of both traditions were significant. Hu Peihui J °ö æø (1782-1849) from Anhui, an 1819 jinshi ¶i ¤h who rose to high office, openly called on literati to adopt Han-Song syncretism as a means to overcome the battle lines between them.[23]

The debate over eclecticism versus sectarianism was a common characteristic of both Qing literati and Tokugawa classical scholars in the early nineteenth century. The difference was the political context. During the Qianlong °® ¶© reign (1736-95), the Qing dynasty had successfully balanced the demands of advocates of Han Learning and the competing views of champions of Song Learning. During the Tokugawa Kansei reign period of 1789-1800, however, the shogunate had become a key supporter of Song Learning over its rivals, leaving the eclectic scholars on the defensive and subject to the charge of official heterodoxy.

3. Ôta Kinjo ¤Ó ¥Ð ÀA «° on Tokugawa Classicial Learning

Following his discussion of the three major streams in Chinese classical learning, Ôta Kinjo then took up classical learning in the Chinese tradition as it was practiced in Tokugawa times. Ôta saw the latter traditions in light of his understanding of the history of classical studies in China. From Ôta's point of view, Tokugawa classical scholars were building upon the learning and expertise of their Chinese counterparts. First on Ôta's agenda was the Ancient Learning School (Kogakuha ¥j ¾Ç ¬£), championed initially by Itô Jinsai ¥ì Ãà ¤¯ ÂN (1627-1705), and then further developed by Ogyû Sorai ²ý ¥Í Ìu ±t (1666-1728) and his considerable following.

Surprisingly, perhaps, Ôta Kinjo severely criticized Itô Jinsai for his "lack of erudition" (¤£ ³Õ) and for the fact that he was "not very good at evidential research" (¤£ ªø ¦Ò ÃÒ). Nevertheless, Ôta acknowledged that Itô's behavior as a model scholar-literatus had been exemplary. This critique of Itô's classical learning and his lack of philological expertise by a Tokugawa scholar conversant with Qing Learning should help put to rest recent claims by modern Japanese scholars that Itô Jinsai's "School of Ancient Learning" may have influenced the development of classical learning, particularly evidential research, in Qing China. Often this recent claim is defended by scholars who compare Itô's classical scholarship with that of Dai Zhen in late eighteenth century China.[24] Although there were similarities in their focus on "ancient studies" (guxue ¥j ¾Ç in China, kogaku in Japan), Dai Zhen's use of evidential research techniques, particularly phonology, in his scholarship was never matched by Itô Jinsai, as Ôta Kinjo correctly noted in 1804.[25]

Ôta considered Ogyû Sorai's classical scholarship a major improvement over Itô Jinsai's, but Ôta quickly added that Sorai had been dissolute in his behavior and erratic in his teachings. According to Ôta, Ogyû Sorai had "understood the value of evidential research studies, but Sorai's own kôshôgaku had never been very outstanding" (¨ä ©Ò ¦Ò ÃÒ ©¹ ©¹ ¤£ ºë). Here we have another comment by Ôta Kinjo, a voice from within the historical development of Tokugawa classical studies, that clearly gainsays claims by modern Japanese scholars in the twentieth century that Sorai and his followers also may have influenced Qing dynasty evidential research. Again, we find that Sorai's philological expertise was no match for that of the Qing scholars that Ôta Kinjo was familar with.[26]

Ôta Kinjo attacked Sorai and his followers for their heterodox ideas. By stressing Xunzi ¯û ¤l over Mencius ©s ¤l , Sorai had, according to Ôta, overthrown the orthodox transmission of classical doctrine from Mencius to the present via both Han and Song Learning. Ôta noted: "Although Jinsai attacked [the authenticity of] several Classics, his findings did not entail any heterdoxy. Now Sorai seems to honor the Classics, but his findings place him at the edges of heterodoxy." Compared to Song classical scholars, who mixed Buddhist with classical doctrine, Sorai's approach wandered even further from the Chinese literati mainstream. At least the teachings of the Song scholars still accorded with the Way of the sages. "Sorai, although he honored the sacred Classics, still what he saw in them was base indeed. It was the equivalent of the [Legalist] teachings of Guan Zhong ºÞ ¥ò and Shang Yang °Ó »ß."[27]

According to Ôta, Tokugawa scholars of the Chinese Classics had fallen into serious doctrinal error for over sixty years as a result of the acclaim that Sorai and his followers received in Japanese scholarly and shogunal circles. Classical scholars now stressed utilitarian values, which for Ôta meant that they were overlooking the moral teachings of Confucius and Mencius. In the aftermath of the Kansei proscription of heterodoxy, Ôta sought in his scholarship and teaching to redirect Chinese classical learning away from the heterodox conclusions reached by followers of "Ancient Learning" in Japan. His eclectic position was based in part on his sense that Song Learning, although tainted by Daoist and Buddhist accretions, was still vastly superior to the Legalism disguised as classical learning that the Sorai School bandied about. The issue for Ôta was the choice between utility and profit versus morality and righteousness. Ôta and other members of the Eclectic School saw their efforts as a direct refutation of the dangerous heresies Sorai had introduced into Tokugawa scholarly circles. In this sense, he agreed with the proponents of the Kansei reform of Tokugawa schools and the classical learning taught there. Morality, not utility, was the key to scholarship and education.[28]

The Eclectic School's critique carried over to Qing Learning. Ôta spoke for the setchû gakuha when he pointed out the limitations of pure erudition. Any body of knowledge that lost track of its ethical underpinnings was as potentially dangerous as Sorai's heresies, according to Ôta. Broad erudition without moral training and the daily practice for virtue was a dead-end. Qing Learning thus needed to be complemented by Song Learning. Evidential research had to be enriched by moral cultivation. True practical learning (jitsugaku ¹ê ¾Ç) was not an empty mouthing of utilitarian values, as for Sorai and his school, but according to Ôta the daily practice of virtues that had been heralded by the sages, Confucius, and Mencius in antiquity.[29]

Writing after the Kansei prohibitions had been announced, eclectics such as Ôta Kinjo wanted to have Chinese classical learning both ways. Ôta's position represented a post-reform compromise between Song and Han Learning, but the compromise was still potentially dangerous. Leaders of the Kansei prohibitions, who had futilely tried to stamp out anything that smacked of anti-Cheng-Zhu heterodoxy in the late eighteenth century, had also lumped the eclectics together with the Sorai school. Doctrinal niceties were not the shogunate's forte. Matsudaira Sadanobu's cultural and political enterprise was to restore educational confidence in the Cheng-Zhu persuasion of classical orthodoxy and thereby eliminate the prevalence of all forms of heterodoxies, whether "Ancient Learning" or eclecticism, which in his view had ruined public morality among urbanized elites. Heterodoxy was forbidden for all students of classical learning in Tokugawa Japan. At the same time, educational reform was undertaken, according Robert Backus, "in the expectation that it would improve the morale and performance of the bureaucracy by training the character and abilities of the men who were to staff it."[30] However, the leaders of the Kansei anti-heterodoxy campaign, as well as eclectics such as Ôta Kinjo, had left out of the debate over Qing learning, the role that Ancient Learning philology and Chinese-style Han Learning was playing in the emergence and development of Tokugawa nativism.[31]

The growth of sectarian divisions in Tokugawa academic circles had forced the leaders of the Kansei prohibitions to link the eclection position to the same heterodox doctrines propagated by the Sorai school, which Ôta Kinjo tried to decouple. Initially it mattered little to the shogunate's cultural spokesmen that followers of the Eclectic School, such as Ôta Kinjo, had in fact effectively gainsaid the Sorai position and were in effect returning in part to the moral teachings of Cheng-Zhu Song Learning. Purity required strict adherence to orthodoxy. By daring to supplement the latter with the kôshôgaku of Qing learning, eclectics in the minds of the Kansei purists had "joined the pack of bête noires who were dismembering Confucianism."[32]

The eclectics had the last word, however. Despite the Kansei prohibitions, the vitality of Ôta Kinjo's classical learning could not be easily tossed aside. A research methodology that depended on the latest developments in Qing Learning and evidential research increasingly gained ground. Empirical research proved as irresistable in Tokugawa Japan as in Qing China. As the Sorai version of "Ancient Learning" lost ground among classical scholars because of orthodox and eclectic assaults, the eclectics were able to increase their voice in the expression and reproduction of classical learning in the early nineteenth century, which included the publication of scholarly works drawing on the books from China that represented the latest discoveries of Qing Learning. One such discovery pertained to the authenticiy of the Old Text (guwen ¥j ¤å; J. kobun) Classics, orthodox in China since the Later Han «á º~ (A.D. 25-220) dynasty.

4. Qing Classicism and the Old Text Documents Classic Controversy

The slow but steady emergence of evidential research studies in Qing China as a self-conscious field of academic discourse was predicated on the centrality of philological research to: 1) determine the authenticity of classical and historical texts; 2) unravel the etymologies of ancient classical terms; 3) reconstruct the phonology of ancient Chinese; and 4) clarify the paleography of Chinese characters. These trends, which began in the late Ming and influenced early Tokugawa scholars such as Itô Jinsai and Ogyû Sorai, climaxed under the Qing. As a representative example of the overall direction in Qing evidential studies, many kaozheng scholars claimed, for instance, that the Old Text portions of the Documents Classic were forgeries from the third century A.D., and not the work of the sage-kings of antiquity.

This textual controversy became a cause celebre among Han Learning scholars, at the same time that the civil examination system used Old Text passages on the "human mind and the mind of the Way" to test candidates' knowledge of the Song Learning orthodoxy. In the chapter in the Documents Classic entitled "Counsels of Yu the Great" ("Da Yu mo" ¤j ¬ê ÂÓ), the distinction between the "human and mind of the Dao" was enunciated for the first time. The sage-king Shun µÏ had admonished the soon-to-be-crowned Yu ¬ê: "The human mind is precarious. The mind of the Dao is subtle. Have absolute refinement and singleness of purpose. Hold fast the mean" (¤H ¤ß ±© ¦M¡A¹D ¤ß ±© ·L¡A±© ºë ±© ¤@¡A¤¹ °õ ³Ö ¤¤¡C).[33]

Taken together, these two passages from the Documents Classic became key pillars of the orthodox Dao Learning position during the Yuan, Ming, and Qing dynasties. In a culture that drew its ideals from a past golden age populated by sage-kings of unquestioned wisdom, orthodoxy expected classical verifications for its present articulation. Accordingly, Cheng Yi had drawn the explicit bifurcation between the human mind (renxin ¤H ¤ß) as uncontrolled desire and the mind of the Dao (daoxin ¹D ¤ß) as heavenly principle: "The human mind equals human desires; therefore it is very precarious. The mind of the Dao equals heavenly principle; therefore it is extremely subtle. Only through refinement can the [mind of the Dao] be observed. Only through singleness of purpose can it be preserved. In this manner only can one hold to the mean."[34]

Zhu Xi, building on Cheng Yi's Northern Song interpretation, gave the renxin daoxin ¤H ¤ß ¹D ¤ß (J. jinshin dôshin) passage a new theoretical twist in the Southern Song by subsuming the distinction into his own philosophy of principle: "Those who speak of the precariousness of the human mind mean that it is the sprout of human desires. The subtlety of the mind of the Dao conveys heavenly principle."[35] Zhu was suggesting that his bifurcation between li ²z (principle) and qi ®ð (pneuma, energy) had its counterpart in Shun's distinction between the mind of Dao and the human mind. The former could be described as ethical, that is, the source of moral principles, and the latter as human, that is, the source of desires and hence of evil as well. To the degree that Zhu Xi's concepts of li and qi were mutually exclusive, and thus mutually irreducible, his position could be interpreted as introducing an antagonism between moral principles and the material world of human desires.

In his famous1189 preface to his Zhongyong zhangju ¤¤ ±e ³¹ ¥y (Phrases and sentences in the "Doctrine of the Mean"), which became required reading for all young men preparing for civil examinations, Zhu made more explicit his reason for linking the distinction between the moral and human mind to his philosophy of lixue ²z ¾Ç(studies of principle). Moreover, he added to the distinction of the mind of Dao and the human mind the parallel distinction between "public" (gong ¤½) and "private" (si ¨p) enunciated in the "Zhou guan" ©P©x ("Offices of Zhou") chapter of the Documents Classic: "If one does not know how to control the mind, then it is precarious. The more precarious [the human mind becomes] the more subtle the subtle [mind of the Dao] becomes. The public-mindedness of [universal] principles thus has no way to overcome the personal concerns of one's human desires. One must cause the mind of the Dao always to be the master of the person and the human mind always to obey it."[36]

Zhu Xi moved freely between the Four Books and Five Classics, treating them holistically as the basis for the thought-world of the classical age. Zhu's efforts culminated with Cai Shen ½² ¨H (1167-1230), his student, who used the "human and mind of the Dao" passage for a holistic interpretation of all the chapters in the Documents, a view that became required in the Yuan and Ming civil examination curriculum. Students were expected to memorize the Cheng-Zhu position on the Classics and elaborate on it for imperial examiners, but even the latter increasingly recognized by the eighteenth century that many orthodox views were philologically suspect.[37]

Since the Song dynasty, many doubts had been expressed concerning the provenance of the Old Text chapters of the Documents, but it was not until Yan Roju's research and the definitive conclusions he drew in his unpublished but widely distributed manuscript (in Qing China but not in Tokugawa Japan--see below) entitled Evidential analysis of the Old Text Documents (Shangshu guwen shuzheng ©| ®Ñ ¥j ¤å²¨ ÃÒ) that the question was considered settled.[38]

Based on Yen's demonstrations that the Old Text portion was not authentic, some officials sent memorials to the throne in the 1690s and again in the 1740s calling for elimination of the Old Text chapters from the official text used in the civil examinations. Each time the proposals were set aside by the court. Hui Dong ´f ´É (1697-1758), the doyen of Han Learning in Suzhou Ĭ ¦{, had renewed Yan Roju's attack on the Old Text chapters in the 1740s. Hui noted that it had taken several centuries for suspicions concerning the Old Text Documents to lead anywhere conclusive. Hui Dong's Han Learning followers continued research on the Old Text chapters, picking up where their mentor had left off. Changzhou's Sun Xingyan ®] ¬P l (1753-1818), with his definitive study of the variances between the classical recensions of the Old and New Text Documents brought to completion the attack on the spurious Old Text chapters. Sun's analysis of Later and Former Han «e º~ (206 B.C. - A.D. 6) sources marked one of the high points of Han Learning prestige during the Qing dynasty.[39]

At the confluence of classical studies, legitimation of imperial power, and public policy, the conservative position vis-á-vis the Classics taken by Song Learning advocates represented their cultural solidarity with the imperial orthodoxy of the Ming and Qing dynasties. The Han Learning threat to the orthodox Old Text Classics threatened the shared consensus enshrined since the early Ming in the civil examination curriculum. Many refused to accept the textual findings of evidential research scholars. For example, Zhuang Cunyu ²ø ¦s »P (1719-88), a Hanlin Academy academician frequently assigned to supervise provinical examinations, and later a leader in the reemergence of New Text classicism, noted while serving as a court secretary to emperor in the 1740s that if the long accepted Old Text chapter known as the "Counsels of Yu the Great" were impugned, then the cardinal doctrine of the "human mind and mind of the Dao," as well as Gao Yao's (minister to Emperor Shun) legal injunction, which stated "rather than put to death an innocent person, you [Shun] would rather run the risk of irregularity," would be subverted. These were teachings, Zhuang contended, that depended on their classical sanction. Accordingly, on ideological grounds, Zhuang Cunyu attempted to set limits to the accruing kaozheng research in the Han Learning mainstream by insulating the classics from such criticism. Such conservative efforts in Qing China in the mid-eighteenth century predated the late eighteenth century Kansei anti-heterdoxy campaign in Japan.[40]

Still, the Qing civil examinations continued to cite the passage on the "human mind and mind of the Dao" from the Old Text "Counsels of Yu the Great" chapter with no indication of the philological controversy surrounding its authenticity. Examiners and students faithfully recapitulated the Cheng-Zhu interpretation of the transmission of the mind of the sage-kings. During both the regular 1730 and special 1737 metropolitan examinations, for example, policy questions raised in the third session dealt with the "human mind and mind of the Dao" passage. For the first policy question of 1730, examiners explicitly brought up the distinction between the moral and human mind while asking candidates to discuss the metaphysical attributes of the Supreme Ultimate (taiji ¤Ó ·¥). The answer prepared by Shen Changyu ¨H ©÷ ¦t (1700-44), the top finisher on the metropolitan and secundus on the palace examination, was reprinted in the official record and was rated by one of the chief examiners as "learning having a basis." Shen's exemplary essay presented the Song literati view of cosmology whereby the Supreme Ultimate gave rise to yin and yang, which in turn produced the five evolutive phases and the world of myriad things.

Shen's essay explored how the relation between nature and the mind corroborated the Cheng-Zhu distinction between the human mind and mind of the Dao. Without the moral categories derived from nature, the mind remained unaffected by its roots in the Supreme Ultimate. The practice of benevolence required "nurturing one's nature" by "having singleness of purpose and holding fast to the mean." Otherwise, Shen concluded, the "human mind" would reign, and one's heavenly nature containing moral principles would be lost. Rhetorically presenting his answer to the Yongzheng Emperor (r. 1723-35), Shen appealed to the "orthodox studies" (zhengxue ¥¿ ¾Ç) upon which his essay was based.[41]

Classical predispositions in China began to change in the late eighteenth century, however. Provincial and metropolitan examiners, for instance, began to test technical kaozheng topics previously outside the civil curriculum. In chronological terms, however, policy questions based on Han Learning crested in the early nineteenth century, a generation after its intellectual triumph among southern literati in the late eighteenth century. In the 1810 Jiangnan ¦¿ «n provincial examination for candidates from Anhui ¦w À² and Jiangsu ¦¿ Ĭ provinces, for instance, the first of the third session's policy questions straightforwardly raised the issue of the authenticity of portions of the Documents Classics.

The examiners opened their query by immediately raising the debate concerning the relation of the "Preface" ("Xu" §Ç) to the original hundred-chapter version of the Documents, which had long been attributed to Confucius. The examiners asked: "Why hadn't the preface been included in the [original] listing of the hundred chapters?" Next, candidates were asked to explain why during the Former Han dynasty there were discrepancies over how many chapters (28 or 29) of the New Text version of the Documents text had survived the Qin ¯³ (221-207 B.C.) "burning of the books" policy. Following this, the candidates were required to explicate the perplexing circumstances whereby Kong Anguo ¤Õ ¦w °ê (156-74? B.C.), a descendent of Confucius and a Han Erudite of the Classics, had prepared his own "Preface" for a version of the Documents that added 29 more Old Text chapters from a recently discovered text of the Documents to the earlier New Text version. "Why," the examiners asked, "had 59 chapters been listed for this version when there should have been only 58?"

After dealing with Former Han sources, the examiners turned to the Later Han dynasty classicist Zheng Xuan ¾G ¥È (130-200), the "patron-saint" of Qing dynasty Han Learning, whose scholia listed the 100 chapters in the original but lost Documents in a different order from Kong Anguo's version. "Why this discrepancy?" the candidates were asked. Subsequently, issues related to Tang and Song handling of the Documents text were raised. Why had Kong Yingda ¤Õ ¿o ¹F (574-648), then in charge of Tang efforts to settle on authoritative texts for the classical examination curriculum, labelled a third version of the Documents from the Han dynasty a forgery? Why had Zhu Xi voiced suspicions concerning the unusual phraseology (for Han dynasty writings) of Kong Anguo's commentary and preface to the Documents?[42]

The organization and content of this query reveal the degree to which the philological discoveries associated with Han Learning and evidential research had begun to filter from literati publications into the civil examination system. Although still a test of cultural and political loyalty, whereby the Qing reign was praised by the examiners for nourishing classical studies, this exploration of the textual vicissitudes surrounding the Documents Classics required precise information that would demonstrate to the examiners that the candidate was aware of the authenticity controversy surrounding this particular Classic. Rather than a test of cultural orthodoxy, however, the question raised potentially corrosive issues that could challenge orthodox "truths." One of the key Old Text chapters now thought by many literati to be a forgery was the "Da Yu mo," which contained classical lessons on the basis of which the theories of "orthodox statecraft" (zhitong ªv ²Î) and "orthodox transmission of the Way" (daotong ¹D ²Î) had been constructed.[43]

Such textual concerns were not unique to the Yangzi delta, although the academic community there had been pioneers in reviving Han Learning concerns and appropriating kaozheng research techniques for classical and historical studies. Changes in civil examination questioning were occurring empire-wide, principally as a result of Qing appointments of provincial examiners, who frequently came from the Yangzi delta and thus were conversant with the latest research findings of classical scholars there. Yangzi delta scholars had long been the most successful on the metropolitan and palace examinations in Beijing and thus were the most likely to gain appointment to the Hanlin Academy and the Ministry of Rites. Most who served as provincial examination officials were chosen from the latter two overlapping institutions in the metropolitan bureaucracy. The impact of kaozhengxue also extended beyond the boundaries of the Qing dynasty. Tokugawa scholars such as Ôta Kinjo eagerly awaited the arrival of the most recent classical works published during the Qing dynasty.[44]

5. Ôta Kinjo and the Old Text Documents Classic Controversy

In the seventh chapter (ken ¨÷) of his Kyûkeitan, Ôta Kinjo took up the Documents Classic (J. Shôsho ©| ®Ñ), a topic he had raised in more detail in his earlier work entitled Kabekyô bensei ¾À¸g ¿ë¥¿ (Discerning the Correctness of the Walled Classics). This latter title referred to the Chinese Classics that had been allegedly discovered in one of the walls in King Gong of Lu's ¾| ®¥¤ý (r. 154-127 B.C.) residence when the latter expanded it into a palace upon taking the throne there. By tradition, the residence had also been Confucius' ancestral home, and at the time of the Qin "burning of the books" scholars in Lu had purportedly placed the Classics in the wall of Confucius' residence to preserve them for posterity. Their rediscovery in the early decades of the Former Han dynasty later touched off a movement to restore these "old texts," which were thought more authentic than the contemporary "new texts" (jinwen ¤µ¤å; J. kimbun) that had been reconstituted, some from memory, when the Han dynasty replaced the Qin in 207 B.C.[45]

By choosing the title "Discerning the Correctness of the Walled Classics," Ôta Kinjo had effectively enunciated in Tokugawa scholarly circles the paleographical (mojigaku ¤å¦r ¾Ç; C. wenzixue) origins of the Old Text versus New Text debate then prominent in Qing dynasty scholarly circles. By 1800, Qing evidential research scholars had reached a general consensus on the authenticity of the Old Text versions of the Classics that we will rely on to analyze how much of this Han Learning evidential research Ôta Kinjo had grasped in his research on the Old Text Documents.

In his analysis of the New and Old Text chapters of the Documents Classic, Ôta Kinjo concluded that there had been three different versions of that Classic since the Former Han dynasty. The first, transmitted by Fu Sheng ¥ñ³Ó after the fall of the Qin dynasty in 207 B.C., was the "New Text" version in 28 or 29 chapters, which had lost more than 70 chapters of the original. The second version was the Old Text version transmitted by Kong Anguo during the reign of Emperor Wu (Wudi ªZ «Ò, r. 140-87 B.C.), which added some sixteen new chapters and were part of the "Old Text" Classics later championed by the Han imperial scholar Liu Xin ¼BÝõ (45 B.C. - A.D. 23) during the reign of Emperor Ping ¥ (A.D. 1-7) and during Wang Mang's ¤ý²õ usurpation between 9 and 23 A.D. The third, according to Ôta, was a forged version of the Old Text version with 25 chapters, which had been allegedly rediscovered by the scholar Mei Ze ±ö ðó circa A.D.307-323, and presented to the Imperial Court of the Eastern Jin ªF ®Ê (317-385) dynasty. Of these three versions, Ôta Kinjo considered only the New Text and original Old Text versions authentic. Ôta regarded Mei Ze's version of the Old Text chapters as a later forgery, but he contended that the original Old Text version of the Documents, which according to Ôta was finally lost in the Tang dynasty, had also been authentic. This meant that the only authentic version that had survived from antiquity was the New Text rendering.[46]

In final remarks appended to his discussion of the Documents Classic in the Kyûkeitan, Ôta indicated that many Han classical scholars during the Yuan and Ming dynasties had already attacked the authenticity of the Mei Ze version of the Old Text chapters and upheld the authenticity of Fu Sheng's New Text version. By his own admission, Ôta was not breaking new ground on these issues, but he did think he was contributing something new to the accrued classical research coming from China. In Ôta's words he had "discovered something that earlier people had not" (µo «e ¤H ©Ò ¥¼ µo), namely, that the original Old Text chapters associated with Kong Anguo had been authentic. Ôta's use of this six character phrase to describe his own findings was a formulaic rendering of how Qing scholars long associated with kaozhengxue in the seventeenth and eighteenth centuries had described their own discoveries.[47]

But whose works from Qing China did Ôta Kinjo and the other eclectic scholars in Edo in the early eighteenth century have access to? How could he be so sure that his discovery concerning this question of authenticity was new, when the controversy had raged among eighteenth century Han Learning scholars and produced many studies of the issue in Qing China? And why didn't Ôta refer to the famous study by Yan Roju, discussed above, which had circulated in the early eighteenth century in China and become the authoritative study of the Old Text Documents? To answer these questions, we can draw on Ôta's own account of how the Old Text Documents controversy had evolved in China after the fall of the Han dynasty.[48]

Ôta Kinjo's account of past scholarship in China began with the suggestion that the Old Text forgery discovered by Mei Ze in the early fourth century had likely been composed by a follower of the post-Han classicist Wang Su ¤ýµÂ(195-256). Wang had sought to gainsay the dominant position of Zheng Xuan's eclectic classical learning, which combined New and Old Text views of the Classics, in the Later Han literati world. Wang Su claimed, for example, that one of his students, who happened to be a descendant of Confucius, had presented him with a text known as the Kongzi jiayu ¤Õ¤l ®a »y (School sayings of Confucius), which conveniently contained passages supporting Wang Su's views. Ôta did not cite sources from China to support this claim, but we know that Qing scholars such as Yan Roju also entertained this view of Wang Su's role in classical forgeries.[49] Zheng Xuan had wanted to synthesize the New and Old Text schools, and other Later Han literati likewise had attempted to mediate between the Old and New Text traditions. In this effort, they were later surpassed by Wang Su, who seems to have had a hand in the preparation of the more down-to-earth portrait of Confucius presented in the School Sayings of Confucius, a text purportedly dating from the Former Han but which did not take its final form until the third century A.D.[50]

Ôta noted that numerous other works, including the School Sayings of Confucius and the Kongcongzi ¤ÕÂO¤l (Kong family master' anthology),[51] suddenly appeared after the fall of the Han, and most could be associated with Wang Su and his followers. In fact, Zheng Xuan was criticized by Wang Su for accepting too much of the New Text tradition. Later, Wang Su's views were accepted as authoritative during the Tang dynasty as the basis for imperial ritual, and the Old Text Classics were once and for all declared orthodox. The forgeries of the post-Han era had become canonical.

In Ôta's view, the Song scholar Wu Yu §d Ùµ(1124 jinshi degree) had been the first in China to question the authenticity of the Mei Ze version of the 25 Old Text chapters of the Documents. Ôta added that Zhu Xi also had been suspicious of the Kong Anguo commentary for the Old Text chapters, which he linked to the other forgeries that suddenly appeared in the early years of the Eastern Jin dynasty. Here, again, although he did not cite his Qing sources, Ôta Kinjo was following Qing accounts of the development of doubts concerning the provenance of the Old Text portions of the Documents Classic and the Kong Anguo preface and commentary, which suspiciously resembled post-Han writings. We know of Ôta's dependance on Qing scholarship here because he next mentioned the findings of Mao Qiling ¤ò©_ ÄÖ (1623-1716), a staunch defender of the authenticity of the Old Text Documents. In fact, Mao had been the chief rival Yan Roju had faced in convincing other Qing scholars that the Old Text chapters associated with Mei Ze were later forgeries.[52]

Much of Ôta's subsequent account took the form of a refutation of Mao Qiling's most famous work on the Documents controversy entitled Guwen shangshu yuanci ¥j ¤å©| ®Ñ Þ µü (Words on the injustice done to the Old Text Documents), which had itself been constructed by Mao as a point-by-point refutation of Yan Roju's charges of forgery in his Evidential Analysis of the Old Text Documents. Again, it is curious that Ôta Kinjo made no mention of Yan Roju's findings that Mao had tried to refute. Ôta had access to the published version of Mao Qiling's study of the Documents controversy, which was included in the Kangxi ±d º³ edition circa 1699 of Mao's complete works known as the Xihe heji ¦è ªe ¦X ¶°(Mao Qiling's combined works), which had made its way to Japan as part of the Zhejiang-Nagasaki trade, most likely because Mao was a prominent Zhejiang scholar from Xiaoshan ¿½¤s , a town very close to Hangzhou ªC ¦{, the capital of the province.

The reasons why Ôta had not yet heard of Yan Roju's study were twofold. First, Yan's magnum opus had been passed around only in manuscript form when it was first completed in China. It was not printed until 1745. Yan's manuscript was the only version that Mao Qiling saw, and because Yan was a rival scholar from Jiangsu province, his work may not have been widely available in Zhejiang until the nineteenth century. Many Jiangsu scholars such as Hui Dong, who championed Han Learning in Suzhou in the mid-eighteenth century, had not seen it either. Indeed, Ôta Kinjo had not even seen Hui Dong's 1770s study entitled Guwen shangshu kao ¥j ¤å ©| ®Ñ ¦Ò (Study of the Old Text Documents).

Second, although he corresponded with Yan, Mao was very careful to avoid any mention of Yan Roju in his Guwen shangshu yuanci. Known for his arrogance and a desire to win debates at all costs, Mao had his work quickly printed circa 1699 to establish his position among Qing scholars empire-wide. Yan's manuscript, available to Mao, was never cited. Rather, Mao used his own disciple, Li Gong §õÜã (1659-1733), as his foil. Li Gong, for example, studied ritual music under Mao Qiling in 1698, but Li also met with Yan Roju in in 1699 on his way home to North China. In Mao's study of the Documents, Li Gong queried Mao about certain suspicious points concerning the Old Text chapters, which Mao refuted, he thought, in convincing fashion.[53]

Mao Qiling had maintained that the Mei Ze version of the Old Text chapters of the Documents was authentic and that it was authenticated by the historical literature. Ôta Kinjo thus thought he was refuting the findings of a major Qing scholar when Ôta claimed that only the original 16 chapter Old Text version had been authentic. In Ôta's view, Han classical scholars such as Zheng Xuan had never seen the later 25 chapter version. In addition, he argued that Mao Qiling had overly relied on Sui and Tang accounts to verify the Mei Ze version of the Old Text chapters. Ôta tried to demonstrate that such post-Jin dynasty medieval accounts were unreliable. Tang scholars in particular, Ôta pointed out, had been unable to discern the tracks of the forger because they had not compared the Mei Ze version with textual emandations from the original 16 chapter version prepared by Han classicists.[54]

Ôta noted that the uncanny resemblance between Wang Su's commentary for the Documents and the Mei Ze version had also been raised by Tang scholars such as Lu Deming ³° ¼w ©ú (556-627) and Liu Zhiji ¼Bª¾´X (661-721). Ôta added that even Kong Yingda, the classical scholar-official in charge of the Tang dynasty project to prepare orthodox versions of the Classics in the seventh century, had indicated in his notes that he also suspected that Wang Su had secretly seen the forged commentary for the Mei Ze version before Wang had prepared his own commentary. For Ôta, this indicated that Tang scholars had gotten the sequence of textual events wrong and not recognized that the 25 chapters of the Mei Ze version had been prepared by a follower of Wang Su, who likely had relied on Wang's own commentary for his commentary. Because Kong Yingda and other Tang scholars were unreliable sources, Mao Qiling's position, which was based squarely on these Tang accounts, was untenable. The coup de grace for Ôta was that Qing scholars such as Quan Zuwang ¥þ¯ª±æ (1705-55) had even accused Mao Qiling himself of forgery.[55]

Ôta Kinjo next took up the chronology for the discovery of the Old Text Classics in the early years of the Former Han dynasty. The discovery of the Classics by King Gong of Lu took place in 154 B.C., yet the bibliography section of the Hanshu º~ ®Ñindicated that this had occurred during the reign of Wudi, which commenced in 140 B.C. However, because 154 B.C. was likely the correct date, Ôta maintained that this chronology cast significant doubt on the Kong Anguo commentary and preface. Anguo had been born in 156 B.C. and for him to present the Old Text chapters and his commentary to the imperial court in the last years of the Wudi reign, circa 87 B.C., as the "Preface" attributed to him claimed, he would had to be an old man. Yet the Hanshu also indicated that Kong Anguo had died young, which meant that Anguo could not have lived to prepare the "Preface" or commentary.

For his source, Ôta cited the study by the Qing scholar Zhu Yicun ¦¶ÂU´L (1629-1709) entitled Guwen shangshu kao ¥j ¤å ©| ®Ñ ¦Ò (Study of the Old Text Documents). In addition, he referred to the widely read study by the eighteenth century Han Learning scholar Wang Mingsheng ¤ý »ï ²± (1722-97) entitled Shangshu houan ©| ®Ñ «á ®× (Late cases dealing with the Documents Classic). Their conclusions had been that a member of Anguo's family after his death, and not Kong Anguo himself, had presented the Old Text Classics to the Former Han imperial court. The preface and commentary attributed to Anguo had subsequently been forged. It was also curious to Ôta that Sima Qian ¥q °¨ ¾E (145-86? B.C.), although he had studied under Kong Anguo and used the original 16 chapters of the Old Text tradition, did not cite a single phrase or sentence in the Shiji ¥v °O (Records of the Grand Historian) from the Mei Ze version of the Documents or from the Anguo commentary to it. Little did Ôta Kinjo realize that Yan Roju had presented this position most authoritatively in China circa 1700.[56]

Ôta's analysis requires some further elucidation of a textual history that he, like his Qing informants, took for granted. Liu Xin, mentioned above as a key figure in the Han dynasty discovery of the Old Text Classics, had mastered the classical repertoire when his father Liu Xiang ¼B ¦V (80-9 B.C.), an eminent Former Han classical scholar, was appointed to the Imperial Secretariat. After weathering the political storms of the day, Liu Xiang was charged with cataloging works in the Imperial Library archives during the reign of Emperor Cheng ¦¨ (r. 28-25 B.C.). Liu Xin assisted his father in this project, which was the equivalent in its day of the Siku quanshu ¥| ®w ¥þ ®Ñ (Complete works in the Four Treasuries [of the Imperial Library]) in the late eighteenth century. During his work in the imperial archives, however, Liu Xin unearthed previously neglected works composed in ancient styles of calligraphy, i.e., "Old Text," which were said to have been discovered over a century earlier in Confucius' former residence. Among these texts, Liu Xin listed the I-li ¶h § (Leftover chapters of the rites) in thirty-nine chapters and sixteen chapters of the Documents Classic. Liu also allegedly recovered the Zuozhuan ¥ª ¶Ç, which he claimed had been prepared as a commentary to Confucius' Spring and Autumn Annals (Chunqiu ¬K ¬î) by Confucius' disciple Zuo Qiuming.

Thinking he had in his hands the authentic versions of the Classics written under Confucius' own direction, Liu Xin contended that these Old Text Classics were superior to the texts then used in the Han Imperial Academy to authorize official appointments as Erudites of the Five Classics (Wujing boshi ¤ ¸g ³Õ ¤h). His chance to promulgate them as orthodox texts in the Imperial Academy came when Wang Mang, Xin's friend since his student days, recommended Liu to Emperor Ai «s (r. 6 B.C.-A.D. 1) to finish Liu Xiang's work on a comprehensive bibliography for the Imperial Library. In his famous "Letter to the Erudite of the Chamberlain for Ceremonies," Liu Xin described the provenance of the Old Text Classics, an account that Ôta Kinjo frequently referred to, and thus worth translating in part here:[57]

In antiquity, after [the legacy of the sage-kings] Yao and Shun had long been corrupted, the Three Dynasties [Xia, Shang, and Zhou] successively arose. Sagely emperors and enlightened kings reappeared one after another. The Way of the sages was clearly illumined. After the Zhou dynasty declined, [court] ritual and music were no longer correct. The Way of the sages became impossible to preserve intact. Consequently, Confucius worried that the Way would not be put into effect, and so he travelled to all the states of the empire. After he returned to Lu from Wei, [court] music became correct. The Ya and Song [songs of the Poetry Classic] were all made appropriate. He prepared the Change [Classic], wrote a preface for the Documents, and compiled the Annals to commemorate the Way of the emperors and kings. When Confucius perished, his esoteric words were cut off; when his seventy disciples died, the great meanings were betrayed. . . .

When the Han dynasty arose, it was separated from the time of the sagely emperors and enlightened kings by a long expanse of time. The Way of Confucius moreover had been cut off. Laws and institutions had no precedents to follow. At the time, there was only Shu Suntong [¨û ®] ³q, fl. ca. 221-206 B.C.], who was able to define in general terms what [the correct] rituals and ceremonies should be. In the empire, there remained intact only the divinations of the Change; no other books had survived [the Qin dynasty].

By the time of Emperor Hui ´f [r. 194-188 B.C.], the [Qin] proscription against owning books was rescinded. However, leading officials remained military men such as Zhou Bo [©P «k, d. 169 B.C.] and Guan Ying Äé À¦ [n.d.], who paid no heed [to the Classics]. During the reign of Emperor Wen ¤å [r. 179-157 B.C.], for the first time the clerk Chao Cuo ´Â ¿ù [d. 155 B.C.] was assigned to study and receive the Documents under the tutelage of Fu Sheng. The Documents had just been retrieved from the wall [in his home where it had been hidden], but its bamboo slips had been mixed up because the strings holding them together had decayed. Today, that version survives, but classical scholars transmit the readings from it and no more. . . .

When King Gong of Lu destroyed Confucius' residence in order to build a palace there, [works written in] Old Text were discovered in the walls of the residence. There were thirty-nine leftover chapters on ritual and sixteen chapters of the Documents. After the Tianhan ¤Ñ º~ era [100-97 B.C.], Kong Anguo presented them [to the Imperial Academy], but because of the witchcraft trials then rocking the court, they were not officially accepted. Even the [commentary to] the Spring and Autumn Annals by Zuo Qiuming, which was an old work written in more than twenty bundles of ancient script that had survived in the imperial archives, had not yet been unearthed.

Earlier, Liu Xin had requested that the erudites accept the Old Text works he had found as authoritative Classics required for study in the Imperial Academy. In effect, Liu Xin challenged the bureaucratic "rice bowls" of Former Han officialdom. Their loyalties remained to the "New Texts" established as orthodox since Dong Zhongshu ¸³ ¥ò µÎ (179?-104? B.C.) under Emperor Wu. Liu's letter to the erudites reflected the bitterness he felt after his proposal to the erudites had been dismissed. Furious that his views were not considered important enough to merit even discussion, Liu Xin attacked the erudites for their shortsightedness. Liu's famous letter cost him his position at court. For the remainder of the Former Han dynasty, Liu Xin was persona non grata at the capital.

Upon the death of Emperor Ai in A.D. 1, Wang Mang, then well-placed in imperial politics, became the power behind the throne. Wang immediately recalled Liu Xin to the

capital to serve as an advisor and high official. Under Liu Hsin's direction, the Old Text

Classics, which now included the Zuozhuan, the Mao recension of the Poetry Classic, the Rites of Chou, and the extra sixteen Old Text chapters of the Documents, were made orthodox and replaced the New Text Classics in the Imperial Academy. In 9 A.D., Wang Mang overthrew the Han dynasty and came to power himself as first emperor of the Xin ·s (New) dynasty. The change in dynasty was an opportune moment in both political and intellectual terms. Wang Mang required a classical framework that would legitimate his seizure of dynastic power from the Liu family; Liu Xin, even though he was a descendent of the Liu house, required political backing to maintain the Old Text Classics in the Imperial Academy. In exchange for classical legitimacy, Wang Mang granted Liu Xin what he had been fighting for. Now at the head of hundreds of scribes and clerks in the Imperial Academy, Liu Xin controlled the orthodox voice of the state. Under Liu's direction, classical texts were edited and compiled in accordance with the Old Text sources he had discovered. Henceforth, classical scholars, whether in China, Korea, or Japan, would begin with the Classics that Liu Xin and his staff had put in place as the venue for imperial legitimacy. Some would later argue that even Liu Xin's version of the Old Text chapters, now lost, had been forged.

During the last years of Wang Mang's rule, Liu Xin was implicated in a plot to assassinate Wang and take power himself. When the plot was discovered, however, Liu Xin committed suicide, just a few months before Wang's Xin dynasty fell and the Later Han dynasty succeeded it. Although Old Text Classics associated with Wang Mang and Liu Xin were discredited during Emperor Guangwu's ¥ú ªZ reign (25-57), the New Text Classics were never restored to the position of eminence they had enjoyed during the Former Han. At meetings convened in A.D. 79 at the White Tiger Hall to achieve a classical consensus concerning the official Canon, the principal issue was no longer a choice of New Text versus Old Text Classics. Both were now irrevocably part of the Han tradition of canonical texts.

By the second century A.D., scholars such as Jia Kui ¸ë ¶f (30-101), Xu Shen ³\ ·V (58-147), Ma Rong °¨ ¿Ä (76-166), and most notably Zheng Xuan, as described above, combined New and Old Text Classics and commentaries in order to synthesize the earlier traditions associated with the Classics. Later Han dynasty classical scholars reacted against the Former Han and Later Han scholars who had dabbling into the occult aspects of prognostication and apocrypha texts. Wang Su and his followers brought this scholarly current to a climax, but as we have seen the charges of forgery associated with texts lost after the fall of the Han also date back to this era.

Thus, Ôta Kinjo's account built on this Han dynasty story of textual transmission to show how the forger of the Mei Ze version had been able to manipulate the information about the troubled transmissions of the New and Old Text chapters of the Documents and thereby had successfully substituted a forgery in the place of the missing Old Text chapters. Ôta concurred with Qing scholars who painstakingly traced the textual origins of the classical phrases and sentences that had been lifted from other sources and were worked into the Mei Ze version of 25 Old Text chapters. The forger had been an outstanding classical scholar. Ôta included among these deceptions the forger's successful lifting of the famous "human mind and the mind of the Way" passage from a parallel passage in the writings of the pre-Han philosopher Xunzi ¯û ¤l (fl. 298-238 B.C.). In addition, the forger had used many other texts and references to fill out his forgery and make it classically credible. Thus, the Old Text issue touched on the authenticity of important passages that made up both the Qing and Tokugawa classical orthodoxy.[58]

Building on the research of Zhu Yicun, Wang Mingsheng, and other Qing evidential research scholars, Ôta Kinjo was not only able to controvert Mao Qiling's conservative position on the Mei Ze version of the Old Text Documents. He was also able to demonstrate that the Former Han version in 16 chapters had existed in several different recensions, all of which had been authentic. Ôta thought that only Wang Mingsheng and Xu Qianxue ®} °® ¾Ç (1631-94) had realized that the original Anguo version and another version written in lacquer by Du Lin §ù ªL (d. A.D. 47) were both the same.[59]

In his concluding remarks on the Documents Classic, we get a sense of the excitement that Ôta Kinjo must have felt each time he got his hands on a new classical work from China via the Nagasaki trade. Ôta wrote that his research on the Documents began when he was 17 or 18, circa 1783, in discussions with the Edo scholar Itô Gaugaku ¥ì ªF ÆR ©¨, whose father was a follower of the Sorai school. Ôta claimed that even then he had recognized that the current Old Text chapters were forgeries. After studying the issue for another ten years, Ôta argued that he understood the exact reasons why the chapters were false. Later, after getting his hands on Mao Qiling's Guwen shangshu yuanci, he wrote his account, the aforementioned Kabekyô bensei, to refute Mao's position.[60]

After completing his research and thinking he had made his own contributions to the authenticity debate as it had evolved in Qing China, Ôta must have greeted each Qing work he received in the early nineteenth century, when he was approaching 50, with some anticipation. Had anyone explained the problem as he had? Had the works of any scholars in China, besides the hints provided by Xu Qianxue and Wang Mingsheng, recognized the authenticity of the original 16 chapters of the Old Text chapters that Liu Xin had championed in the Former Han and under Wang Mang?

Then, finally, it arrived via Nagasaki. A copy of the 1745 printed edition of Yan Roju's Shangshu guwen shuzheng had arrived in Japan, after Ôta had completed his detailed Kabekyô bensei but some time just before the Kyûkeitan was ready for printing in 1815. Ôta poignantly wrote:

At the very end, I got a copy of Yan Roju's Guwen shuzheng (Kobun shôshô). Upon reading it I realized that my theory [that the original Old Text chapters had been authentic] had already been discovered by Roju. Early Qing scholars who had doubted the authenticity of the [later] Old Text [version] were numerous, and Roju was the ancestral scholar for this. It is only because his book arrived by sea in a trading vessel very late, that I earlier said [that I had "discovered something that earlier people had not"]. Those who read these assessments later can have no doubt about how it happened.[61]

In many ways, Ôta Kinjo was publicly apologizing for his classical pretentions. But he was also declaring that his independently derived analysis had confirmed Yan Roju's conclusions. At least, Ôta had refuted without knowing it Yan's scholarly nemesis: Mao Qiling. While Ôta had initially not realized what had actually inspired Mao's vituperations about the authenticity of the Old Text chapters, he had clearly and forcefully argued against them, point by point. There was a certain symmetry and integrety to Ôta's use of kôshôgaku techniques in his classical studies, particularly his willingness to acknowledge the priority of Yan Roju in the genealogy of scholars in China and Japan who had worked on the Old Text Documents.[62]

**********

According to Nakamura Kyûshirô, kôshôgaku in Tokugawa Japan had always depended on classical currents in China for its growth and elaboration. Nakamura has argued, for example, that Ogyû Sorai had depended on late Ming scholars such as Li Panlung §õ Ãk Às (1514-70) for Sorai's elaboration of Ancient Learning in Japan, which means that the Sorai school was based on intellectual currents that had flowered in China 150 years earlier. Similarly, Nakamura maintained, evidential research in Tokugawa Japan was about a century behind developments in China. Thus from Nakamura's perspective, Ôta Kinjo's Kabekyô bensei and Kyûkeitan were composed 100 years after Yan Roju's seminal Shangshu guwen shuzheng.[63]

Ôta Kinjo's evidential studies and their dependance on classical works arriving from China before and after the Kansei anti-heterdoxy campaign indicate, however, the intellectual limits to the Tokugawa shogunate's efforts to shore up the Cheng-Zhu Dao Learning orthodoxy. Although the Song Learning orthodoxy moved forward in the years following the Kansei reforms, its proponents had clearly failed to check its Han Learning alternative still coming in through Nagasaki via the import and spread of new classical works and novel methodologies from south China, the heartland of evidential research studies since the mid-eighteenth century.[64] Kôshôgaku as a methodology was no one's monopoly in China or Japan. It could be used to attack Cheng-Zhu learning, as Sorai had in part done, or to reaffirm it with reservations, as Ôta Kinjo and the eclectics did. It could also be employed by nativist scholars in Japan who sought to purify ancient Japanese chronicles and poetry masterpieces of their sinitic encumbrances and thereby restore the true "spirit" of Japanese antiquity.[65]

The uses of philology as the proper scholarly tool to affirm and verify classical knowledge was superseding the content of that knowledge. There have long been suggestions that distinguished post-Meiji Japanese historians such as Shigeno Yasutsugu « ³¥ ¦w ö (1827-1910) in Tokyo and Naitô Kônan ¤º Ãà ´ò «n (1866-1934) in Kyoto learned to apply the methodology of German Rankean history by integrating western learning with their earlier training in kôshôgaku. In the process, the Tokyo and Kyoto University traditions of historical research achieved maturity. As in Qing China, classical scholars in Tokugawa Japan were in part providing the nativist foundations for a tradition of precise, empirically based research and impartial analysis. In the eighteenth century, Tokugawa scholars still emulated research from Qing China. In the late nineteenth century, however, Qing scholars increasingly learned from intellectual developments in Meiji Japan.[66]

---

[1] See Elman, From Philosophy to Philology: Social and Intellectual Aspects of Change in Late Imperial China (Cambridge: Harvard University Council on East Asian Studies, 1984), and Classicism, Politics, and Kinship: The Ch'ang-chou School of New Text Confucianism in Late Imperial China (Berkeley: University of California Press, 1990).

[2] See Anthony Grafton, Defenders of the Text: The Traditions of Scholarship in an Age of Science, 1450-1800 (Cambridge: Harvard University Press, 1991), and Grafton and Lisa Jardine, eds., From Humanism to the Humanities. Education and the Liberal Arts in Fifteenth- and Sixteenth-Century Europe (Cambridge: Harvard University. Press, 1986).

[3] Elman, From Philosophy to Philology, passim.

[4] See Elman, "The Formation of 'Dao Learning' as Imperial Ideology During the Early Ming Dynasty," in Theodore Huters, R. Bin Wong, and Pauline Yu, eds., Culture & State in Chinese History (Stanford: Stanford University Press, 1997), pp. 58-82.

[5] Elman, Classicism, Politics, and Kinship: The Ch'ang-chou School of New Text Confucianism in Late Imperial China, passim.

[6] See Elman, From Philosophy to Philology, passim.

[7] Hamaguchi Fujiô ÀØ ¤f ´I ¤h ¶¯, Shindai kokyogaku no shisô shi teki kenkyû ²M ¥N ¦Ò ¾Ú ¾Ç ÇU «ä ·Q ¥v ªº ¬ã ¨s (Research on the intellectual history of Qing dynasty evidential studies) (Tokyo: Kokusho kankôkai, 1994).

[8] See also Ôba Osamu ¤j ®x ²ç, Edo jidai ni okeru Chûgoku bunka juyo no kenkyû ¦¿ ¤á ®É ¥N ÇR Æð Æ÷ Çr ¤¤ °ê ¤å ¤Æ ¨ü ®e ÇU ¬ã ¨s (Research on the acceptance of Chinese culture in the Edo period) (Tokyo: Dô bôsha, 1984), and Laura Hess, "The Reimportation from Japan to China of the Kong Commentary to the Classic of Filial Piety" (Seattle: University of Washington Ph.D. dissertation, 1994).

[9] Fujitsuka Chikashi Ãà ¶ï ¾F, Shinchô bunka tôden no kenkyû ²M ´Â ¤å ¤Æ ªF ¶Ç ÇU ¬ã ¨s (Research on the eastern transmission of Qing dynasty culture) (Tokyo: Kokusho kankôkai, 1975). See also Robert Backus, "The Kansei Prohibition of Heterodoxy and its Effects on Education," Harvard Journal of Asiatic Studies 39, 1 (June 1979): 62-63.

[10] Nakamura Kyûshirô ¤¤ §ø ¤[ ¥| ¦, "Kôshôgaku gaisetsu" ¦Ò ÃÒ ¾Ç ·§ »¡ (Outline of Evidential Research), in Kinsei Nihon no Jugaku ªñ ¥@ ¤é ¥» ÇU ¾§ ¾Ç (Confucian studies in early modern Japan), compiled by the Committee in Commemoration of Tokugawa Shukuga's Seventieth Birthday (Tokyo: Iwanami, 1941), pp. 709-11.

[11] See Ôta Kinjo, "Sôron," in Kyûkeitan (Edo and Kyoto, 1815). Cf. Backus, "The Kansei Prohibition," pp. 56-59.

[12] Ôta Kinjo, "Sôron," in Kyûkeitan, 1.2a.

[13] Kyûkeitan, 1.2b-3a.

[14] Kyûkeitan, 1.3b-4a.

[15] Kyûkeitan, 1.4b-6a.

[16] Kyûkeitan, 1.11b.

[17] Kyûkeitan, 1.12b. See also the article by Ôba Osamu on the Nagasaki trade in this volume.

[18] See Elman, A Cultural History of Civil Examinations in Late Imperial China (Berkeley: University of California Press, 1999), chapter 2.

[19] See Herman Ooms, Charismatic Bureaucrat: A Political Biography of Matsudaira Sadanobu, 1758-1829 (Chicago: University of Chicago Press, 1975), passim.

[20] Kyûkeitan, 1.12b-13a. See also Robert Backus, "The Motivation of Confucian Orthodoxy in Tokugawa Japan," Harvard Journal of Asiatic Studies 39, 2 (December 1979): 275-338.

[21] See Qingdai qianqi jiaoyu lunzhu xuan ²M¥N«e ´Á ±Ð¨| ½× µÛ ¿ï (Selections of writings on education from the early Qing period), edited by Li Guojun §õ °ê ¶v et al. (Peking: People's Education Press, 1990, 3 vols.), 3/305-07, 3/326-27.

[22] See Chen's 1807 afterword in Qingdai qianqi jiaoyu lunzhu xuan, 3/392-93.

[23]Qingdai qianqi jiaoyu lunzhu xuan, 3/455-56, 3/459-60,

[24] See the different views recently taken by Okada Takehiko ©£ ¥Ð ªZ «Û, Edoki no Jugaku ¦¿ ¤á ´Á ÇU ¾§ ¾Ç (Confucian learning in the Edo period) (Tokyo: Tosho Press, 1982), pp. 74-110, and Yu Yingshi §E ^ ®É, Lun Dai Zhen yu Zhang Xuecheng ½× À¹ ¾_ »P ³¹ ¾Ç «° (On Dai Zheng and Zhang Xuecheng) (Hong Kong: Longmen Bookstore, 1976), pp. 185-96, on this issue.

[25] On Dai Zhen's evidential research, see Hamaguchi Fujiô, Shindai kokyogaku no shisô shi teki kenkyû, pp. 177-217.

[26] Kyûkeitan, 1.13a-14b. See Fujitsuka Chikashi Ãà ¶ï ¾F, "Mono Sorai no Rongo sei to Shinchô no keishi" ª« Ìu ±t ÇU½× »y ¼x ÇO²M ´Â ¸g ®v (Ogyû Sorai's Proofs of the Analects and Qing dynasty classical teachers), Shinagaku kenkyû ¤ä ¨º ¾Ç ¬ã ¨s 4 (1935): 1-61.

[27] Kyûkeitan, 1.14b-15a.

[28] Kyûkeitan, 1.15b-16b.

[29] Kyûkeitan, 1.17a-b.

[30] Backus, "The Kansei Prohibition of Heterodoxy," pp. 55-59.

[31] See Mark McNally, "Spectral History: Hirata Atsutane and Tokugawa Nativism (Los Angeles: UCLA Ph.D. dissertation in History, 1998), chapters two and three.

[32] Backus, "The Kansei Prohibition," pp. 55-59.

[33] Shangshu tongjian ©| ®Ñ ³q ÀË (Concordance to the Documents Classic) (Peking: Chaori wenxian Press, 1982 reprint), 03/0517-0532 (p. 2). I have followed, with minor changes, the translation in Wing-tsit Chan, "Zhu Xi's Completion of Neo-Confucianism," Études Song-Sung Studies, Ser. 2, No. 1 (1973): 79.

[34] Er-Cheng quanshu ¤G µ{ ¥þ ®Ñ (Complete writings of Cheng Hao and Cheng Yi), Henan Cheng-shi yishu ªe «n µ{ ¤ó ¿ò ®Ñ (Bequeathed writings of Cheng Yi), (Taibei: Zhonghua Bookstore Sibu beiyao edition, 1927-35), 19.7a-7b.

[35] Zhuzi daquan ¦¶ ¤l ¤j ¥þ (Master Zhu [Xi's] Great Compendium), (Shanghai: Zhonghua Bookstore, Sibu beiyao edition, 1927-35), 67.19a.

[36] Zhuzi daquan, 76.21a-22a.

[37] Cai Shen, "Xu" §Ç (Preface) to the Shu jizhuan ®Ñ ¶° ¶Ç (Collected commentaries to the Documents) (Taipei: World Bookstore, 1969), pp. 1-2.

[38] For recent research, see Liu Renpeng ¼B ¤H ÄP, "Lun Zhuzi weichang yi guwen shangshu weizuo" ½× ¦¶ ¤l ¥¼ ¹Á ºÃ ¥j ¤å ©| ®Ñ °° §@(Zhu Xi never doubted the authenticity of the Old Text Documents), Qinghua xuebao ²M µØ ¾Ç ³ø New Series (Taiwan), 22, 4 (December 1992): 399-430.

[39] For discussion, see Elman, "Philosophy (I-li) Versus Philology (K'ao-cheng): The Jen-hsin tao-hsin Debate," T'oung Pao 59, nos. 4-5 (1983): 175-222.

[40] See my Classicism, Politics, and Kinship: The Ch'ang-chou School of New Text Confucianism, chapters 3-5.

[41] Huishilu ·| ¸Õ ¿ý (Record of metropolitan civil examinations), 1730: 41a-43a.

[42] Jiangnan xiangshi timing lu ¦¿ «n ¶m ¸Õ ÃD ¦W ¸ô (Record of successful candidates in the Jiangnan provincial examination), 1810: 9a-9b, in the archives of the No. 1 Historical Archives, Beijing. For purposes of focus, I have not described other important debates here. See for example, Elman, "Ming Politics and Confucian Classicism: The Duke of Chou Serves King Ch'eng," in Mingdai jingxue guoji yantaohui lunwenji ©ú ¥N¸g ¾Ç °ê »Ú ¬ã °Q·| ½× ¤å ¶° (Taibei: Academia Sinica, 1996), pp. 93-171.

[43] For discussion, see my From Philosophy To Philology, pp. 177-80, 200-02, 207-12.

[44] See Elman, "Changes in Confucian Civil Service Examinations From the Ming to Ch'ing Dynasties," in Elman and Alexander Woodside, eds. Education and Society in Late Imperial China, 1600-1900 (Berkeley: University of California Press, 1994), pp. 135-143.

[45] Kyûkeitan, 7.1a. See Elman, Classicism, pp. xxvi-xxx.

[46] Kyûkeitan, 7.13b-14a.

[47] Kyûkeitan, 7.14a-b. See Elman, From Philosophy To Philology, pp. 65-66.

[48] See Elman, From Philosophy To Philology, pp. 177-80.

[49] From Philosophy To Philology, pp. 200-02.

[50] See Jack Dull, "A Historical Introduction to the Apocryphal (Ch'an-wei) Texts of the Han Dynasty" (Seattle: University of Washington Ph.D. dissertation, 1966), pp. 113-38, 152-76, 183-241. On Wang Su, see R.P. Kramers, K'ung-tzu chia-yu: The School Sayings of Confucius (Leiden: E. J. Brill, 1950), pp. 194-96. We should add that the usual view of Old Text as "rationalistic" and New Text as superstitious" is simplistic. Liu Xin, for example, extensively used the apocrypha for his Old Text position, as did Zheng Xuan during the Later Han. See Tjan Tjoe Som, Po Hu T'ung. The Comprehensive Discussion in the White Tiger Hall (Leiden: E.J. Brill, 2 vols., 1949 & 1952), I/141-54.

[51] See Yoav Ariel, K'ung-Ts'ung-Tzu: the K'ung family masters' anthology: a study and translation of chapters 1-10, 12-14 (Princeton: Princeton University Press, 1989).

[52] Kyûkeitan, 7.1a-b. See Elman, From Philosophy To Philology, pp. 200-02.

[53] Kyûkeitan, 7.1a-b, and Mao Qiling, Guwen shangshu yuanci (ca. 1699 edition), 1.3a. See also Li Gong's preliminary remarks included at the opening of the Guwen shangshu yuanci, pp. 1a-2b. Cf. Arthur Hummel, ed., Eminent Chinese of the Ch'ing Period (Taipei: Chengwen Press reprint, 1972), pp. 358, 477, and Elman, From Philosophy, p. 200. Mao's letter to Yan was included in Mao's Xihe heji, but Ôta did not refer to this letter.

[54] Kyûkeitan, 7.1b-3a.

[55] Kyûkeitan, 7.3a-5b. Apparently, Ôta still did not realize that Mao had deliberately avoided mentioning Yan Roju's manuscript, upon which the Guwen shangshu yuanci had been based.

[56] Kyûkeitan, 7.5b-7a. Cf. Elman, From Philosophy To Philology, pp. 177-78.

[57] See Liu Xin's Yi Taichang boshi shu ²¾ ¤Ó ±` ³Õ ¤h ®Ñ, in the Hanshu (History of the Former Han dynasty), compiled by Ban Gu ¯Z ©T (7 vols. Peking, Zhonghua Bookstore, 1962, and Taibei: Shixue chubanshe, 1974), 4/1969-1970.

[58] Kyûkeitan, 7.8a-10b.

[59] Kyûkeitan, 7.11b-13a.

[60] Kyûkeitan, 7.13b.

[61] Kyûkeitan, 7.14b.

[62] On the issue of priority in Qing scholarly circles, see my From Philosophy To Philology, pp. 221-28.

[63] Nakamura Kyûshirô, "Kôshôgaku gaisetsu," pp. 28-29.

[64] Backus, "The Kansei Prohibition," pp. 103-06.

[65] Mark McNally, "Spectral History: Hirata Atsutane and Tokugawa Nativism," chapters 4-7.

[66] Jiro Numata, "Shigeno Yasutsugu and the Modern Tokyo Tradition of Historical Writing," in W. G. Beasley and E. G. Pulleyblank, eds., Historians of China and Japan (Oxford: Oxford University Press, 1961), pp. 264-87. See also Joshua Fogel, Politics and Sinology: The Case of Naitô Kônan (1866-1934) (Cambridge: Harvard Council on East Asian Studies, 1984), pp. 5-16, and D. R. Howland, Borders of Chinese Civilization: Geography and History at Empire's End (Durham: Duke University Press, 1996), pp. 197-241.