Empire and the Circulation of Frontier Intelligence: Qing Conceptions of the Ottomans

Empire and the Circulation of Frontier Intelligence Qing Conceptions of the Otomans Matthew W. Mosca University of Hong Kong S erving in Europe between 1887 and 1890, the Qing diplomat Hong Jun 洪鈞 (1840–1893) made eforts to solve a problem that had been puzzling him: why Turkey was called Khungghar (Ch. Hongga’er 鴻噶爾 ) in China. Before inding an answer that satisied him, Hong consulted Russians and Otoman envoys in Europe, gathered supporting references to British and Prussian practice, and surveyed Chinese sources.1 At irst glance, his research displays the cosmopolitan hallmarks of late Qing geographic scholarship. But upon closer inspection, both the underlying questions—Who, what, or where is Khungghar?— and the search for answers across languages and cultures, through oral and textual inquiries, continued a line of investigation reaching back to I wish to thank James E. Bosson for his support and invaluable assistance throughout this project. Mark C. Elliot has also generously given me the beneit of his expertise, as have the Journal’s anonymous reviewers. he following people have given advice and assistance: David Brophy, Devon Dear, Seunghyun Han, Ying Hu, Renyuan Li, Onuma Takahiro, Jonathan Schlesinger, and Hoong Teik Toh. An earlier version of this research was presented at the annual meeting of the Association for Asian Studies, 2007. Work on this article was supported by two postdoctoral fellowships, irst at the Center for Chinese Studies, University of California, Berkeley, and subsequently at the Hong Kong Institute for the Humanities and Social Sciences, University of Hong Kong, for which I record my gratitude. Any errors that remain are, of course, my own. 1 Hong Jun, Yuanshi yiwen zhengbu 元史譯文證補 , Xuxiu Siku quanshu edition, vol. 293 (Shanghai: Shanghai guji chubanshe, 1995) [hereater XXSKQS], 27xia.3b–4a. Published by the Harvard-Yenching Institute hJAs 70.1 (2010): 147–207 147 148 Matthew W. Mosca the beginning of the dynasty. he present essay explores the meaning of Khungghar, or more precisely what generations of scholars and oicials took it to mean, as part of a larger efort to study the circulation of information within the Qing empire. To understand how knowledge of the outside world circulated, the Qing empire must be considered as an integrated unit. Distinctive regional and intellectual backgrounds certainly inluenced how individual subjects understood foreign countries, but such understandings were formed under the inluence of empirewide networks of information. hese included not only oicial channels of communication but also private ones that were created and sustained, though oten tenuously, by personal contacts within the apparatus of the Qing central government. Past studies of how the vast and variegated Qing empire was brought under central control have tended to emphasize the efective way Manchu rulers—especially the Kangxi (r. 1661–1722), Yongzheng (r. 1723–1736), and Qianlong (r. 1736–1796) emperors—adapted administrative structures and rituals to the cultural norms of their subject peoples and divided the domain into compartments united only by their common submission to the same imperial center.2 Viewing the empire in terms of the circulation of information, this article seeks by contrast to emphasize connections between subjects. he diferent cultural segments of the empire, despite the barriers posed by linguistic and intellectual heterogeneity, had become integrated enough to exert an inluence on one another. Knowledge was received, distributed, and reformulated between regions and communities that at 2 he description of the Qing empire as composed of discrete cultural units, each administered diferently by the court, is common among studies of what has been called “New Qing History.” In one inluential description of the imperial ideology underlying Qing governance, Pamela Kyle Crossley has argued that rulers, above all Qianlong, aspired to be universal monarchs of “a domain in parts” by legitimizing their rule to each constituent unit “simultaneously” in the cultural and political idioms most suited to secure their allegiance. Because the emperor was “the sole point where all speciics articulated” and also “culturally null” (i.e., not personally preferring any one of the models of rulership prevailing within the empire), subjects were not confronted by the other cultural orientations within the Qing domain. To the contrary, Crossley argues, Qianlong tried to enforce boundaries to preserve what he considered to represent the pure cultural and ethical norms of diferent groups from cross-fertilization. See A Translucent Mirror: History and Identity in Qing Imperial Ideology (Berkeley: University of California Press, 1999), esp. pp. 1–52, 221–80. For a critique of the model of the Qing as a “compartmentalized empire” and its failure to adequately explain imperial cohesion, constructed chiely from an economic standpoint, see Kwangmin Kim, “Saintly Brokers: Uyghur Muslims, Trade, and the Making of Qing Central Asia, 1696–1814” (Ph.D. diss., University of California, Berkeley, 2008), pp. 25–37. Frontier Intelligence in the Qing 149 irst glance may appear to have been efectively isolated. he low of information was limited, to be sure, but signiicant enough that evolving views on many topics of geographic, historical, or strategic significance—of which Khungghar was only one—cannot be elucidated unless one considers the empire as a single circuit for the transmission of knowledge. Needless to say, no homogeneous perspective on any intellectual or political question emerged among Qing subjects, but it is also misleading to seek isolated Tibetan, Mongol, Manchu, or Chinese perspectives. he overarching imperial structure opened channels of communication, which stimulated new perspectives that cannot be atributed to a single group of subjects or cultural bloc. To a large extent, the exchange among groups occurred beyond the conscious mediation and deliberate policy of the Qing government, through private channels created unintentionally as a by-product of the formal administrative structure to which they ran parallel. For reasons of state, the central government gathered knowledge through multilingual bureaucratic agencies reporting to the Grand Council. Occasionally, the stafs of court agencies were ordered to transform these administrative documents into oicial compendia. hese documents and court publications are now the best understood channels of information circulation within the empire. Yet the duties of iling and compiling these materials that were given to Manchu, Mongol, and Han Chinese staf inspired private, “of-duty” inquiries, leading to alternative channels of information transfer not required by bureaucratic regulations. Court and frontier oicials used their spare time to gather information from informants, read archives for their own ediication, question colleagues of other ethnic and linguistic backgrounds, and in some cases record their indings in privately authored manuscripts or printed works that would complement oicial productions and thus circulate information not found elsewhere. he heterogeneity of the Qing imperial elite led information to circulate in paterns distinct from those in European colonial empires. In addition to their loyal Manchu, Mongol, and Han bannerman, Qing rulers summoned to Beijing, permanently or in regular rotation, Chinese oicials, Mongol princes, Tibetan Buddhist clerics, Eastern Turkestani begs, and Jesuit missionaries.3 his diverse elite was 3 A substantial literature has emerged to describe the diversity of the Qing court, of which the following titles are only a sample. On bannermen, see Mark C. Elliot, he 150 Matthew W. Mosca fragmented because they had diferent bureaucratic career tracks, social circles, languages, religious and cultural perspectives, and educational backgrounds. Information commonplace to one group might be unknown to another. his is in contrast to the circulation of information in British India described by C. A. Bayly, who, drawing on the work of Manuel Castells, argues that an “information order” connected “the colonial state’s surveillance agencies” with the “autonomous networks of social communicators” in Indian society.4 In this model, the most critical gaps in knowledge transmission were those separating the “political intelligence” sought by the ruling elite from the “indigenous knowledge” of its subjects.5 In the Qing case, by contrast, members of the imperial elite were more isolated from colleagues of other cultural backgrounds than from subjects from the same cultural realm. To identify a Qing “information order” underlying the empirewide low of knowledge one must therefore concentrate on the small number of sites and media that allowed information to cross the linguistic and cultural barriers.6 his article will concentrate particularly on the low of intelligence from the empire’s Inner Asian frontier into the hands of Han Chinese scholars and oicials. Over the course of the dynasty, the mechanisms opening or constricting these channels changed profoundly. For approximately the irst century of Qing rule, the Manchus and Mongols who administered the frontier and recorded information in their own languages held a virtual monopoly on frontier intelligence. Soon ater 1800, published debates on frontier geographic and political afairs emerged almost exclusively from the hands of Han literati experts, as would be the case for the remainder of the dynasty. Two Manchu Way: he Eight Banners and Ethnic Identity in Late Imperial China (Stanford: Stanford University Press, 2001); for contact between Mongol nobles and the Qing court see Ning Chia, “he Lifanyuan and the Inner Asian Rituals in the Early Qing (1644–1795),” LIC 14.1 (1993): 60–92; for Jesuits at court see Louis Pister, Notices biographiques et bibliographiques sur les jésuites de l’ancienne mission de Chine, 1552–1773 (San Francisco: Chinese Materials Center, 1976); for Tibetan lamas see “he Qing Court’s Tibet Connection: Lcang skya Rol pa’i rdo rje and the Qianlong Emperor,” HJAS 60.1 (2000): 125–63. 4 C. A. Bayly, Empire and Information: Intelligence Gathering and Social Communication in India, 1780–1870 (Cambridge: Cambridge University Press, 1996), pp. 1–4. 5 Bayly, p. 2. 6 On information within the oicial channels of the Qing central government see Beatrice S. Bartlet, Monarchs and Ministers: he Grand Council in Mid-Ch’ing China, 1723– 1820 (Berkeley: University of California Press, 1991). Frontier Intelligence in the Qing 151 major changes during the intervening half century (1750–1800) led to this radical reordering of the low of information. First, the administrative and literary demands of the conquest of the Western Regions (1755–1759) brought about an unprecedented conjunction of interests between the elite Manchu and Mongol administrators of frontier areas and their Han colleagues at court. Increasingly the former transmited their knowledge, verbally and in writing, and the later actively sought out and recorded such information. Second, when the death of the Qianlong emperor in 1799 allowed the Han scholarly elite to publish on frontier topics without fear of violating taboos, a huge corpus of research materials appeared in print. Up to then, knowledge about the frontier had been largely the preserve of Manchu and Mongol administrative personnel and the Chinese scholars who had access to them through court posts; now, through printing, it was opened up to literati of any oicial or scholarly background. Under the dominance of Chinese scholarly standards and assumptions, research agendas and modes of argument came to be framed by the methods of textual scholarship. In the eyes of textual researchers, continued access to irsthand frontier intelligence was unnecessary, and the hearsay on which it was based appeared suspect. Investigating the evolving circulation of frontier knowledge also allows one to reconsider the corpus of available historical sources regarding Qing interactions with Inner Asia. Due to China’s highly developed print and manuscript culture, information put into Chinese was more successfully propagated than that in other languages of the empire. herefore, it is quite possible that much information has been lost that was once widely familiar among the non-Chinese segment of the elites of the empire; and, disturbingly, that we may know far less about the Qing court and its Inner Asian activities than we think. A more thorough exploration of Manchu- and Mongolian-language archival materials will, one hopes, ameliorate this problem over time, but at present published sources—overwhelmingly in Chinese—still underpin much of our understanding of Qing Inner Asia. Placing these Chinese sources, on which scholars have perforce relied heavily, within the context of the overall Qing information order will illuminate their limitations. his article approaches these questions about the Qing information order through the study of how the single term Khungghar, 152 Matthew W. Mosca roughly indicating the Otoman emperor and his lands, circulated within the Qing imperial elite. Because this term was native to none of the languages and peoples of the empire, yet was known to scholars and oicials of Chinese, Mongol, Manchu, and Tibetan backgrounds, it represents a comparatively neutral case that allows consideration of a wide cross-section of the empire. Our primary focus throughout will be on the speciic links between two large constellations within the Qing elite: the comparatively integrated Manchu-Mongol-Tibetan sphere on the Inner Asian frontier, and the world of Han Chinese scholar-oicials who, until the very end of the dynasty, were excluded from high posts in frontier administration. Lumi or Khungghar? Two Realms of Early Qing Frontier Knowledge In the decades ater 1644, in parallel with their conquest of China, the Manchu rulers of the Qing empire consolidated and expanded their Inner Asian holdings, especially in Manchuria and Mongolia. heir involvement in the political world of the northern and western Mongols required political contacts extending to Tibet and the shores of the Caspian Sea.7 Strategic need compelled the court to gather news in many languages across an enormous expanse of territory, but this information was slow to reach Han Chinese scholars, who continued to use Ming-era (1368–1644) sources to study Central Asia. As a result, diferent understandings about the frontier circulated in two distinct, relatively independent spheres, Chinese and Mongol-Manchu; only gradually were connections forged between these two spheres that permited the empirewide circulation of information. One new object of interest that emerged in the Kangxi reign was a place or person called Khungghar, indicating the Otoman empire or its ruler.8 Denis Sinor has identiied this word as the Mongolian form 7 For a review of Qing diplomacy among the Mongols before 1759, including the mission to Ayuki Khan of the Torghud near the mouth of the Volga, see Peter C. Perdue, China Marches West: he Qing Conquest of Central Eurasia (Cambridge: Harvard University Press, 2005), pp. 133–299. 8 A loanword, this term had many variant spellings in Mongolian, Manchu, Tibetan, and Chinese. D. Sinor, citing the dictionary of J. E. Kowalewski, identiies the two Mongolian forms Qungγar and Küngγar; see his “Qungγar: a curious Mongol appellation of the Turks,” in Varia Eurasiatica: Festschrit für Professor András Róna-Tas (Szeged: Department of Altaic Studies, 1991), p. 165. L. S. Puchkovskiĭ, relying partly on the work of Walther Frontier Intelligence in the Qing 153 of the Otoman Turkish hunkār, meaning “sovereign,” adopted by the Otoman emperor Murad II (r. 1421–1451) and later becoming “a general term applied to sultans, mainly by foreigners.”9 Early occurrences of this term in Qing sources conform to this meaning, but in time its signiicance became a subject of considerable debate. his paper will retain it untranslated, in order to preserve the multiple interpretations given to it by individual authors. For much of the Ming period, the term roughly equivalent to Khungghar had been the Chinese word Lumi 魯迷 , a transliteration of the name Rum or Rumi, which was successively used for Rome, the eastern Roman Empire centered on Byzantium, parts of Muslim Anatolia, and then the Otoman lands.10 Between 1423 and 1618, at least thirteen missions from Lumi arrived at the Ming court. Most came ater the Otoman Empire had raised its proile in Central Asia, both by conquering Mecca and by the sultans’ adoption of the title of Caliph, claiming “rulership of the entire Islamic world.”11 As early as 1547, a detailed description of the route from China to Istanbul (Lumi cheng 魯迷城 ) was in the possession of Chinese geographers.12 Lumi was accorded entries in many oicial and unoicial Heissig, identiies the forms Güngger, Küngker, and Küngkür; see his Mongol’skie, Buri͡ atMongol’skie i Oĭratskie Rukopisi i Ksilografy Instituta Vostokovedeni͡ia (Moscow-Leningrad: Izdatel’stvo Akademii Nauk SSSR, 1957), pp. 41–42. Here I follow Sinor and use the form Qungγar throughout, modiied as Khungghar. 9 Sinor, pp. 165. Sinor gives a thorough account of past speculation over the origin of the term. Paul Pelliot also takes up the etymology in less detail in his Notes critiques d’histoire kalmouke (Paris: Librairie d’Amérique et d’Orient, 1960), pp. 88–89 n. 258. 10 Cemal Kafadar, “A Rome of One’s Own: Relections on Cultural Geography and Identity in the Lands of Rum,” Muqarnas: An Annual on the Visual Culture of the Islamic World 24 (2007): 7–25. Variants of Lumi date to the Song. he much older term Fulin 拂菻 and its variants, oten indicating Byzantium/Constantinople/Istanbul, might also derive from “Rome.” See Donald D. Leslie and Kenneth H. J. Gardiner, Roman Empire in Chinese Sources (Rome: Bardi, 1996), pp. 281–82. 11 Haneda Akira 羽田明 , Chūō Ajia shi kenkyū 中央アジア史研究 (Kyoto: Rinsen shoten, 1982), pp. 349–54; Wada Hironori 和田博徳 , “Mindai no teppō denrai to Osuman teikoku: Shinkifu to Seiiki tochi jimbutsu ryaku” 明代の鐵砲傳來とオスマン帝國 : 神器 譜と西域土地人物略 , Shigaku 史學 31 (1958): 692–719. On Otoman activity in Central Asia in this period, see Rana von Mende-Altaylı, Die Beziehungen des Osmanischen Reiches zu Kashghar und seinem Herrscher Ya’qub Beg, 1873–1877 (Bloomington: Indiana University Research Institute for Inner Asian Studies, 1999), pp. 1–5; Colin Imber, he Otoman Empire, 1300–1650: he Structure of Power (New York: Palgrave Macmillan, 2002). Central Asian merchants sometimes posed as ambassadors to China in order to access Chinese markets; see Joseph F. Fletcher, “China and Central Asia, 1368–1884,” in he Chinese World Order: Traditional China’s Foreign Relations, ed. John King Fairbank, with contributions by Ta-tuan Ch’en and others (Cambridge: Harvard University Press, 1968), pp. 206–9. 12 Wada, pp. 701–10. 154 Matthew W. Mosca Ming geographic works.13 Ater the Qing conquest, these sources continued to inform Chinese descriptions of foreign lands. Accounts of Lumi can be found in the Ming shi 明史 (History of the Ming dynasty; 1739) and other oicially edited early Qing scholarly works.14 Authors of private geographic writings also drew on this legacy. Gu Yanwu reproduced the 1547 itinerary in his Tianxia junguo libing shu 天下郡國 利病書 (Book concerning the advantages and disadvantages of administrative units within the realm; preface 1662), while Gu Zuyu 顧祖禹 (1631–1692) included Lumi in his list of “foreigners to the southwest” in his Dushi fangyu jiyao 讀史方輿紀要 (Essential record of geography for reading histories, 1678).15 Also cited in the early Qing was Giulio Aleni’s (1582–1649) Zhifang wai ji 職方外紀 (Record of countries outside the jurisdiction of the Bureau of Operations, 1623), which described Turkey (Du’erge 度爾格 ) though ignoring its power and Islamic character.16 A few new descriptions of the frontier became available in Chinese before 1750. hese early irsthand accounts of frontier areas, including northern Manchuria, Mongolia, and Tibet, generally consisted of brief, unsystematic observations made by authors (almost without exception Han Chinese) who had either been exiled or were serving on a military expedition or imperial journey to Inner Asia (see the Appendix). Presumably their Mongol and Manchu counterparts had more extensive frontier experience and proiciency in local languages; nonetheless prior to the publication, in 1723, of Tulišen’s (Ch. Tu-li-shen 圖 理琛 , 1667–1741) Yiyu lu 異域錄 (Record of foreign regions), virtually no descriptions of the frontier or foreign lands emerged from the hand 13 Ming huidian 明會典 , Yingyin Wenyuange Siku quanshu edition, vol. 617 (Taibei: Taiwan shangwu yinshuguan, 1983) [hereater SKQS], 98.15a; Zhu Siben 朱思本 and Luo Hong xian 羅洪先 , Guang yutu (Taibei: Xuehai chubanshe, 1969), p. 424. Other references can be found in Mao Ruizheng 茅瑞徵 , Huang Ming xiangxu lu 皇明象胥錄 (Taibei: Huawen shuju, 1968), 7.25b–26b; and Yan Congjian 嚴從簡 , Shuyu zhouzi lu 殊域周咨錄 (Beijing: Zhonghua shuju, 1993), 15.497–98. 14 You Tong 尤侗 , (Ming shi) Waiguo zhuan (明史 ) 外國傳 , in “Ming shi” dingbu wenxian huibian “明史 ” 訂補文獻彙編 , ed. Xu Shu 徐蜀 (Beijing: Beijing tushuguan chubanshe, 2004), 6.8b; Ming shi 明史 (Beijing: Zhonghua shuju, 1974), 332.8626–27; (Qinding) Gujin tushu jicheng (欽定 ) 古今圖書集成 (Shanghai: Tushu jicheng yinshuju, 1884), Fangyu huibian, bianyidian, juan 86. 15 Wada, p. 701; Gu Zuyu, Dushi fangyu jiyao, Yutu yaolan 輿圖要覽 (Taibei: Hongshi chubanshe, 1981), 6:4.5686. 16 Giulio Aleni [Ai Rulüe 艾儒略 ], Zhifang wai ji (Beijing: Zhonghua shuju, 2000), 1.48–58. Frontier Intelligence in the Qing 155 of a non-Han author outside of oicial channels.17 he early accounts of the frontier available to Chinese scholars did not relect the voluminous frontier knowledge that was systematically gathered at the Qing court. A rit divided the frontier-oriented writings of Han scholars from the data circulating among those Manchu and Mongol oicials who were guiding imperial policy toward Inner Asia. At the Qing court, intensive diplomacy with the Khalkha, Zunghar, and Torghud Mongols brought atention for the irst time to the powerful country of Khungghar. Yet most oicial and unoicial geographic and statecratoriented writings produced before 1750 ignored the knowledge of Manchu and Mongol oicials. he process by which this rit was overcome, and by which the term Khungghar came to replace Lumi, was set in motion as this knowledge became available in Chinese, itfully before 1750 and more extensively thereater. he Torghud Mongols, a branch of the Oirat or western Mongols, played a central role in bringing Khungghar to the atention of the Qing court. hey had migrated early in the seventeenth century to the steppe around the mouth of the Volga, near Russia and the Crimean khanate, an Otoman vassal state.18 he Torghud generally allied themselves with Russia, but they tried to preserve their strategic autonomy by sending envoys to Persia, the Crimean khanate, and other groups, while maintaining religious ties with the Dalai Lama in Tibet, and political contact with the Qing court and the Zunghars.19 Ayuki Khan (r. 1669– 1724) sent the irst Torghud embassy directly to the Otoman court in 1680. Evidently using Muslim ambassadors, he professed loyalty to the 17 Manchu frontier oicials submited memorials that described, among other things, frontier conditions and peoples. Evelyn Rawski and Pamela Crossley, in their article “A Proile of the Manchu Language in Qing History,” HJAS 53.1 (1993): 78 n. 46, give the instances of Umuna and Funingga as authors of “travel writings in Manchu” in the form of oicial reports to the government on frontier conditions. he present article deines “travel writing” more narrowly as the voluntary composition of a work intended for circulation outside oicial channels. Prior to the mid-Qianlong period, aside from Tulišen’s, the only Manchu-authored frontier account I have found is the Saibei jicheng 塞北紀程 , by Maska (Ch. Ma-si-ka 馬思喀 ). 18 On steppe politics in this region between 1480 and 1800, see Michael Khodarkovsky, Russia’s Steppe Frontier: he Making of a Colonial Empire, 1500–1800 (Bloomington: Indiana University Press, 2002), pp. 77–183. In some cases, the term Khungghar referred to the Crimean khanate as well as the Otoman Empire proper, according to Sinor, p. 167. 19 Michael Khodarkovsy, Where Two Worlds Met: he Russian State and the Kalmyk Nomads, 1600–1771 (Ithaca: Cornell University Press, 1992), pp. 90–96. 156 Matthew W. Mosca Otomans and announced anti-Russian views.20 Although Ayuki later repaired his relations with Russia, he continued to send embassies to the Otoman sultan in 1704 and 1709.21 hus, by the time they were drawn into the emerging conlict between the Qing and the Zunghars, the Torghud were familiar with the geopolitical situation stretching from Mongolia to the Black Sea and were an important conduit for intelligence.22 Contact with the Torghud allowed Qing frontier geographers to expand the scope of their knowledge signiicantly. No other source of information about Khungghar existed before such contact, because no access was available to Western Mongolia and Turkestan—which would become the most fruitful routes of information about Khungghar ater their conquest in 1755—and even contact with Russia was rare. One of the earliest sources to describe Russia was the Chusai jilüe 出塞紀略 (Brief record of a journey beyond the frontier) by Qian Liangze 錢良擇 (1645–ca. 1707), who accompanied the high Manchu oicial Songgotu (Ch. Suo-e-tu 索額圖 , d. 1703?) on a mission to the Russian border. his book recorded Russian claims that their neighbors—Europe, the Khalkha Mongols, the Zunghars, and Persia—all feared and served them. Among these neighbors Qian also included “the Muslims,” but without specifying either Turkey or Khungghar.23 Khungghar irst entered the Manchu and Chinese geographic lexicon ater the Qing embassy of 1712–1715 went via Russia to Ayuki Khan of the Torghud. his is when Tulišen, a subordinate member of this embassy, learned of “the King of Turkey, Khungghar Khan” (Ch. Tuliyesike guowang Gongka’er han 圖里耶斯科國王拱喀爾汗 ; M. Turiyesk’o gurun i Gungk’ar han), whom he referred to several times in his Yiyu lu.24 According to Tulišen’s account, the governor of Siberia, Prince M. P. Gagarin, listed Khungghar Khan as one of the many rulers then at war.25 Gagarin, Tulišen notes, also described how Tsar Peter the Von Mende-Altaylı, p. 3. Khodarkovsky, Where Two Worlds Met, pp. 74–153. 22 Zhang Weihua 張維華 and Sun Xi 孫西 , Qing qianqi Zhong-E guanxi 清前期中俄 關係 ( Jinan: Shandong jiaoyu chubanshe, 1997), pp. 184–85; Khodarkovsky, Where Two Worlds Met, pp. 134–38. 23 Qian Liangze, Chusai jilüe (Beijing, Zhonghua shuju, 1991), p. 27. 24 Tulišen, Lakcaha jecen de takūraha babe ejehe bithe/Kōchū iikiroku: Tulišen’s I-yü-lu, ed. Imanishi Shunjū 今西春秋 (Tenri: Tenri Daigaku Oyasato Kenkyūjo, 1964), pp. 135, 344. 25 Tulišen, p. 107. 20 21 Frontier Intelligence in the Qing 157 Great had seized Azov (M. Adzoo hoton) from Khungghar Khan and restored it to him only ater peace was concluded.26 Tulišen further tells how King Charles XII of Sweden (M. Karulusi, Siiyesk’o gurun i han) was defeated by Russia and forced to take refuge in the Khungghar domain.27 he term Tuliyesike guo, the description of historical events, and the clear delineation of countries neighboring Russia leave no doubt that “Khungghar Khan” referred to the ruler of the Otoman Empire. Tulišen appears to have based his account on irsthand interviews with Gagarin; indeed he seems to have picked up a smatering of Russian on his journey.28 However, given that the term Khungghar was in use among the Mongols around this time, it is likely that the phrase “King of Tuliyesike, Khungghar han” was a compound of Russian and Mongolian usage, relecting either linguistic assistance lent to the embassy by Mongols, or a Manchu melding of Russian and Torghud terminology. Tulišen located the Otoman Empire northwest of Russia (alongside Sweden, Portugal, and England), rather than listing it among those (including the Torghud, Zunghars, Persia, and Bukhara) said to lie south of it.29 His map showed Turkey due west of Moscow, far to the northwest of Beijing.30 his conception of Khungghar’s location would vex nineteenth-century Han scholars as they tried to identify it. Around the same time as Tulišen’s account, the Manchu oicial Kui-xu 揆叙 (1674?–1717) also mentioned the Otoman Empire in his Xiguangting zazhi 隙光亭雜識 (Assorted notes of Kui-xu), where he stated: “What is called Khungghar (Hongke’er 烘克爾 ) in Mongolian, and ‘Turkey’ (Du’erke 都兒克 ) in the West (Xiyang), is a large Muslim country. Russia once paid it tribute.”31 Kui-xu relied on Mongol and Western (speciically, Jesuit) sources and seems not to have been inluenced by Tulišen. he two authors used entirely diferent transcriptions, and their descriptions do not overlap in content. Furthermore, Kui-xu’s suggestion that Russia once paid tribute to Khungghar Tulišen, p. 161. Tulišen, p. 135. 28 Tulišen, p. 23. 29 Tulišen, p. 136. 30 Tulišen, pp. 48–51. For this map see Leo Bagrow, “he First Russian Maps of Siberia and their Inluence on the West-European Cartography of N.E. Asia,” Imago Mundi 9 (1952): 83–93. 31 Kui-xu, Xiguangting zazhi, XXSKQS, vol. 1146, 2.44b. 26 27 158 Matthew W. Mosca contradicts the spirit of Tulišen’s account, which recorded Russian victories but ignored defeats. hese works represented two views of the Otoman Empire that would rival each other well into the nineteenth century: a negative view based on Russian sources, and a positive estimation derived from other Central Eurasian sources.32 Tulišen and Kui-xu, the two men who brought frontier information to the atention of Han oicials, were Manchu. As far as we know, no Han oicial was sent on the embassy through Russia. he sort of social or bureaucratic gathering where Kui-xu probably encountered his Mongol informant would have been inaccessible to his Chinese counterparts. For example, Kui-xu’s brother, Singde (Ch. Xing-de 性 德 , 1655–1685), apparently went on a mission in 1682 to the region around the Etsina River to investigate conditions among the Oirat, including the possibility that they had allied with Russia.33 While at court, Singde took notes about Jesuits, introducing some of them with “Westerners say” (Xiren yun 西人云 ), perhaps implying personal contact with Catholic missionaries.34 Kui-xu learned about frontier afairs through well-informed contacts of this kind. What made Tulišen and Kui-xu exceptional was their determination to write about frontier afairs in Chinese. Few other Manchus in the Kangxi period wrote privately concerning their experiences and thoughts about the frontier, a genre dominated by Han Chinese civil oicials. High Manchu oicials who had dealt with Russia and the Mongols, such as Songgotu, Pengcun, Maci, Sabsu, and Fuciowan, limited their writings to oicial reports. Some patronized frontier writings by Han authors. Zhang Penghe’s 張鵬翮 (1649–1725) account of an abortive mission to the frontier to negotiate with Russia originated in Maci’s (Ch. Ma-qi 馬齊 , d. 1739) desire to have the episode chronicled in Chinese. Qian Liangze composed his Chusai jilüe while in Songgotu’s employ. Fang Guancheng 方觀承 (1698–1768), a Chinese Kui-xu’s rendering of “Turkey” as Du’erke does not quite match Aleni’s Du’erge 度爾 格 or Verbiest’s Du’erge 度兒格 (Zhifang wai ji,1.48; Kunyu tushuo 坤輿圖說 , SKQS, vol. 32 594, 2.1b). Extant Jesuit texts provided no information that could connect Khungghar with Turkey; most likely Kui-xu’s knowledge came from a personal encounter with a Jesuit. 33 Liu Dehong 劉德鴻 , Qingchu xueren diyi: Nalan Xingde yanjiu 清初學人第一 : 納蘭 性德研究 (Beijing: Zhongguo shehui kexue chubanshe, 1997), pp. 341–63. Unlike Kui-xu, Singde did not record his Inner Asian experiences in a book of jotings, although he did compose some poetry on the subject. 34 Singde [Xing-de 性德 ], Lushuiting zazhi 淥水亭雜識 , Qingdai biji congkan 清代筆 記叢刊 , vol. 1 ( Jinan: Qi Lu shushe, 2001), 2.1b. Frontier Intelligence in the Qing 159 scholar who accompanied Fu-peng 福彭 (d. 1748) on campaign to Uliyasutai as a private secretary, wrote about this experience and other frontier topics.35 Manchu oicials promoted the writing of private chronicles of frontier afairs, but, before 1750, few atempted original compositions of their own. Most likely this was because private geographic accounts were generally expected to be composed in the formal style of those whom Benjamin Elman has termed the “writing elite.”36 Almost every Qing-period account of the outside world was penned by someone whose education had concentrated on the writing skills necessary to pass the civil service examinations. Most Manchus, like most Han, lacked the classical literacy that was evidently considered essential for geographic writing. Moreover, it seems that they were both less likely than their Han counterparts to derive a sense of exoticism from frontier travel and less familiar with earlier traditions of frontier writing.37 By contrast, Tulišen and Kui-xu had the requisite education to compose private frontier-oriented works in Chinese. he former was competent in Manchu and Chinese and worked on the oicial translation of the Tongjian gangmu 通鑑綱目 (Essentials of the Comprehensive mirror in aid of government) into Manchu. He produced both a Chinese and a Manchu edition of his Yiyu lu, each of which had a strikingly diferent fate.38 he Chinese version circulated in manuscript and print, collecting prefaces (seven for the irst edition alone) from eminent Han scholars. It was reprinted at least twice by Tulišen, subsequently included in the Siku quanshu manuscript collection, and further reprinted in at least eight collectanea. By contrast, the Manchu printing, which appeared around the same time as the irst Chinese edition, atracted no prefaces except Tulišen’s, was never reprinted, See Arthur W. Hummel, ed., Eminent Chinese of the Ch’ing Period (Washington: United States Government Printing Oice, 1943–44). 36 For a deinition of the “writing elite” see Benjamin Elman, A Cultural History of Civil Examinations in Late Imperial China (Berkeley: University of California Press, 2000), pp. 276–77. 37 Emma J. Teng has argued that frontier travel writing was atractive to Han literati because of the relative freshness of the subject mater (although some frontiers were fresher than others) and identiies among Han authors a “passion for ‘distant travels.’” Taiwan’s Imagined Geography: Chinese Colonial Travel Writing and Pictures, 1683–1895 (Cambridge: Harvard University Asia Center, 2004), pp. 19–20. 38 A Mongolian translation of the work was also produced at some point. See Sinor, p. 168. 35 160 Matthew W. Mosca and sank into obscurity.39 he lack of demand for Manchu- and Mongol-language descriptions of the contemporary frontier is likewise suggested by another fact: none of the approximately eighty other Manchu and Mongol oicials who had also served on Tulišen’s mission and two further embassies to Russia and the Torghud in the Yongzheng period seem to have writen about their service.40 Kui-xu’s linguistic facility can be atributed to changing cultural conditions: the Kangxi-period lowering of Chinese-language poetry composed by Manchus who increasingly participated in the Chineselanguage civil service examination system; the emperor’s close relations with Han scholars; and his enthusiasm for Confucian learning and literary projects. Some Manchus atained high levels of literary sophistication, and became disciples of the eminent Han scholars who had reconciled themselves to Manchu rule and were entering oicialdom.41 Kui-xu was part of this trend. His father, the prominent oicial Mingju (Ch. Ming-zhu 明珠 , 1635–1708), was a patron of Chinese scholars. Kui-xu studied under the scholar-poets Wu Zhaoqian 吳兆騫 (1631–1684) and Cha Shenxing 查慎行 (1650–1727). His elder brother Singde, who surpassed Kui-xu in fame as a poet, was also an acolyte of prominent Han oicials.42 Yet few of these Kangxi-period Manchu poets participated in military afairs, and thus frontier information was not a major subject of their sophisticated Chinese writings.43 hat Tulišen and Kui-xu were the only Kangxi-period men to mention Khungghar in their personal writings indicates the novelty of the term, and the scarcity of individuals able to transfer frontier information from one social or bureaucratic sphere to another within the empire. Given the tenuous connection between the Manchu elite and the Chinese literate elite, Khungghar could have been widely known at 39 Imanishi Shunjū outlines the history of the two editions in his introduction to Tulišen’s work, pp. 7–20. 40 he second mission, departing from Beijing in 1729, included a total of ity-eight people. he third embassy, departing in 1731, had a staf of twenty-three. he leaders of this mission were Manchus (some might have been Mongol); the ethnicity of the other staf is unknown. See Zhang Weihua and Sun Xi, pp. 312–19. 41 On the rise of Manchu literary production in the Kangxi period, see Zhang Jiasheng 張佳生 , “Kangxi chao Manzu wenxue xingsheng de yuanyin” 康熙朝滿族文學興盛的原 因 , Manzu yanjiu (1995.1): 52–57. 42 For Kui-xu’s connections, see Hummel, ed., p. 663. For Singde see Lynn Struve, “he Hsu Brothers and Semioicial Patronage of Scholars in the Kang-hsi Period,” HJAS 42.1 (1982): 254. 43 Zhang Jiasheng, pp. 52–57. Frontier Intelligence in the Qing 161 court and on the frontiers while leaving almost no trace in contemporary Chinese writings. Indeed, this seems to have been the case in the Yongzheng period. Court Interest in Khungghar and Its Strategic Signiicance in the Yongzheng Period In the Yongzheng period, information about the lands west of the Qing frontier was sought with unprecedented vigor by high-ranking Manchu court oicials. Yet almost all our knowledge about this interest comes from the writings of Jesuits and, in one case, a Russian, who were in personal communication with them. A rit in the circulation of information kept Chinese literati largely unaware of Khungghar despite the extensive inquiries about it made by their non-Han colleagues at court and elsewhere, and the appearance of leeting references to it in Mongolian-language chronicles. his section surveys the origins of this new interest in Khungghar against the backdrop of warfare with the Zunghars, and what the Qing court discovered as it pursued knowledge of foreign lands. Ater his irst embassy went to the Torghud, the Kangxi emperor’s interest in lands beyond his western frontier deepened. At the end of his reign he ordered oicials to make inquiries on the Russian frontier about the tsar’s relations with Sweden and the Zunghars, and to get to the botom of a rumor that Ayuki was quarreling with a subject of Khungghar.44 Before his death he ordered Jesuit missionaries to prepare a map of lands between the Qing frontier and the Caspian Sea based on intelligence reports.45 Having inherited these concerns about the western frontier, the Yongzheng emperor in 1725 gave the Jesuits access to some Torghud oicers (oiciers kalmouks) in Qing service and “several Tartar [probably Manchu] itineraries” in order to make a new map of the same region.46 44 Memorial of Rasi [La-xi 喇錫 ] et al. (Oct. 17, 1722), in Kangxichao Manwen zhupi zouzhe quanyi 康熙朝滿文朱批奏摺全譯 (Beijing: Zhongguo shehui kexue chubanshe, 1996), p. 1512. 45 Henri Bernard, “Les étapes de la cartographie scientiique pour la Chine et les pays voisins (depuis XVIe jusqu’à la in du XVIIIe siècle),” MS 1.2 (1935–1936): 466. 46 Joseph Brucker, “La Mission de Chine de 1722 à 1735,” Revue des Questions Historiques, 162 Matthew W. Mosca With the arrival of a Russian embassy to negotiate the Treaty of Kiakhta, the tsar’s political and geographic position relative to West and Central Asian powers became of renewed interest. Early in 1727, Yongzheng’s brother Yin-xiang 胤祥 —that is, Prince Yi (1686–1730)—summoned Jesuits to the palace “to interrogate them for a long time about the borders of Turkey, Persia, and Russia,” and the mutual relations between these empires.47 Soon thereater, they were commissioned to produce a map “of Siberia and Russia up to Petersburg.”48 he interest in Russia’s relations with its neighbors, particularly Sweden and the Muslim empires of Asia, continued ater the Russian embassy had departed. Fr. Antoine Gaubil (1689–1759), a French Jesuit, reported two years later that Yin-xiang continued to ask about the Turks and Persians, and consulted “lamas and other travelers from Kashmir, Ladakh, and the source of the Ganges [in western Tibet],” along with Muslims in Beijing who boasted of the extent of Islamic power even in Europe, a clear reference to the Otomans.49 In 1728, yet another map of Russia’s frontier with “Asiatic Turkey (la Turquie d’Asie), Persia, and Tartary” was commissioned.50 he Jesuits came away with the impression that, during the projected Qing war against the Zunghars, Yongzheng strongly desired Russia to be sidelined by a war on its western frontier. Wrote Gaubil: “hey would very much have liked Persia and Turkey, and even Sweden, to make war on the Russians.”51 In 1729, the Yongzheng emperor began planning a campaign to destroy the Zunghars, who, incidentally, were also well aware of Otoman power.52 As part of his preparations, he sent a second embassy to the Torghud, along with the irst formal delegation to the Russian court. Departing Beijing in 1729 and not returning until 1732, this mis29 (April 1881): 515; Antoine Gaubil, Correspondance de Pékin, 1722–1759, ed. Renée Simon (Geneva: Droz, 1970), p. 173. According to Brucker, these men had arrived via the Syr Darya basin, and were possibly Torghud from the Volga. 47 Brucker, p. 516. Brucker and Gaubil wrote in French; the translations are my own. 48 Gaubil, pp. 171–75. 49 Bernard, p. 467; Gaubil, pp. 236–37. 50 Brucker, p. 516; Gaubil, p. 235. 51 Brucker, p. 516; Gaubil, p. 237. 52 For details of this campaign, see Perdue, pp. 249–53. In 1722 the Zunghar leader Tsewang Rabtan asked the Russian envoy Ivan Unkovskiĭ who, the Turkish sultan or the Chinese khan, was more powerful. his reference has been discovered by Onuma Takahiro 小沼孝博 , “‘Kongga’er guo’ xiaokao: 18 zhi 19 shiji Ou-Ya dongbu Aosiman chao renshi zhi yiduan” “控噶爾國 ” 小考 : 18 至 19 世紀歐亞東部奧斯曼朝認識之一端 , Minzu shi yanjiu 民族史研究 8 (2008): 156. Frontier Intelligence in the Qing 163 sion had two sections: one sent to the Russian court, and the other to ofer imperial greetings to the new Torghud khan, Tseren Dondug.53 In 1731 the Qing court sent a second mission to Russia, but was barred by Russia from sending a further mission to the Torghud.54 In 1732, members of this second embassy asked their Russian escort about “the nature of the Turkish state and the possibilities of a Manchu mission being sent through Russia to Turkey.”55 Based on questions put to him, Fr. Gaubil had also come to suspect that this embassy was relaying to Beijing intelligence “about the Swedes, the Turks, and the Persians.” Gaubil himself was asked by Yin-xiang about the land route linking Europe to China via Turkey and Persia.56 Whether the Qing court was seeking an Otoman connection to solicit aid from Muslims under Zunghar rule, as Mark Mancall has speculated, or whether they wished to use Turkey to entangle Russia in the west, subsequent events completely derailed the Qing court’s strategy. Without Russian consent for their embassies, the Qing court temporarily lost contact with the Torghud. Yin-xiang, the coordinator of court intelligence gathering, died suddenly in 1730; and the Qing loss of a major batle to the Zunghars in 1731 interrupted its eforts to destroy them. Of these two Yongzheng-era missions to Russia scarcely any trace appears in Chinese-language writings from the Qing period, although some account of them was given in unpublished Manchu documents.57 he only surviving evidence of the court’s knowledge of Khungghar in this period is found on oicial maps produced between 1725 and 1730 that cover the entire Anatolian peninsula and part of Otoman Europe. he originals of these maps, which are held in the First Historical Archives in Beijing, are not currently available to foreign researchers, and have not been published in high resolution. On the best available image, of a manuscript map from 1729–1730, the contours of the Black and Caspian Seas are virtually identical to those given on the subsequent Qianlong edition of the map.58 his later map, 53 Zhang Weihua and Sun Xi, pp. 309, 314. See also Mark Mancall, “China’s First Missions to Russia, 1729–1731, Papers on China 9 (1955): 88. 54 Zhang Weihua and Sun Xi, pp. 319–20. 55 Mancall, p. 90. 56 Gaubil, pp. 236–37. 57 Zhang Weihua and Sun Xi cite the Manchu ledger, “Archives concerning Russia” (Eluosi dang 俄羅斯檔 ). 58 To compare these maps, see Aomen lishi ditu jingxuan 澳門歷史地圖精選 (Beijing: 164 Matthew W. Mosca completed around 1760 and printed in copperplate by 1775, shows “the city of Constantinople, seat of the Khan of the Khungghar kingdom” (Hongga’er guo zhi han suoju Gongsidangdinebole hetun 紅噶爾國之汗 所居拱斯當底訥伯勒和屯 ), on the European side of the Bosphorus and terms Asia Minor Du’erjia guo 都爾佳國 .59 Once more we ind the Mongolian term Khungghar, probably from a “Kalmyk” (i.e., Torghud) source, used together with a geographic term, in this case Constantinople, from a Russian or Jesuit source. Indicating the chasm between the two intellectual worlds, Chinese and Inner Asian, of the empire in the Yongzheng period is that Khungghar was virtually unknown to Chinese sources but began to be discussed in Mongol chronicles around this time. he earliest work was the 1725 G’angg’a-yin urusqal (Flow of the Ganges), a genealogy composed by the Mongol aristocrat and Qing oicial Gomboǰab (ca. 1670–1750). According to this chronicle, the fourth son of Chinggis khan’s second son Chaghatai (d. 1242) was “Kungghar, who lived as ruler in the country of Rum and established his capital in the city of Istanbul.”60 As L. S. Puchkovskiĭ has noted, however, Kungghar and the other rulers mentioned in this work were not Chaghatai’s sons, but rather Central Asian rulers of later periods.61 Given that the name Khungghar does not seem to occur in earlier Mongol historiography, Gomboǰab might irst have encountered it during his service in the central government in the Kangxi and Yongzheng reigns.62 Whatever the source of his claim, it was accepted by later Mongol chroniclers, including the 1739 Altan kürdün mingγan gegesütü bičig (Book of the golden wheel with a thousand spokes) by the Tibetan Buddhist cleric Huawen chubanshe, 2000), p. 46; Da Qing yitong yutu 大凊一統輿圖 (Beijing: Quanguo tushuguan wenxian suowei fuzhi zhongxin, 2003), pp. 119–20. Compare also the treatment of the Crimean peninsula in these two maps (pp. 97–98), which both show two cities in the same locations. 59 Da Qing yitong yutu, pp. 119, 139. Hetun is the Manchu hoton, walled city. Dr. H. T. Toh has suggested to me that Du’erjia likely derives from the Arabic Turkiya, possibly via a Persian or Turkish source. 60 Gomboǰab, Činggis eǰen-ü altan uruγ-un teuke g’angg’a-yin urusqal neretü bičig orosiba (Kökeqota: Öbör Mongγol-un arad-un keblel-ün qoriy-a, 1981), pp. 56–57. 61 Puchkovskiĭ, pp. 41–42. I am indebted to James Bosson for drawing my atention to this source. 62 On Gomboǰab’s life and career, see Walther Heissig, Die Familien-und Kirchengeschichtsschreibung der Mongolen, Teil I: 16.–18. Jahrhundert (Wiesbaden: Oto Harrassowitz, 1959), pp. 113–17. Frontier Intelligence in the Qing 165 Dharma and, via that source, the Bolur erike (Crystal rosary; 1774– 1775) of Rasipungsuγ.63 Even in the nineteenth century, both Mongolian and Tibetan-language histories reiterated Gomboǰab’s view that Khungghar was the ruler of Rum. None of these sources seem to have inluenced Chinese historiography or geography.64 In sum, over the course of the Yongzheng reign, Manchu oicials made extensive inquires about Turkey, and Mongol chroniclers began to mention Khungghar as the historical ruler of Istanbul. his frontier intelligence was unavailable to Chinese scholars, however. One reason for this was the convention that oicial Chinese geographic writing could grant atention only to foreign countries that had past or current formal relations with the court. hus, despite appearing on Qing maps, Khungghar, whose connection to Lumi went unrecognized, did not have its own entry in such works as the Da Qing yitong zhi 大清一統志 (Uniied gazeteer of the Great Qing realm) and was mentioned only incidentally in the entry on Russia.65 More signiicant, however, was the dearth of Manchu and Mongol oicials with the literary ability or incentive to transmit their knowledge to an educated Chinese audience. he Rising Prominence of Khungghar on the Frontier and at Court in the Qianlong Period Between 1755 and 1759, the Qing government conquered the Western Regions and stationed soldiers and administrators farther west than ever before. Understanding and safeguarding these new territories required that oicials conduct extensive inquiries about surrounding 63 Siregetü Güüsi Dharma, Altan Kürdün Mingγan Gegesütü Bičig, ed. Walther Heissig (Copenhagen: Ejnar Munksgaard, 1958), p. 27; Walther Heissig, Bolur Erike, ‘Eine Kete aus Bergkristallen’: Eine Mongolische Chronik der Kienlung-zeit von Rasipungsuγ (1774/75) (Peiping: Fu-jen University, 1946), pp. 163–64. For brief biographies of Gomboǰab and Rasipungsuγ, see Christopher P. Atwood, Encyclopedia of Mongolia and the Mongol Empire (New York: Facts on File, 2004), pp. 43–44, 208–9. 64 Inter alia, references can be found in the Subud Erike: A Mongolian Chronicle of 1835, ed. Yang Haiyang (Cologne: International Society for the Study of the Culture and Economy of the Ordos Mongols, 2003), p. 60; and, in Tibetan, in the 1819 Hor Chos-‘byung, in George Huth, ed. and trans., Geschichte des Buddhismus in der Mongolei (Strassburg: Karl J. Trübner, 1892), pp. 19, 29. 65 Qinding Da Qing yitong zhi 欽定大清一統志 , SKQS, vol. 483, 423.5a. 166 Matthew W. Mosca lands, and they were also able to put informal questions to longdistance merchants and other informants for their own ediication. As these eforts to gather intelligence progressed, Khungghar achieved unprecedented prominence. Reports made on the frontier and at court described it as the major power of the Muslim world, and one of the mightiest kingdoms on earth. his section will examine the political background of this rise in Khungghar’s proile, the sources and methods employed to collect information, and the speciic achievements assigned to Khungghar, including military dominance over Russia. Khungghar’s rising prominence can be tracked largely through privately writen descriptions of the Western Regions by Manchu oicials serving there. he earliest of these survives in two manuscripts: one published under the modern title Xiyu dili tushuo 西域地理圖說 (Illustrated explanation of the geography of the Western Regions), and an almost identical unpublished version entitled Huijiang zhi 回疆 志 (Treatise on the Muslim frontier), atributed to a Manchu bannerman, Yong-gui 永貴 , who served as an administrator in Kashgar until recalled to Beijing by an edict in September 1763.66 he identiication of Yong-gui as its author is supported by internal evidence, which suggests that its author was a Manchu bannerman who stopped work on the manuscript around October 1763.67 Even before Yong-gui wrote his treatise, by the time of the Qing conquest of the Western Regions, tales of a vast and rich Muslim city in the west had reached the frontier. As early as 1720, Feng Yipeng 馮一 鵬 had heard in Ningxia of a city among the “turbaned Muslims” with outer walls of brick that took forty-eight days to circumambulate.68 he Qing general Joohūi 兆惠 (1708–1764) was said to have heard durRuan Mingdao, ed., Xiyu dili tushuo zhu (Yanji: Yanbian daxue chubanshe, 1992); Yong-gui, Huijiang zhi, manuscript copied in 1893 by Li Wentian, now in National Central Library, Taibei, item number 210.8 04104. For Yong-gui’s career in this period see Guochao qixian leizheng chubian 國朝耆獻類徵初編 (Taibei: Mingwen shuju, 1985), 138:25.17b–20a. 67 Ruan Mingdao 阮明道 observes from the occasional use of Manchu script to transcribe proper names and frequent mistakes in the Chinese that the author was a Manchu. See his “Youguan ‘Xiyu dili tushuo’ de liangge wenti” 有關《西域地理圖說》的兩 個問題 , in Zhongguo lishi yu dili lunkao 中國歷史與地理論考 (Chengdu: Ba Shu shushe, 2002), pp. 117–20. 68 Feng Yipeng, Saiwai zazhi 塞外雜識 (Shanghai: Shangwu yinshuguan, 1936), p. 15. 66 Frontier Intelligence in the Qing 167 ing his campaigns of a country called Gong (Gong guo 龔國 ) whose copper city walls were ive hundred li in circumference.69 Ji Yun 紀昀 (1724–1805), exiled to Urumqi in 1770, also came across such reports, and skeptically noted: “It is said that the ancestral country of the Muslim regions has walls made of copper. hose in the western Muslim regions say that this city of copper is ten thousand li to their east, while those in the eastern Muslim region say that this city of copper is ten thousand li to their west. Each venerates that which is far away— nobody has ever reached this place.”70 For Yong-gui, this fabulous city was Bukhara. Ater warning his readers that “because there has never been traic back and forth its geography and circumstances cannot be known in depth,” he proceeded to report being told that: he boundaries of its territory are unrivaled. . . . Russia is still its province and owes it an annual tax. he size of the capital city of Bukhara is diicult to measure. It is surrounded by over three hundred gates, each ive days journey from the next. . . . he land is extremely vast, and gold and [precious] stones emerge there in profusion.71 When it came to describing Khungghar, Yong-gui was more restrained: “he territory of the Khungghar (Kongka’er 孔喀爾 ) Muslims, sometimes called Laum [i.e., Rum, here rendered in Manchu script]. . . . Its land is northwest of Bukhara, but since there has never been traic back and forth the minutiae of its geography are diicult to ascertain.”72 Yong-gui’s incomplete manuscript was heavily revised by the Manchu oicial Su-er-de 蘇爾德 , and an account of foreign countries composed in 1779 by a certain Wu-cheng-ge 五誠格 was added to it. Like Yong-gui’s work, information about distant lands was based on hearsay. According to his preface, Wu-cheng-ge spent his spare time on the frontier “repeatedly inquiring of the greatest foreign merchants about the customs, languages, topography, and routes of the lands they passed through; I took notes on the information they translated and Zhao Yi 趙翼 , Yanpu zaji 簷曝雜記 , XXSKQS, vol. 1138, 1.20b. Ji Yun, “Yueweicaotang biji” zhuyi 《閱微草堂筆記》注譯 (Beijing: Zhongguo huaqiao chubanshe, 1994), 12.1150–51. 71 Xiyu dili tushuo zhu, 6.123–24. 72 Xiyu dili tushuo zhu, 6.124. 69 70 168 Matthew W. Mosca gathered them into the ‘Record of Hearsay Concerning Foreign Lands’ (Yiyu chuanwen lu 異域傳聞錄 ) in a single fascicle.”73 In describing Khungghar, Wu-cheng-ge appropriated many of the atributes Yong-gui had assigned to Bukhara. According to his informants, it was extremely rich and populous. In size it eclipsed Russia, which it ruled as a vassal state (shuguo 屬國 ), collecting an annual tribute of ive hundred each of adolescent boys and girls.74 Wu-cheng-ge stated that Rumu 如木 was an alternative name for Khungghar; and he located it four or ive months’ journey northwest of Kashgar, beyond Russia’s western border. his view can be traced back to Tulišen and Feng Yipeng and probably relects the route of the hajj-pilgrimage, the primary link between Central Asia and the Otoman Empire in the seventeenth and eighteenth centuries.75 One of the main routes to Mecca led northwest from Bukhara to Astrakhan, passed through Russian territory, emerged into the Crimean khanate, and crossed the Black Sea to Anatolia before reaching its destination. he entire route took a minimum of about six months, so the igure of four to ive months that Wu-cheng-ge estimates for traveling from Kashgar to Khungghar seems plausible.76 he northwestern trajectory of the irst stage of this journey, and the passage through Russia into Crimean and Otoman territory, together explain the tenacious view that Khungghar was located northwest of Russia. he claim that Russia paid it annual tribute in people might derive from the Otoman practice of devshirme, or forced “collection” of Christian youths into the service of the sultan (although this was a domestic practice and had ceased by the late eighteenth century),77 the tribute Russia once paid to the Crimean khanate, or perhaps the many captives seized by Crimean raids.78 Although the decline of Otoman power accelerated over the course of the eighteenth century relative to the Russian and AustroXinjiang Huibu zhi 新疆回部志 , Siku weishoushu jikan edition, part 9, vol. 7, 4.806. Xinjiang Huibu zhi, 4.811. 75 Suraiya Faroqhi, Pilgrims and Sultans: he Hajj under the Otomans, 1517–1683 (London: Tauris, 1994), pp. 139–42. 76 R. D. McChesney, “he Central Asian Hajj-Pilgrimage in the Time of the Early Modern Empires,” in Safavid Iran and Her Neighbors, ed. Michael Mazzaoui (Salt Lake City: University of Utah Press, 2003), pp. 132–33. 77 Colin Imber, he Otoman Empire, 1300–1650: he Structure of Power, pp. 134–42. 78 Alan W. Fisher, he Russian Annexation of the Crimea, 1772–1783 (Cambridge: Cambridge University Press, 1970), pp. 19–21. 73 74 Frontier Intelligence in the Qing 169 Hungarian empires, it was precisely in this period that Qing sources emphasized the dominance of Khungghar over Russia.79 his perspective was laid out starkly in the most widely read and inluential of the privately composed frontier geographies, the Xiyu wenjian lu 西域 聞見錄 (1777), compiled by Qi-shi-yi 七十一 (zi Chunyuan 椿園 ), who was of the “Changbai clan” and the Manchu Plain Blue Banner.80 Although the author gave no indication of his sources concerning Khungghar, these were presumably hearsay from foreigners (whom he consulted about other countries), or possibly from Torghud or other nomadic peoples. His lengthy account described Khungghar, located in the northwest, as the largest Muslim state. he capital was Wulumu 務魯木 (i.e., Rum), a city “extremely vast; to traverse it north to south by horse would take over ninety days.” his metropolis ruled a network of cities, some with one hundred thousand citizens. Its soil was rich in gold and silver, and poverty was unknown. It had rites “just like China, absolutely incomparable to the bestial behavior of the various nations of the Western Regions.”81 Qi-shi-yi’s account stressed both the military prowess of the Otoman forces and their regional dominance: Russia was originally their vassal state (shuguo) and had been so for a number of years. Ater 1756 Russia ceased to pay tribute and tax. For seven years Russia was not reprimanded but nonetheless atacked Khungghar, which sent forth troops. hey did great batle, and the entire Russian army was overcome and destroyed. . . . he Čaγan Khan [Russian tsar] again raised a levy of one hundred thousand men and also made use of several tens of thousands of crack Torghud troops. Once more Russia did batle with 79 he 1699 Treaty of Karlowitz is widely seen as a watershed in the Otoman military decline vis-à-vis European powers. Rifaat A. Abou-el-Haj, “he Formal Closure of the Otoman Frontier in Europe: 1699–1703,” JAOS 89.3 (1969): 467–75. 80 For biographical details, see Qingchao xu wenxian tongkao 清朝續文獻通考, XXSKQS, vol. 819, 27.244. According to his own account, Qi-shi-yi was born in or around Beijing (Yan), and had traveled throughout the empire. he Qing shilu records that he passed the jinshi exam in 1754. His own account places him in the Western Regions in 1775. Changbai shi 長白氏 seems to be a geographic marker rather than a clan name per se; see Mark C. Elliot, “Manchus as Ethnographic Subject in the Qing,” in Empire, Nation, and Beyond: Chinese History in Late Imperial and Modern Times—A Festschrit in Honor of Frederic Wakeman, ed. Joseph W. Esherick, Wen-hsin Yeh, and Madeleine Zelin (Berkeley: Institute for East Asian Studies, 2006), p. 31. 81 Qi-shi-yi, Xinjiang yutu fengtu kao 新疆輿圖風土攷 (Taibei: Chengwen chubanshe, 1968), pp. 54–55. 170 Matthew W. Mosca Khungghar, and once again it was greatly defeated. he Torghud were thus much afraid, and in 1771, they abandoned Russia and pledged allegiance to China. Several hundreds of thousands of Khungghar frontier troops crossed the border, driving all far before them, and put direct pressure on the Russian capital. he Čaγan Khan was terriied. He acknowledged himself as a subject (chengchen 稱臣) and sued for peace. It was determined that in addition to the ordinary payment, Russia would every year provide a contribution of ive hundred young men and women. . . . Some say that upon the western frontier of Khungghar they have many more subject states, which pay annual tribute just as Russia does.82 his account of Khungghar’s size, power, and wealth was widely read. Later Han Chinese scholars of frontier afairs found these strong claims implausible and in need of correction, but impossible to ignore. he Qianlong emperor himself also made personal inquiries about Khungghar. he earliest such reference I have found occurs in a geographic essay about India of 1768, in which he drew on Buddhist cosmology to describe the world as composed of three “great countries” ranged around Mt. Kunlun on the continent of Jambudvipa: China, India, and Khungghar.83 It deserves note that, in his private writings, Qianlong acknowledged other “great countries” with no suggestion that China was hierarchically superior or that these countries paid him tribute. It is not entirely clear how Khungghar came to occupy such an eminent place in Qianlong’s world view, in the same class as China and historically prominent India. One possible explanation is that Qianlong was inluenced by Muslim Central Asian informants who were in contact with the Qing court during the conquest of the Western Regions. As Onuma Takahiro has recently established through his work in Manchu-language archives, dignitaries from the Kazakhs and Khokand informed the court between 1757 and 1759 that they considered the Qing and Khungghar to be parallel powers, one dominating the east and the other the west.84 Given the Qianlong emperor’s use of the Buddhist geographic framework of Jambudvipa, it is also possiQi-shi-yi, p. 55. Yuzhi wenji 御製文集 , Second Compilation, SKQS, vol. 1301, 21.2b–5a. his essay is dated in Gugong suocang Hendusitan yuqi tezhan tulu 故宮所藏痕都斯坦玉器特展圖錄 (Taibei: Gugong bowuyuan, 1983), p. 9. 84 Onuma, pp. 154–55. 82 83 Frontier Intelligence in the Qing 171 ble that he was inluenced by geographic ideas circulating among highranking lamas at court who were familiar with knowledge current in Tibet. George Bogle, an English envoy to Tibet, found the third Panchen Lama (1737–1780) curious and well informed about both Turkey and Russia, and eager to know the progress of warfare between them.85 In a conversation in November 1774, according to Bogle’s record, the Panchen Lama told him that in Tibetan histories, “Hindustan was formerly reckoned the greatest Empire in the world, next to it [Otoman] Turkey [Rum], and ater that China.” 86 In 1789, the Tibetan geographer ‘Jigs-med-gling-pa likewise wrote that the Mughal and Otoman (Tib. Rum) empires were the two most powerful in the world.87 he Panchen Lama seems to have derived information about the Otoman empire from both Muslim and Torghud informants. In his semi-mythical world geography, entitled Shambhala’i-lam yig (Guidebook to Shambhala) he refers to Otoman lands as “the land of Rum Sham” (Tib. Rum Sham gyi yul, Sham from the Arabic aš-Šām, indicating historical greater Syria)—a term probably used by an Islamic informant. However, when referring to Mecca, then under Otoman rule, he remarked that “the tribes dwelling in the north, such as the Torghud, call it Khung-du-khur padshah or Khung-khur padshah.”88 he reference to the emperor (padshah) of Khungghar makes clear that the Torghud had in mind the larger Otoman empire in which Mecca was situated, not just the city itself. Not long ater listing Khungghar as one of the world’s “great countries,” the Qianlong emperor had the opportunity to learn more about it directly from the Torghud khan, Ubasi (Ch. Wo-ba–xi han 渥巴錫 汗 , 1745–1774). his encounter emerged from the mass migration of the majority of the Torghud from the mouth of the Volga to Qing territory in 1771—an exodus whose origins the Qing government wished to investigate. Because it had occurred during the 1768–1774 war between 85 Alastair Lamb, ed., Bhutan and Tibet: he Travels of George Bogle and Alexander Hamilton, 1774–1777 (Hertingfordbury, U.K.: Roxford Books, 2002), p. 256. 86 Lamb, ed., p. 221. 87 Michael Aris, ‘Jigs-med-gling-pa’s ‘Discourse on India’ of 1789: A Critical Edition and Annotated Translation of the Lho-phyogs rgya-gar-gyi gtam-brtag-pa brgyad-kyi me-long (Tokyo: International Institute for Buddhist Studies of ICABS, 1995), pp. 38–39. 88 Albert Grünwedel, ed. and trans., Der Weg nach Sambhala (Śambalai lam yig) des driten Gross-Lama von bKra śis lhun po bLo bzan dPal ldan Ye śes (Munich: Königl. Bayer. Akad. der Wissenschaten, 1914), p. 59. 172 Matthew W. Mosca Russia and the Otoman Empire, the court’s inquiries drew atention to political developments in western Eurasia, and Khungghar’s role in the Torghud light.89 In a memorial of April 25, 1771, Frontier Pacifying Assistant General of the Right Čebdanǰab (Che-bu-deng-zha-bu 車布登札布 ) and his colleagues relayed to the Qing court a Russian report of the Torghud’s light. hereupon the Qing government dispatched a Mongol oicial to the far north of the Qing domain to gather more information from the leaders of the Uriyangqai banner of Altan Naγur, who frequently visited kin in Russian territory. As it happened, one banner leader had just returned from a trading trip to the Russian Uriyangqai. He informed the Qing oicial that the Torghud had served Russia victoriously in a recent war against the khan of Khungghar, and that Russia had rewarded them by granting their request to return to their old homeland. he high Qing oicials who relayed this version of events considered it disinformation that the Russians had peddled to avoid alarming the tsar’s other Mongol subjects about the Torghud revolt. It seemed highly unlikely to Qing oicials that the Torghud, had they been genuinely victorious, would have declined material rewards from the Russians in favor of enduring a diicult trek back to their original homeland. It was more plausible, according to Qing oicials, that Khungghar had defeated the Torghud, who thereupon absconded to avoid Russian reprisals.90 Further evidence for this opinion was soon provided by the Ili general Iletu (Ch. Yi-le-tu 伊勒圖 ), who heard during inquiries among Kazakh merchants that the Torghud had led for fear that they would have to follow the Russians into batle against Khungghar. If this Kazakh report were accurate, noted the general, then Russian boasts of victory were the complete fabrication that Čebdanǰab had suspected.91 When the Torghud arrived upon the frontier, a diferent version of events emerged. A Torghud Buddhist cleric reported that his people had long wished to lee to the Qing empire but had feared Russian reprisals. Ater winning victories for Russia in two successive campaigns, 89 On the background of the Torghud migration, see Khodarkovsky, Where Two Worlds Met, pp. 224–35. 90 “Chafang Tu’erhute huigui zouzhe xuanyi” 查訪土爾扈特回歸奏摺選譯, Lishi dang’an (1988.2): 36–39. 91 Rescripted 36/5/14. Manwen Tu’erhute dang’an yibian 滿文土爾扈特檔案譯編 (Beijing: Minzu chubanshe, 1988), pp. 9–10. Frontier Intelligence in the Qing 173 they took advantage of the chaos generated by the Russo-Otoman war to escape. Ubasi himself conirmed that his troops won major victories against Khungghar, but added that they chafed at the Russian refusal to divide prisoners equitably.92 In both versions of events, the Russo-Otoman war triggered the Torghud migration; the Torghuds’ diiculty in adjusting to Russian customs and their atraction to the Qianlong emperor’s patronage of the Gelukpa school of Tibetan Buddhism formed the underlying cause. Nonetheless, the Qing court’s public proclamations suppressed the role of the war with Khungghar as a cause of the Torghud migration. Although the Qianlong emperor did not elaborate on the political background of the migration in his personal writings, he included details of the inquiries about Khungghar that he gathered during interviews with Ubasi Khan. For instance, during one meeting Ubasi and other leading Torghud ofered the Qianlong emperor a symbolic tribute of swords. A certain Tu-er-du-bai Khwāja 圖爾都拜和卓 identiied the writing on one sword as “scriptural poetry” (jingshi 經詩 ) writen in the “Muslim script” (here presumably Arabic), praising the sword and proclaiming the invincibility of its bearer. Ubasi recalled that the sword had belonged to his father. hough uncertain of its provenance, he noted: “When my father [Dondug Dashi, r. 1741–1761] was in Russia, men from Khungghar would go back and forth, always bearing goods in tribute. Perhaps this was a sword given by the men of Khungghar.” Qianlong then ordered Yu Minzhong 于敏中 (1714–1779) to make a record of this “Khungghar script.”93 In sum, between 1760 and 1780, Khungghar gained prominence in accounts of the frontier and the lands beyond. It was only ater the Qing expansion into the Western Regions that Khungghar, distinguished by its size, wealth, and military power, was placed in the irst rank of world powers by the emperor and frontier oicials. he Otoman empire and its wars with Russia were known to Qing subjects residing on the Inner Asian frontier, and to merchants and nomadic peoples arriving from abroad; and their knowledge was transmited to the emperor and highranking Manchu and Mongol oicials charged with administering the frontier. Some information was gathered through oicial inquiries, but much came to light through informal inquiries and conversations. 92 93 Manwen Tu’erhute dang’an yibian, pp. 23–24, 109–11. Manwen Tu’erhute dang’an yibian, pp. 132, 156–57. 174 Matthew W. Mosca Circulation of Frontier Information among Han Literati in the Late Qianlong Period At present, studies of the conveyance of Inner Asian information to Chinese readers ater the conquest of the Western Regions concentrate on two types of sources: the prodigious volume of oicial publications issued by the Qing court to describe its new lands; and the poems, travel diaries, and jotings composed by exiled Han oicials.94 Before 1760, Khungghar had been extensively researched at court while remaining almost unknown to Chinese scholars and oicials. Ater 1760, knowledge about the frontier reached the Han literati in unprecedented volume, as the case of Khungghar corroborates. However, the transmission of knowledge about Khungghar draws our atention to a distinct type of source, accounts of the frontier writen privately by Manchu administrators. Oicial surveys, such as the Qinding huangyu Xiyu tuzhi 欽定皇 輿西域圖志 (Imperially certiied illustrated gazeteer of the Western Regions in the imperial domain), restricted their atention to lands like Bukhara or Afghanistan, with whom some contact had occurred during or ater the conquest period.95 Han exiles, who had comparatively short tenures on the frontier and limited contact with foreigners, could not by their own resources produce thorough, gazeteer-like works. By contrast, Manchu (and some Mongol) frontier oicials gained from their administrative experience a more thorough knowledge of Inner Asian conditions than that possessed by exiles. Having become comfortable in writing Chinese in a literary context, men like Qi-shi-yi produced accounts that were free from formal conventions and broached topics, such as Khungghar, largely ignored in oicial works. 94 On the role of Han exiles in promoting frontier studies, see Joanna Waley-Cohen, Exile in Mid-Qing China: Banishment to Xinjiang, 1758–1820 (New Haven: Yale University Press, 1991), pp. 138–62; L. J. Newby, “he Chinese Literary Conquest of Xinjiang,” Modern China 25.4 (1999): 451–74. Both scholars draw on Nailene Chou, “Frontier Studies and Changing Frontier Administration in Late Ch’ing China: he Case of Sinkiang, 1759–1911 (Ph.D. diss., University of Washington, 1976); Dorothy V. Borei, “Images of the Northwest Frontier: A Study of the Hsi-yü Wen Chien Lu,” American Asian Review 5.2 (1987): pp. 26–46. James Millward pays great atention to cartography, court publications, and historical geography in his thorough consideration of the origins of frontier studies, “‘Coming onto the Map’: ‘Western Regions’ Geography and Cartographic Nomenclature in the Making of Chinese Empire in Xinjiang,” LIC 20.2 (1999): 61–98. 95 Huangyu Xiyu tuzhi, SKQS, vol. 500, 46.mulu. Frontier Intelligence in the Qing 175 Although he represents something of a special case, Qianlong’s personal prose and poetry should be included in this category of writings. Whereas formal court compilations were constrained by convention to ignore Khungghar as a non-tributary state, the emperor’s prodigious output of essays and verse contained a number of references to it. Because references to Khungghar in Manchu-language operational documents were not translated in oicial compendia, it is only the Qianlong emperor’s own literary productions that have informed historians—and Han literati in his own day—of the prominent role Khungghar played in his mental world view as a “great country” and “the largest entity within the Muslim regions.”96 Yet Han literati were not simply the passive beneiciaries of a trend among Manchus and Mongols to write more extensively in Chinese about their knowledge of the frontier. Rather, they were active partners, or at least assistants, in this enterprise. Han literati collaborated in recording, polishing, and circulating the indings of their Manchu and Mongol colleagues, with the result that frontier knowledge reached the Chinese-reading public. his collaboration depended on contact between literati and frontier experts in professional or social situations, as well as on a shared interest in Inner Asian afairs on both sides. his section will explore the conditions that enabled Qi-shi-yi’s Xiyu wenjian lu in particular to become an efective medium for spreading information about the frontier. Work on scholarly projects as part of their bureaucratic service led Han literati increasingly to share an interest in the frontier with their Manchu and Mongol counterparts. Chinese oicials rarely served as high-level administrators of non-Chinese regions in the Qing period, but, as the empire expanded, they became increasingly involved in military planning, especially logistics.97 Above all, they became involved in archiving and compiling materials for the campaign histories, geographies, and supporting works that began to appear in large numbers ater the conquest of the Western Regions. At the start of the campaign in 1755, Qianlong irst ordered a gazeteer of the region (provisionally completed in 1761 as the Qinding huangyu Xiyu tuzhi), and then in 1759 Yuzhi shiji 御製詩集 , 4th compilation, SKQS, vol. 1307, 10.14b–15a. For the role of elite scholars in military afairs, see Iona D. Man-Cheong, he Class of 1761: Examinations, State, and Elites in Eighteenth-Century China (Stanford: Stanford University Press, 2004), pp. 179–96. 96 97 176 Matthew W. Mosca added a military chronicle, the Pingding Zhunga’er fanglüe 平定準噶爾 方略 (History of the campaign to pacify the Zunghars). To standardize names used in these works the Qinding Xiyu tongwen zhi 欽定西域 同文志 (Imperially certiied treatise on standardized transcriptions of [names concerning] the Western Regions) was subsequently compiled. hese three works were then thoroughly revised, to achieve their inal form in 1782.98 Meanwhile, the Siku quanshu manuscript project extended eforts to standardize non-Chinese names and terms to earlier dynastic histories, leading to the Qinding Liao Jin Yuan sanshi guoyu jie 欽定遼金元三史國語解 (Imperially certiied explanation of [terms from the respective] dynastic languages in the standard histories of the Liao, Jin, and Yuan dynasties). So profoundly did the Qing military campaigns inluence the work assignments and outlook of oicialdom that Joanna Waley-Cohen has argued that over the course of the eighteenth century the court tried to recalibrate the balance between military and civil values, imbuing Han oicials with martial traits.99 he literary projects emerging from the eighteenth-century war campaigns required deep knowledge of frontier afairs and languages, especially Manchu. To meet this need, many of the Han literati elite who obtained their jinshi degree and were retained for court service took crash courses in Inner Asian languages, topography, and other subjects, either as editors or as clerks in the Grand Council. he primary agency propagating frontier knowledge among Han oicials was the Oice of Military Archives (Fanglüe guan 方略館 ), established in 1749. As Beatrice Bartlet noted, because campaign histories and other compilations had to be prepared largely from undigested central government records, this bureau supervised both archival management and oicial publication. It drew much of its supervision, staf, and materials from the Grand Council, with which it was closely ailiated.100 Senior oicials, evenly balanced between Manchu and Han, super98 Enoki Kazuo, “Researches in Chinese Turkestan during the Ch’ien-lung 乾隆 Period, with Special Reference to the Hsi-yü-t’ung-wên-chih 西域同文志 ,” in Studia Asiatica: he Collected Papers in Western Languages of the Late Dr. Kazuo Enoki (Tokyo: Kyuko-shoin, 1998), pp. 458–71. 99 he argument that the Qing government from the Kangxi through the Qianlong reigns deliberately “recast culture in a more martial mold” is laid out in Joanna WaleyCohen, he Culture of War in China: Empire and the Military Under the Qing Dynasty (London: I. B. Tauris, 2006), pp. 93–97. 100 Bartlet, pp. 225–28. Frontier Intelligence in the Qing 177 vised the compilation of campaign histories, while junior Han bureaucrats did the bulk of the actual research, writing, and organization.101 Court assignments on editorial projects thus fostered specialized knowledge about Inner Asia. Yu Minzhong, who recorded the “Khungghar script” on Ubasi’s sword for the Qianlong emperor, began his career as a compiler in the Hanlin Academy, where he is said to have mastered Manchu exceptionally well, and went on to edit many of the emperor’s personal compositions as well as oicial works dealing with the frontier.102 Wang Chang 王昶 (1725–1806), who became a clerk in the Grand Secretariat in 1758, thereater worked on several multilingual projects including the Xiyu tongwen zhi and a collection of sacred Buddhist incantations (dhārani). Qian Daxin 錢大昕 (1728–1804), one of the greatest scholars of his day and an active court editor in the Qianlong period, had studied Mongolian to some extent for his personal research on Yuan history.103 Chu Tingzhang 褚廷璋 (jinshi 1763), through his work in the Oice of Military Archives on the Xiyu tuzhi and Xiyu tongwen zhi, became (according to the testimony of his colleague Wang Chang) “extremely well versed in the topography of the Zunghar and Muslim regions.”104 Moreover, in his spare time he used materials culled from his editorial duties to write poems about cities in Xinjiang, “in order to supplement items not contained in histories.”105 Liu Xigu 劉錫嘏 (jinshi 1769), who worked at the Hanlin Academy and Oice of Military Archives, gained such luency in Manchu that he became indispensible for standardizing Chinese transcriptions of foreign words required in the histories of the Liao, Jin, Yuan, and Ming dynasties.106 he activities of the Oice of Military Archives, and of the court’s editorial agencies more broadly, expanded the number of Han literati with expertise about frontier afairs. However, because editorial service 101 Yao Jirong 姚繼榮 , Qingdai fanglüe yanjiu 清代方略研究 (Beijing: Xiyuan chubanshe, 2006), p. 68. 102 Wulanqimuge 烏蘭其木格 , “Qing Qianlong chao Hanzu mingchen—Yu Minzhong shuping” 清乾隆朝漢族名臣 —–于敏中述評 , Nei Menggu shifan daxue xuebao (Zhexue shehuikexue ban) 內蒙古師範大學學報 (哲學社會科學版 ) 33.2 (2004): 108. 103 Huang Zhaoqiang 黃兆強 , Qingren Yuan shi xue tanjiu—Qingchu zhi Qing zhongye 清人元史學探究—–清初至清中葉 (Banqiao: Daoxiang chubanshe, 2000), pp. 99–100. 104 For Chu Tingzhang see Guochao qixian leizheng chubian, vol. 149, 129.24a. 105 He-ning 和寧 , Sanzhou jilüe 三州輯略 (Taibei: Chengwen chubanshe, 1968), 8.305. 106 Zuanxiu Siku quanshu dang’an 纂修四庫全書檔案 (Shanghai: Shanghai guji chubanshe, 1997), vol. 1, pp. 485–86. 178 Matthew W. Mosca on campaign-related works was a duty assigned by their Manchu superiors, it is unclear whether Han Chinese scholars were genuinely interested in the material. Nonetheless, evidence about the backgrounds of those scholars who helped prepare Qi-shi-yi’s Xiyu wenjian lu when of duty, or who copied and circulated portions of it (particularly passages mentioning Khungghar) in their personal writings, demonstrates that many scholars who worked on these projects were interested in Inner Asia afairs. Editorial projects opened channels of unoicial communication that made Han bureaucrats aware of Khungghar. Qi-shi-yi began his career by earning a jinshi degree in a class that included such eminent scholar-oicials as Qian Daxin and Ji Yun. he examination system and work in the central government most likely expanded opportunities for making the acquaintance of Han oicials, who then undoubtedly contributed to the success of his work. Chen Kangqi 陳 康祺 (1840–1890) noted that the Xiyu wenjian lu had such a ine style because “probably, it was in reality polished (runse 潤色 ) by the President of the Board of Punishments, Ruan Kuisheng 阮葵生 [1727– 1789].”107 Ruan, according to his near contemporary Fu-qing 福慶 (ca. 1747–1819), contributed a preface to the work and published it.108 Although I have been unable to locate this edition, circumstantial evidence supports Fu-qing’s claim: Ruan included excerpts from Qi-shiyi’s writings—including his account of Khungghar—in his own book of jotings, Chayu kehua 茶餘客話 (Talks with guests over tea).109 How Ruan met Qi-shi-yi is not known, but Fu-qing describes Qi-shi-yi as a staf member of a bureau within a ministry (bucao 部曹 ), suggesting that he spent part of his career working as a mid-level bureaucrat in Beijing.110 Ruan had developed an avid interest in frontier afairs as a byproduct of his court service. Ater earning his jinshi degree in 1761, he 107 Chen Kangqi, Langqian jiwen chubi 郎潛紀聞初筆 (Beijing: Zhonghua shuju, 1997), 7.155. 108 Fu-qing, Yiyu zhuzhici 異域竹枝詞 , Congshu jicheng chubian edition, vol. 3277 (Shanghai: Shangwu yinshuguan, 1937), p. 1. According to Fu-qing, Ruan’s edition was published under the title Xinjiang jishi zhengxin lu 新疆紀實徵信錄 . 109 Ruan Kuisheng, Chayu kehua, XXSKQS, vol. 1138, 13.12a. 110 Fu-qing, p. 1. Bucao was a general term used in the Qing to refer to siguan 司官 or siyuan 司員 , the staf of bureaus (qingli si 清吏司 ) within ministries, and certain other positions. Frontier Intelligence in the Qing 179 irst served as secretary in the Grand Secretariat, and then concurrently as a compiler (zuanxiu guan 纂修官 ) on the Zunghar campaign history and a clerk in the Grand Council. In 1773, he joined the Siku quanshu project and worked on revisions to the Xiyu tuzhi and Xiyu tongwen zhi.111 While on duty he read voraciously in Qing archives to satisfy his own curiosity, even soliciting night shits from his coworkers to peruse them more extensively. His large book collection and wide circle of acquaintances allowed him to pursue extensive personal researches into the Western Regions.112 he same patern of court editorial service holds true for scholars in the Qianlong period who read Qi-shi-yi’s work and incorporated passages into their own writings. Zhao Yi’s (1727–1814) 1796 Nianershi zhaji 廿二史劄記 (Notes on the twenty-two standard dynastic histories) listed Khungghar among the Muslim countries in Asia, citing as his source the Yiyu suotan, a variant name for the Xiyu wenjian lu.113 Starting of in 1756 as a clerk in the Grand Council, Zhao was then transferred to compilation work in the Hanlin Academy before going on to become one of the irst private chroniclers of the Qianlong military campaigns. Another high-ranking oicial, Guan Ganzhen 管幹 珍 (1734–98), also drew on Qi-shi-yi’s account to make reference to Khungghar in a work entitled the Zhifang zhi 職方志 (Treatise on the [geographic records preserved in the] Bureau of Operations).114 Guan was a jinshi of 1766 and worked as a compiler in the Hanlin Academy. He later supervised the compilation of a work on Mongol genealogy that contained some of Ubasi’s remarks about Khungghar.115 Hong 111 Wang Zeqiang 王澤強 , Ruan Kuisheng nianpu 阮葵生年譜 , Huaiyin shifan xueyuan xuebao 淮陰師範學院學報 28.1 (2006): pp. 14–18. 112 Wang Zeqiang, “Luelun Qingdai biji mingzhu Chayu kehua de wenxian jiazhi” 略論 清代筆記名著 ‘茶餘客話 ’ 的文獻價值 , Xibei minzu daxue xuebao (Zhexue Shehui kexue ban) 西北民族大學學報 (哲學社會科學版 ) 2008.2: 113–18. 113 Zhao Yi, Nianershi zhaji, XXSKQS, vol. 453, 34.22a. 114 Guan Ganzhen’s comments are cited by Yu Zhengxie in his Guisi cungao 癸巳存 稿 , in Yu Zhengxie quanji 俞正燮全集 , 3 vols. (Hefei: Huangshan shushe, 2005), 2:6.223. Guan’s original work appears no longer to be extant. 115 his work, the Qinding waifan Menggu Huibu wanggong biaozhuan 欽定外藩蒙古回 部王公表傳 (Imperially certiied genealogical tables and biographies of the Mongol and Muslim aristocracy of the outer regions), recorded that Ubasi had presented a sword to court, stating: “It is said that his ancestor Ayuki received it from Khungghar (Honghuo’er 洪豁爾 ). . . . Since ancient times, it has never had intercourse with China. herefore, it has not been among the tributary countries (bu li Zhifang 不隸職方 ). . . . Ayuki pastured along the Ecil [Volga] River and had intercourse with Khungghar, thus he obtained this 180 Matthew W. Mosca Liangji 洪亮吉 (1746–1809) borrowed from Qi-shi-yi to give Khungghar an entry in his Qianlong fu-ting-zhou-xian tuzhi 乾隆府廳州縣圖 志 (Illustrated treatise on the sub-provincial administrative divisions of the Qianlong era).116 Although he did not become a Hanlin compiler until late in the Qianlong reign, he had worked as an editor at court in earlier decades. Not every reference to Khungghar came from the pen of a court editor: Wang Dashu 王大樞 , who relied heavily on Qi-shi-yi for the account of Khungghar and other foreign countries that Wang gave in his Xizheng lu 西征錄 (Record of westward service), was drawn to frontier afairs ater being exiled to Ili in 1788.117 Nonetheless, the earliest scholars known to have read and copied Qi-shi-yi’s work were chiely Han Chinese oicials connected to the Grand Council and editorial projects. Zhao, Guan, Hong, and Wang seem to have read manuscripts of Qi-shi-yi’s works, but information about Khungghar was also propagated among oicials at court through oral transmission, as when the poet Yuan Mei 袁枚 (1716–1798) heard a story from Umitai 伍彌泰 . A member of the Mongol Plain Yellow banner, Umitai not only held posts in the capital but also served on the frontier, where he met with the Panchen Lama and Torghud envoys, one or both of whom was likely his source of knowledge about Khungghar; there is also some reason to believe that he had visited Russia.118 he story Yuan Mei sword,” SKQS, vol. 454, 102.2b. he sword was presumably given to Ayuki when he formally submited to the Otomans, who bestowed them in ceremonies of vassalage. Swords were presented to Crimean khans together with requests to campaign on behalf of the Otomans (Fisher, pp. 15–16), and a sword was granted to Ya‘qub Beg at the time of his formal subordination to the Otomans in 1873; see Hodong Kim, Holy War in China: he Muslim Rebellion and State in Chinese Central Asia, 1864–1877 (Stanford: Stanford University Press, 2004), pp. 151–52. 116 Hong Liangji, Qianlong fu-ting-zhou-xian tuzhi, XXSKQS, vol. 627, 50.21b. 117 Wang Dashu, Xizheng lu, Guji zhenben youji congkan edition (Beijing: Xianzhuang shuju, 2003), vol. 13, 3.6875–78. 118 Umitai served two stints as an amban (resident Qing oicial supervising local afairs) in Tibet, once between 1756–1759, and again between 1773–1776. During the irst of these appointments he had contact with a Torghud embassy that visited the Dalai and Panchen Lamas; see Zhongguo diyi lishi dang’anguan suocun Xizang he Zangshi dang’an mulu (Man, Zangwen bufen) 中國第一歷史檔案館所存西藏和藏事檔案目錄 (滿、藏 文部分 ) (Beijing: Zhongguo diyi lishi dang’anguan, 1999), pp. 146–47. He had another potential opportunity to hear about Khungghar when he accompanied the Panchen Lama on part of his journey to Beijing in 1779–1780; Qing shi gao 清史稿 (Beijing: Zhonghua shuju, 1976), vol. 36, 323.10823. he circumstances in which he visited Russia were less clear. he scholar Yu Hao 俞浩 cited a work by him entitled “Record of a Mission Frontier Intelligence in the Qing 181 heard from Umitai, although ostensibly about India, clearly relates to Khungghar. Recorded in Yuan’s Zi bu yu 子不語 (hat of which Confucius did not speak; 1781), a collection of fantastic and bizarre tales, it began: “Over four thousand li southwest of Further Tibet there is Wulumu, that is, what the Buddhist Canon terms ‘Central India.’ ” He elaborated that this land was the home of the Buddha, and described its magniicent palaces and abundant gold and silver. Yuan continued: “I have heard that, in the irst year of the Yongzheng reign [1723], Russia wished to seize their land and sent forth over ten thousand troops, driving before them several hundred ferocious elephants to come do batle. he Buddha maintained prohibitory mantras, and sent many thousands of poisonous pythons to go forth and block them. he Russians were terriied, and asked to come to terms . . . [the Buddha] decreed that because this land had few people each decade they should then come and ofer 500 young boys and girls, who would be ordered to mate. To the present it is still thus.”119 Russia’s defeat and consequent payment of ive hundred adolescent boys and girls, along with a reference to Wulumu, puts this story irmly in the tradition of frontier lore about Khungghar. It is possible that Umitai had seen Qi-shiyi’s work before telling this story to Yuan, or that Yuan himself had also read the 1777 edition of Xiyu wenjian lu before puting down Umitai’s report in his Zi bu yu. Yuan Mei was not in government service at the time of the Zunghar campaigns, and how he met Umitai is unknown. Yet it is well documented that he had studied Manchu in 1740 at the Hanlin Academy (his failure to master it resulted in his expulsion), and that he numto Russia” (Shi Eluosi ji 使俄羅斯記 ) in his own 1848 book on frontier history, but this appears no longer to be extant. Yu, presumably on the basis of this work, states that Umitai went to the frontier in 1730 to meet a Russian embassy to discuss frontier afairs. his date falls within the period when Qing envoys visited St. Petersburg, but Yu does not hint that Umitai was on such a mission; Xiyu kaogu lu 西域考古錄 (Taibei: Wenhai chubanshe, 1966), 2:18.14a–b. According to an anecdote recorded by Yuan Mei, Umitai told Yan Changming that he had “gone on a mission to Ele 鄂勒 in the Yongzheng period.” he somewhat fantastic anecdote involves an overland journey to the icy Northern Sea with a party of Westerners (Xiyang ren 西洋人 ). Arthur Waley considers Ele to be a variant transcription of Eluosi 鄂 /俄羅斯 , the standard Qing name for Russia; Yuan Mei: Eighteenth Century Chinese Poet (New York: Grove Press, 1956), pp. 124–26. he Chinese text can be found in Yuan Mei quanji 袁枚全集 , 8 vols. (Nanjing: Jiangsu guji chubanshe, 1993), 4:21.418–19. 119 Yuan Mei, Yuan Mei quanji, 4:21.419. 182 Matthew W. Mosca bered among his patrons the high-ranking Manchu oicials Ortai and Injišan.120 As with those who read Qi-shi-yi’s manuscript, Yuan Mei and other Han oicials who learned frontier lore from Umitai likewise had experience as editors in the Qing court. An example is Yan Changming 嚴長明 (1731–1787), who heard an anecdote about Umitai’s journey into Russia. Ater receiving a juren degree by special decree in 1762, Yan served in the Grand Council and as a compiler of the Zunghar campaign history and geographical works. hrough service in polyglot enterprises, this literatus from Nanjing acquired, according to Qian Daxin, a luent reading ability in both Mongolian (including the Todo script used by the Oirats) and Tibetan. Making use of this ability, the court ordered him to serve in the Sutra and Dharani Bureau ( Jingzhou guan 經咒館 ) and to work on the translation into Chinese of the Mongol chronicle Erdeni-yin tobči (he precious summary; Ch. Menggu yuanliu 蒙古源流 ).121 Umitai was also the source of a story about Tibet recorded by Ji Yun in his Yueweicaotang biji, where he added the encomium that the Mongol “never spoke recklessly in his entire life.”122 Ji had worked extensively on court editorial projects and was briely exiled to Urumqi. Although the circumstances in which these Han scholars met Umitai are obscure, it is clear that encounters occurred in the social milieu of Qing state service. Poetry and literature played a signiicant role in Han-Manchu social exchanges. As Yuan Mei commented, “Recently, Manchus have come exceedingly to surpass Han in their literary pursuits; though he may command an army, none among them is incapable of writing poetry.”123 In sum, the rising interest in intelligence from Inner Asia was closely linked to changing paterns of bureaucratic employment. To succeed in court service, Han scholars had to adapt themselves to working within the framework of large, polyglot court editorial projects, many of which also employed Manchu, Mongol, and Tibetan scholars. To process and digest the oicial archives of a multi-ethnic empire they had to master historical, geographic, and linguistic details 120 J. D. Schmidt, Harmony Garden: he Life, Literary Criticism, and Poetry of Yuan Mei (1716–1798) (London: RoutledgeCurzon, 2003), pp. 14–16. 121 Guochao qixian leizheng chubian, vol. 151, 146.4a. 122 Ji Yun, Yueweicaotang, 6.262. 123 Yuan Mei, Suiyuan shihua buyi 隨園詩話補遺 , in Yuan Mei quanji, 3:7.717. his line is translated more elegantly but less literally in Waley, p. 28. Frontier Intelligence in the Qing 183 about the frontier. he compilation projects generated the irst signiicant cohort of Han literati to grapple intellectually with their dynasty’s vast Inner Asian possessions and the world beyond. For many scholars, professional duty stimulated personal curiosity and prompted them to seek additional information about the frontier. hey consulted not only oicial archives and publications, but also the private writings of frontier administrators. Aiding the transfer of knowledge were personal relationships between Han oicials and their Manchu and Mongol colleagues. As the example of Khungghar demonstrates, these relationships constituted new channels through which frontier information reached the ears and eyes of Han Chinese. Gaps in the information order among the imperial elite were narrowed signiicantly. he Qianlong-Jiaqing Transition and the Qing Information Order he term Khungghar, known on the Qing frontier in the early eighteenth century and familiar to elite Han literati serving at court ater 1760, had, by the early nineteenth century, become a subject of research and debate within the broader Chinese scholarly community. his continuity of interest was sustained despite fundamental changes in the channels circulating frontier-related intelligence, both between Han scholars and Manchu and Mongol frontier oicials, and within the world of Chinese language scholarship. Ater 1800, the low of fresh information from beyond the Inner Asian frontier into the sphere of Chinese scholarship diminished. At the same time, information that had been recorded by court scholars in the eighteenth century became more widely available to the Chinese scholarly readership. In consequence, knowledge about the frontier was pursued in a new way, through textual scholarship. Using the case of Khungghar to illustrate this shit, we will pay particular atention to why it remained on the collective research agenda of eminent Han scholars until the fall of the dynasty. At the end of the eighteenth century, the environment in which frontier studies were conducted began to change. Koon-piu Ho has pointed out that three successive emperors, Kangxi, Yongzheng, and particularly Qianlong, were generally suspicious of private scholarship 184 Matthew W. Mosca and sought to replace it as far as possible with oicial projects.124 Discussions of frontier maters were particularly suspect. he Qianlong emperor’s extreme sensitivity to potential slights against Manchus caused many Ming-era works on frontier management to be banned. Oicial literary surveillance, which was particularly active between 1776 and 1782, watched for the slightest sign of disrespect toward the empire’s non-Han rulers.125 Under these conditions, even loyal Chinese subjects hesitated to publish in a ield so sensitive that a misconstrued passage could have grave repercussions. Ater Qianlong’s death, restrictions were relaxed. Seunghyun Han’s analysis of local gazeteers and similar writings in Jiangnan demonstrates that, although the Qianlong court largely succeeded in silencing private geographic and historical works in the later half of the eighteenth century, an “emancipation” in the Jiaqing (r. 1796–1820) and Daoguang (r. 1820– 1850) reigns allowed both new and older (sometimes banned) works to be published.126 During the Jiaqing reign, the intellectual environment for publishing works related to the frontier likewise grew freer. To illustrate this, the Appendix outlines the publication histories of ity-one privately writen works that describe the empire’s Inner Asian frontier and were composed between 1644 and the Qianlong emperor’s death in 1799. hese works represent, though not exhaustively, a range of genres (poetry, travel records, and comprehensive descriptions). he Appendix excludes oicially ordered gazeteers and books compiled under imperial instruction, whose publication was undertaken by the state. It includes only books composed before 1799, whose publishing histories can relect changing conditions before and ater the Qianlong emperor’s Koon-piu Ho 何冠彪 , “Qingdai qianqi junzhu dui guan si shixue de yingxiang” 清 代前期君主對官私史學的影響 , Hanxue yanjiu 漢學研究 , 16.1 (1998): 155–84. 124 125 R. Kent Guy found the Qianlong-era censorship campaign an efort to “to expunge from the historical record signs of early Sino-Manchu conlict and Chinese disrespect for Manchu custom, heritage and tradition”; see his he Emperor’s Four Treasuries: Scholars and the State in the Late Ch’ien-lung Era (Cambridge: Harvard University Council on East Asian Studies, 1987), p. 166. As Wu Zhefu 吳哲夫 proved, denigrating the Manchus, the Jurchens, other foreign peoples, or Qing rule and policies, were among the major reasons for books to be banned in this period; see his Qingdai jinhui shumu yanjiu 清代禁 燬書目研究 (Taibei: Jiaxin shuini gongsi wenhua jijinhui, 1969), pp. 27–39. Mark Elliot observes, however, that Manchu-related maters were discussed with relative freedom in nineteenth-century jotings (biji); “Manchus as Ethnographic Subject,” pp. 17–37. 126 Seunghyun Han, “Re-inventing Local Tradition: Politics, Culture, and Identity in Early 19th Century Suzhou” (Ph.D. diss., Harvard University, 2005), pp. 94–154. Frontier Intelligence in the Qing 185 death. Of the ity-one titles, twenty-one remained in manuscript until ater 1840. Of the thirty titles that were published, we can consider the number of printed editions (including reprints) of frontierrelated titles that appeared in each of three periods: ten in the Kangxi and Yongzheng reigns; thirteen in the Qianlong reign (to 1799); and thirty-three in the Jiaqing and Daoguang reigns (up to 1840). In short, the irst four decades of the nineteenth century saw the publication or reprinting of about two and a half times as many titles as the sixdecade-long Qianlong reign. A close look at the Qianlong period itself further shows that, of the twelve titles whose publication years we can ascertain, four were published in 1755 or earlier, and six ater 1790. In the span within these dates, which represents both the period of greatest oicial literary activity regarding the Western Regions and the period of most stringent censorship, we ind only two privately printed titles about Inner Asia. In this period substantial manuscript works concerning the Western Regions were produced but not printed. By contrast, in the ive decades ater 1790 we ind thirty-nine printed editions of frontieroriented titles. In the decades ater Qianlong’s death, the widespread availability in print of older works about the frontier, including many now published for the irst time, changed the way information about the frontier circulated and was analyzed. First, it meant that scholars far from the frontier and the court could satisfy their curiosity about Inner Asian lands and peoples. Second, it meant that textual scholarly techniques could be applied to frontier knowledge. In the eighteenth century, court scholars simply copied into their writings information, which itself was oten hearsay gleaned on the frontier, from Manchu and Mongol oicials. By contrast, at that time in the more developed ield of Classical studies, there was a trend toward evidential scholarship (kaozheng 考證 ), which valued the careful analysis of a wide range of textual evidence, oten through collaborative research eforts and well-deined debates. Under these conditions, textual researchers vested authority in scholarship that employed specialized techniques and participated in cumulative scholarly dialogue.127 By approaching Benjamin A. Elman, From Philosophy to Philology: Intellectual and Social Aspects of Change in Late Imperial China (Cambridge: Harvard University Council on East Asian Studies, 1984), pp. 97–99. 127 186 Matthew W. Mosca the frontier exclusively through texts, Han scholars who had never visited the Inner Asian frontier could, ater 1800, become acknowledged experts, and could even take it upon themselves to refute or correct claims made by seasoned Manchu and Mongol frontier administrators. Because of this trend toward text-based research, the perceived value of such works as the Xiyu wenjian lu, which were based on personal frontier experience, steeply declined; and studies that were not based on academic and editorial efort no longer appeared. his limited the low of intelligence from Manchu and Mongol oicials to Chinese-reading scholars. he last example of a Manchu recording fresh information on the subject of Khungghar in Chinese occurred in the Xiaoting zalu 嘯亭雜錄 (Miscellaneous records of Zhao-lian; 1815) of the prince Zhao-lian 昭璉 (1780–1833), who learned about it from a colleague named Bai-shun 百順 . Bai-shun, who held the rank of Commander-general of the Guards Brigade, an elite unit protecting imperial palaces, himself claimed to have visited the frontier of Khungghar, and had been told by an unnamed source that the people of that country claimed to have migrated from the Solon region (in northeastern Manchuria). Zhao-lian speculatively connected them to the group led westward by the Western Liao leader Yelü Dashi 耶律大石 (1087–1143).128 It was not because knowledge of Khungghar disappeared from the Qing frontier that new information ceased to reach Han Chinese scholars. Nineteenth-century writings by Mongols and Tibetans continued to refer to it. Khung-khur and Rumsham were described in a Tibetan historical and geographical work of 1889.129 Even in the early twentieth century, a Russian scholar found that stories about Khungghar 128 Zhao-lian, Xiaoting zalu (Beijing: Zhonghua shuju, 1980), 2.52–53. here are several Bai-shuns, and I have been unable to trace one holding this rank. If we accept his claim to have visited the frontier of Khungghar, then this would seem to connect him with the Yongzheng-era embassies to Russia and the Torghud. Presumably, then, Zhao-lian was told of his claims ater his death. Bai-shun reported that the men of Khungghar were skilled in mounted archery, which might indicate a reference to nomadic Otoman subjects living near the Torghud. Onuma Takahiro speculates that the name Bai-shun might refer to the Qing oicer Shun-de-ne 順德訥 , who in 1757 visited the setlement of Kengger Tura near the Russian frontier. Zhao-lian, in this hypothesis, would then have confused Kengger Tura for Khungghar and represented it as a major country. More evidence would be needed to conirm or disconirm this hypothesis; Onuma, pp. 160–62. 129 Dharmatāla, Chen-po Hor-gyi yul-du dam-pa’i chos ji-ltar dar-ba’i-tshul gsal-bar brjodpa padma dkar-po’i phreng-ba, translated by Piotr Klakowski as Rosary of White Lotuses: Frontier Intelligence in the Qing 187 were still circulating among Mongols around Khobdo.130 Khungghar also continued to come to the atention of Qing frontier oicials. A Manchu memorial reported that Qing border guards had intercepted an Otoman subject trying to cross into the empire from Russia in the Jiaqing period.131 In 1857, the Qing general at Ili reported the statement of an Andijani merchant that “the Russians have recently been ighting back and forth with Khungghar and the English for several years, and they have lost a great amount of territory”—referring here to the Crimean War. 132 he Qing court received, in addition to the memorials mentioned above, ten boxes of Russian books in 1845, whose titles were translated into Chinese as “Campaign History of the Paciication of Khungghar” (Pingding Konggu’er fanglüe 平定空谷爾方略 ) and “An Account of the Country of Khungghar” (Konggu’er guozhi 空 谷爾國誌 ).133 Yet, in contrast to the eighteenth century, the writings of Han literati cite no new evidence about Khungghar from the frontier or the court. In particular, frontier informants of the sort who had been essential to the writings of Qi-shi-yi and other eighteenth-century Manchu oicials ceased to be consulted; instead research came to stress textual sources and standards. Given that new information about Khungghar ceased to low into Chinese, and the credibility of older accounts came under atack, it might be expected that it would be forgoten by Chinese scholars. To the contrary, however, Khungghar came to be studied and discussed by them with unprecedented intensity. here were three intertwined reasons for this. First, a new generation of Han literati emerged who made frontier geography a research priority, despite having no obvious personal or bureaucratic links to Inner Asia. Second, these Being the Clear Account of How the Precious Teaching of Buddha Appeared and Spread in the Great Hor Country (Wiesbaden: O. Harrassowitz, 1987), pp. 49–51. 130 Onuma, pp. 156–57. 131 Jonathan Schlesinger kindly discovered and transcribed for me a Manchu document in the archives in Ulaanbaatar, Mongolia, reporting that a man named Aidimir, a “subject of Kungg’ar” was arrested at a frontier post near Kiatkhta. Interrogated with the aid of “Muslim traders,” he stated that he had been held prisoner in Russia. He found the conditions unbearable and atempted to lee to China along with three Muslim Russian subjects. All four were sent back to Russia. he document (#M2D1–307.4) is dated Sept. 26, 1815. 132 Memorial of Ili General Zha-la-fen-tai 扎拉芬泰 , rescripted Oct. 19, 1857 (XF7/9/2). Chouban yiwu shimo 籌辦夷務始末 , XXSKQS, vol. 416, p. 585. 133 He Qiutao, Shuofang beisheng, XXSKQS, vol. 741, 39.8b–11b. 188 Matthew W. Mosca scholars formed a close-knit community that engaged in its own cumulative scholarly dialogue; repeatedly over the course of decades, they returned to problematic points. Finally, since these Han literati learned about the frontier primarily through published writings, their text-based methodology compelled them to make a detailed analysis even of suspect works like the Xiyu wenjian lu, and reinterpret them through elaborate textual comparisons. For these reasons, the problem of Khungghar’s identity continued to intrigue Chinese scholars for the remainder of the dynasty, as textual methods led them to reach conclusions that were markedly diverse. hus, the critical factor in sustaining interest in Khugghar and the Inner Asian frontier, and in transforming the way these regions were approached, was the training of a new cohort of Han Chinese scholars. he Mongol scholar and frontier administrator Sungyūn (Ch. Songyun 松筠 , 1752–1835) played a pivotal role in preserving a continuity of interest between the scholarly world of the Qianlong period, which favored direct evidence about the frontier, and that of the Jiaqing reign, which preferred textual evidence. Like Qi-shi-yi and Wu-cheng-ge, he took advantage of his contact with foreigners to satisfy his curiosity about foreign conditions. As a biographer commented, “On the day of the arrival of tributary envoys from such countries as the Kazakhs, Kirghiz or Russia, he would call them into his presence and question them on the political situation (zhiluan 治亂 ) in their countries.”134 Assigned to accompany Lord George Macartney (1737–1806) from Beijing to Hangzhou, Sungyūn took the opportunity to increase his knowledge. Macartney found him “a young man of high quality” who “asked many proper questions relative to the riches and power of Russia.”135 Sungyūn recorded the English envoy’s answers in a fu poem entitled “A Concise Record of Frontier Areas” (Suifu jilüe). One interpolated note read: “In 1793, the King of England, which is in the Great Western Ocean, sent an envoy to pay tribute. . . . He had once been stationed on duty in Russia for three years. herefore I made inquiries of him. . . . Russia in the north neighbors the country of Khungghar (Kongka’er guo 空喀爾國 ). It is Muslim. hey commonly ight with Zhao-lian, 4.109. George Macartney, An Embassy to China: Being the Journal Kept by Lord Macartney During His Embassy to the Emperor Ch’ien-lung, 1793–4, ed. J. L. Cranmer-Byng (Hamden, Conn.: Archon Books, 1963), pp. 126–27. 134 135 Frontier Intelligence in the Qing 189 each other over territory. It is passed on that Khungghar is extremely large. It makes city walls out of copper. he eastern and western gates are separated by a journey of some length. I suspected that this was a fable, and inquired about it. Macartney informed me that Khungghar was originally situated in an island in the sea, and trusting to the safety of the water it was as if it had the irmness of a copper wall. his is proof that the narration of the Yiyu lu lacks veracity.”136 When he became Ili general, Sungyūn retained as compiler an exiled magistrate, Wang Tingkai 汪廷楷 (who resided in Ili approximately from 1802 to 1805), but state sponsorship for the project was denied in 1807 on the grounds that the court could instead update the Xiyu tuzhi. Undaunted, Sungyūn privately commissioned a second exiled scholar, Qi Yunshi 祁韻士 (1751–1815), to expand the drat into a book.137 Like other scholars serving at the Qianlong court, Qi had been exposed to frontier afairs when serving as an editorial oicial. A jinshi of 1778, he learned Manchu in the Hanlin Academy and subsequently served as junior compiler (bianxiu 編修 ) in the Oice of State History. His familiarity with Manchu led to an assignment to assist in the Qinding waifan Menggu Huibu wanggong biaozhuan project, where he and another Manchu-trained Han counterpart consulted the oicial survey map of the empire, genealogies preserved in the Court of Colonial Afairs (Lifanyuan 理藩院 ), and piles of Manchu-language routine documents submited by the various Mongol banners.138 Sungyūn’s employment of Qi Yunshi (and later of another excompiler, Xu Song) had a mixed inluence on the circulation of information about Khungghar. Qi Yunshi’s editorial standards, based on the oicial formats used at court, resulted in a work whose coverage 136 he Yiyu lu mentioned here is almost certainly not Tulišen’s account (which does not discuss the size of the walls of Khungghar) but rather Qi-shi-yi’s work, which does. Now generally known as the Xiyu wenjian lu, it had earlier circulated under several titles, some of which began with the term yiyu. For example, Fu-qing, a rough contemporary of Sungyūn, referred to the work as the Yiyu suotan 異域瑣談 . Macartney is probably referring to the peninsular location of Istanbul, or perhaps the Topkapi Palace in particular, with something lost in translation. 137 Enoki Kazuo, “Jo Shō no Seiiki chōsa ni tsuite” 徐松の西域調査について, in Enoki Kazuo chosakushū 榎一雄著作集 (Kyuko shoin, 1992), 2:69–74. 138 A large staf seems to have been working on this project. Presumably, Qi Yunshi and his biographers meant that he was in charge of the day-to-day editorial decisions, while other tasks were carried out by his subordinates. 190 Matthew W. Mosca of foreign lands was far more restricted than Qi-shi-yi’s Xiyu wenjian lu had been. Qi-shi-yi had conceived of his work as a supplement to oicial compilations, reporting accurately on topics they neglected—“the fragmentary afairs of back alleys, and vulgar circumstances of foreign regions.”139 Qi Yunshi preferred textual evidence to hearsay, and criticized the slipshod methods of earlier authors. In particular, he chided previous exiles who cherished “books such as the Suotan and Wenjian lu [both variant terms for Qi-shi-yi’s book] . . . what they record is not free of forced interpretations, factual lapses, and a tendency to like the odd and note down the strange. heir topographies and historical geographies are without factual foundation (wu kaoju 無考據 ) when compared to historical records.”140 Neither Qi Yunshi nor Xu Song, who later succeeded him as editor, sought out hearsay accounts about foreign countries. Disconnected from the Eurasian information circuit, their writings brought no fresh information about Khungghar to the Han literati.141 Although Sungyūn’s project, as executed by Qi Yunshi, limited the scope of inquiry on the frontiers of the Western Regions, it helped promote knowledge of Inner Asia among relatives, friends, and colleagues in China proper. In 1810, Sungyūn was transferred to the governor-generalship of Liang-Jiang, the cultural and academic center of China. Packed in his baggage was a drat of the Xichui zongtong shilüe (Brief account of maters pertaining to the general administration of Qi-shi-yi, p. 3. Qi Yunshi, Xichui yaolüe 西陲要略 (Taibei: Chengwen chubanshe, 1968), p. 1. 141 he work edited by Qi Yunshi, Xichui zongtong shilüe 西陲總統事略 , limited itself to describing the Kazakhs, Kirghiz, and Khokand (Taibei: Wenhai chubanshe, 1965), 11.mulu. he successor to this version was edited by Xu Song and published with imperial endorsement. Although Xu was famous for his irsthand investigation of Xinjiang’s geography during his exile, he speciically excluded from his gazeteer distant countries that were not important for frontier security and describes only the Kazakhs and Kirghiz. Qinding Xinjiang zhilüe 欽定新疆識略 (Taibei: Wenhai chubanshe, 1965), 2:12.1a–2b. Manchus and Mongols continued to write about the frontier, but did so in a consciously scholarly and textual-research oriented fashion. he Mongol frontier oicial He-ning (d. 1821) composed a Huijiang tongzhi 回疆通志 , but this work quoted heavily from existing writings, especially oicial compilations. He omited reference to all foreign countries except the Kirghiz, who were in correspondence with Qing oicials on the frontier: (Taibei: Wenhai chubanshe, 1966), p. 10. he learned Mongol oicial Wo-ren 倭仁 (d. 1871) also mentioned only foreign countries bordering on Xinjiang in his Shache jixing 莎車紀 行 , which likewise cited heavily from textual sources: Wo Wenduan gong (Genzhai) yishu 倭文端公 (艮齋 ) 遺書 , 2 vols. (Taibei: Chengwen chubanshe, 1968), 2:11.31b. 139 140 Frontier Intelligence in the Qing 191 the western frontier), which Qi had completed two years earlier. He sent a copy to Cheng Zhenjia 程振甲 , a former colleague in the Grand Council; Cheng found the work so fascinating that he undertook to edit and print it, together with poems by Qi Yunshi and Sungyūn’s own Suifu jilüe tushi.142 Here again we ind a scholar who had editorial experience in Qianlong’s court promoting frontier studies. he next year, Sungyūn was transferred to the governor-generalship of Liang-Guang. Around the same time, the young scholar Yao Ying 姚 瑩 (1785–1853) accepted an invitation to serve privately in the oice of Guangdong’s provincial education commissioner. His great uncle, Yao Nai 姚鼐 (1731–1815), who had served in editorial posts at court, was an old acquaintance of Sungyūn. hrough this connection Yao Ying met Sungyūn socially and eventually developed an interest in frontier and foreign afairs. Naturally, Yao began by reading the books available to him in Guangdong, but as he worked his way through a mass of oicial geographies, campaign histories, and private works, he noticed numerous contradictions and puzzles. To reconcile these accounts and determine their value he then turned to Sungyūn for help.143 he phenomenon of frontier oicials or court editors successfully propagating their research agendas among private scholars can also be seen in the case of Yu Zhengxie, the irst private Han literatus to take up Khungghar as a textual problem. Yu moved to Beijing and in 1805 became a private assistant to Ye Jiwen 葉繼雯 , an oicial engaged in revising the Da Qing huidian (Collected statutes of the Qing dynasty). Despite his lack of oicial rank, Yu, through this connection, was able to gain access to the Qing archives, and to pursue thereby private geographic and historical inquiries which led him to compose a major essay on the Qing conquest and administration of Mongolia, Xinjiang, and Tibet. he same materials allowed him to write extensively about Russia, completing a study of the Russian banner company in 1806; a general study, “Eluosi shiji” 俄羅斯事輯 (Compilation of afairs pertaining to Russia); and two other essays, “Eluosi changbiangao ba” 142 For a summary of Cheng Zhenjia’s career see Shuyuan jilüe 樞垣記略 (Beijing: Zhonghua shuju, 1984), 18.211. For Cheng’s preface to Sungyūn’s poems, see Xichui zongtong shilüe, pp. 817–18. For the publishing history of this work, see Enoki Kazuo, “Jo Shō,” 2:74–81. 143 Shi Liye 施立業 , Yao Ying nianpu 姚瑩年譜 (Hefei: Huangshan shushe, 2004), pp. 51–53. 192 Matthew W. Mosca 俄羅斯長編稿跋 (Postface to a drat of an extended compilation on Russia; autumn 1806) and “Luocha” 羅剎 ([Explanation of the term] Luocha). hough he had never let China proper, Yu claimed expertise in frontier afairs by “examining oicial and private writings in detail, in order to correct and verify them” (xiangjian guan si zhushu, wei dingzheng zhi 詳檢官私著述 , 為訂證之 ). his method led him to judge Qishi-yi’s Xiyu wenjian lu to be valuable in some ields but “particularly out of accord with reality” in regard to foreign afairs.144 Yu thought the archives and the clerks who kept them, the court survey map, and a host of published geographic texts could bring him closer to the truth. Drawing on Yuan Mei’s story that Russia had once atacked Wulumu (which his reading of Qi-shi-yi led him to identify with the Khungghar khan) and Qianlong’s essay, Yu wove a complex theory arguing that Russia had destroyed Khungghar in the Yongzheng period.145 Flatly refuting Qi-shi-yi, Yu asserted that Russia had again defeated the Otomans in the Qianlong period. As he summarized it, Russia “is in the extreme north, yet some say that the Khungghar khan of its southwestern subject state lies north of it, and they further say that Khungghar can conquer Russia . . . these arguments are all nonsense.”146 Together, the inluence of Yu and Qi stimulated further interest in Khungghar. Qi Yunshi’s son Qi Junzao 祁寯藻 (1793–1866) was related by marriage to a certain Zhang Mu 張穆 (1805–1849). Zhang, shortly ater his arrival in Beijing in 1832, had met Yu Zhengxie, who interested him in the study of historical geography. Qi Junzao later hired Zhang to edit some of his late father’s frontier-related manuscripts. Zhang likely irst encountered the term Khungghar when he helped edit part of Yu Zhengxie’s essays and notes into a work entitled the Guisi leigao 癸巳類稿 (Drats categorized in the guisi year [1833]). As his expertise deepened through his editorial tasks, Zhang himself became intrigued by the problem of Khungghar’s identity. In 1839, he completed a “supplemental collection of afairs” explicitly to correct errors he found in Yu’s work—errors that he identiied by reading the textual sources closely. As He Qiutao 何秋濤 (1824–1862) later observed, most of 144 145 146 Yu Zhengxie, Guisi cungao, in Yu Zhengxie quanji, 2:6.227–30. Yu Zhengxie, Guisi cungao, 2:6.223. Yu Zhengxie, Guisi leigao, 1:9.432–34. Frontier Intelligence in the Qing 193 these “corrections” simply relected Zhang’s reliance on the authority of Sungyūn’s earlier Suifu jilüe. In this case Zhang accepted Sungyūn’s claim that Khungghar was north of Russia; and although he noted that Yu had disputed this conclusion, he did not review Yu’s lengthy argument or explain the process of reasoning that had led him to prefer Sungyūn’s view.147 By 1820, almost all publications about the Qing empire’s Inner Asian frontiers and lands beyond emerged from the private eforts of Han literati. Ater the Qianlong reign, the court both reduced the volume of its publications and relaxed its literary surveillance—two changes that spurred scholars to publish privately. As writen sources about the frontier became more widely available, and textual research techniques became more dominant, claims found in Manchu and Mongol frontier hearsay accounts were no longer taken at face value. Rather, their contents were considered authoritative only if they could be corroborated by Han literati through bookish inquiry. However, private literati interest in the frontier owed much to the patronage and assistance of Manchu and Mongol oicials and Han court editors, who not only imparted their expertise but whose personal connections produced a circle of scholarly interest outside the court. Even as Inner Asian sources ceased to be as widely circulated in Chinese, an interest in Khungghar was kept alive. Scholarship ater the Opium War: Khungghar Interpreted through Western Sources Like the turn toward textual scholarship, the rising inluence of Western sources ater the Opium War (1839–1842) seemed at irst glance to augur ill for the continued relevance of Khungghar; in fact it heightened the interest in the subject. As Western writings translated into Chinese gained in popularity and authority, scholars neither lost interest in Inner Asia nor rejected the earlier Qing sources describing it; rather, they returned, armed with new data, to earlier problems about 147 Zhang Mu, Eluosi shi buji 俄羅斯事補輯 , in He Qiutao, Shuofang beisheng, XXSKQS, vol. 742, pp. 106. 194 Matthew W. Mosca frontier afairs. Han literati now recognized that the meaning of Khungghar was an outstanding puzzle; applying existing textual research techniques to fresh information, they ofered new solutions. he irst scholar to approach Khungghar through newly translated Western materials was Wei Yuan 魏源 (1794–1857). His case illustrates how the research agenda concerning frontier maters that emerged among Han literati at the Qing court was transmited to a new generation of scholars. A chain of events that led Wei Yuan to the subject began with Cheng Tongwen 程同文 (jinshi 1799), a secretary in the Board of War who had served on duty assignment in the Grand Council for over a decade. Like other scholars an editorial assignment—in his case the revision of the Da Qing huidian, in which he edited the section on the Court of Colonial Afairs and the maps of Tibet and Qinghai—prompted him to master not only frontier geography but also Liao, Jin, and Yuan history, evidently learning some Mongolian along the way.148 Cheng passed his interest in frontier afairs on to Gong Zizhen 龔自珍 (1791–1841), the son of his close friend, who became a secretary in the Grand Secretariat. Gong worked briely on the revised imperial gazeteer of the empire, and wrote a number of essays on frontier afairs.149 When Wei Yuan arrived in Beijing as a young scholar, he met both Gong and Yao Ying, who, as mentioned above, had studied frontier geography under Sungyūn. By the time Wei was engaged by He Changling as an editor for his projected Huangchao jingshi wenbian 皇朝經世文編 (Compilation of statecrat writings of the present dynasty; completed late in 1826), he was ready to include his own writings about the Qing frontier alongside those of Chang Tongwen and Gong Zizhen.150 148 According to an entry in the Tongxiang xianzhi, Cheng was “exceptionally strong in his [knowledge of] geographic works, and he always spoke most judiciously about maps of foreign countries and historical and contemporary geographic nomenclature”; Beizhuan jibu 碑傳集補 (Taibei: Mingwen shuju, 1985), 120:7.5b. He wrote works on geography and Yuan history, including Yuanshi yiyin 元史譯音 , the title of which implies some acquaintance with Mongolian; see Fan Kezheng 樊克政 , Gong Zizhen nianpu kaolüe 龔自 珍年譜考略 (Beijing: Shangwu yinshuguan, 2004), p. 32. 149 For a detailed study of Gong Zizhen’s career and relations with Cheng Tongwen, see Guo Liping 郭麗萍 , “Lun Gong Zizhen xibei shidi yanjiu yu Qingdai guanxiu xibei shuji” 論龔自珍西北史地研究與清代官修西北書籍 , Jinyang xuekan 晉陽學刊 (2005.2): 87–91. 150 Huang Liyong 黃麗鏞 , Wei Yuan nianpu 魏源年譜 (Changsha: Hunan renmin chubanshe, 1985), pp. 35–71; Huangchao jingshi wenbian, 3 vols. (Beijing: Zhonghua shuju, 1992), 3:80.1a–81.48a. Frontier Intelligence in the Qing 195 During the Opium War, Lin Zexu sent Wei Yuan the translations of Western-language books and newspapers Lin had commissioned at Canton. Combining these sources with existing materials in Chinese, he completed the Haiguo tuzhi 海國圖志 (Illustrated gazeteer of the maritime countries), which revolutionized the practice of geographic research in late Qing China. At the same time, he also applied information from these new translations to Qing military history, which he reviewed in his Shengwu ji 聖武記 (Record of imperial military campaigns; 1842). In this later book Wei wrote extensively about Russia, and noted the existence of its rival, Khungghar. Following Yu Zhengxie in dismissing the value of Qi-shi-yi’s account, Wei concluded that Russia had in fact always been more powerful than Khungghar. What had happened, in Wei’s opinion, was that the Torghud had become angry with Russia, so that when they arrived on the Qing frontier, “their words all demeaned the Čaγan Khan [i.e., the Russian tsar] and exaggerated Khungghar. . . . Chinese oicials immediately believed this and penned several records. How careless!” Using Western maps to show that no land lay north of Russia, Wei postulated that the country of Tuliya 圖里雅 ,151 which Tulišen had linked to the Khungghar khan, was in fact the place called Pulishe 普里社 , Prussia, in new Western sources.152 his interpretation was later accepted by Wei’s friend Yao Ying.153 In the more famous Haiguo tuzhi, Wei continued to atack what he considered Qi-shi-yi’s credulous errors. He described the supposed Khungghar siege of the Russian capital recorded in the Xiyu wenjian lu as “a nonexistent event; false nonsense.”154 Perhaps the most poignant indication of the parallel decline of Khungghar’s reputation and the authority of eighteenth-century sources was Wei’s emendation of Qianlong’s essay, in which Wei substituted Russia for Khungghar among the triumvirate of “Great Countries.”155 Although Wei took the Shortening this name to Tuliya from Tulišen’s original Tuliyesike serves Wei’s philological purposes in this case, but it is not as disingenuous as it may appear. Yu Zhengxie noted that “sike [i.e. a form of the Russian adjectival suix -skiĭ] is like a Chinese provincial or prefectural seat”; Yu Zhengxie quanji, 1:9.429. Wei was familiar with Yu’s work, and we may assume that he omited the -sike termination on this basis. 152 Wei Yuan, Shengwu ji, XXSKQS, vol. 402, 6.6a–8b. 153 Yao Ying, Kangyou jixing 康輶紀行 , in Siku weishoushu jikan 四庫未收書輯刊 (Beijing: Beijing chubanshe, 1997), part 5, vol. 14, 12.8a. 154 Wei Yuan, Haiguo tuzhi, 3 vols. (Hunan: Yuelu shushe chubanshe, 1998), 3:56.1542. 155 Wei Yuan, Haiguo tuzhi, 1:19.677. 151 196 Matthew W. Mosca time to annotate Qi-shi-yi’s account of Khungghar in detail, he pronounced it to be no beter than a work of iction, devoid of reference value.156 Another early exponent of Western geography, Xu Jiyu 徐繼畬 (1795–1873), also grappled with the problem of interpreting Qi-shi-yi’s work. In his Yinghuan zhilüe 瀛環志略 (Concise treatise on the maritime circuit), Xu considered and rejected Wei’s idea that Khungghar was a European nation. Based on the dates given, he noted, the warfare described by Qi-shi-yi had to be with the country that European sources called Turkey (Tu’erqi 土耳其 ). But then what was “Khungghar”? Xu argued that the Turkish capital was sometimes called Kangsitanyinuoge’er 康思坦貽諾格爾 (Constantinople). he termination -ga’er of Khungghar (Kongga’er in Qi-shi-yi’s Chinese rendering) resembled the termination -ge’er in the Chinese transcription of Constantinople. Xu reasoned that, if the preceding ive characters of the Chinese rendering were abbreviated from Kang-si-tan-yi-nuo into kong, then it appeared that Khungghar was simply short for Constantinople. Xu also noted that Wulumu, said to be the capital of Khungghar, was likely a transliteration of Luoma, Rome. Despite disagreeing with Wei Yuan on other points, Xu concurred with him that Qi-shi-yi’s account of Khungghar’s power was a lie prompted by Ubasi’s animus against Russia, which was then uncritically accepted into the Xiyu wenjian lu.157 Western sources soon penetrated not only studies of foreign countries, but even those concentrating on the Qing empire’s northern and western frontiers. Around 1846, Zhang Mu led the promising Fujianese scholar He Qiutao into the ield of Inner Asian studies, particularly Mongol and Yuan history. He’s interest gravitated to Russia, and in 1857 he completed a massive compendium of all available sources, given the title Shuofang beisheng 朔方備乘 (Complete historical record of the northern lands).158 He Qiutao mentioned Khungghar several times, and concluded that Qi-shi-yi’s account was unreliable.159 Of its treatment of Khungghar he commented, “At that time there was this kind of Wei Yuan, Haiguo tuzhi, 3:56.1544. Xu Jiyu, Yinghuan zhilüe (Shanghai: Shanghai shudian chubanshe, 2001), 4.128–29. 158 For a summary of He Qiutao’s scholarly development see Guo Liping, Jueyu yu juexue: Qingdai zhongye xibei shidixue yanjiu 絕域與絕學 : 清代中葉西北史地學研究 (Beijing: Sanlian shudian, 2007), pp. 260–74. 159 He Qiutao, Shuofang beisheng, XXSKQS, vol. 742, 52.14a. 156 157 Frontier Intelligence in the Qing 197 hearsay, and everyone then set pen to paper and recorded it. hey did not know it was an error.”160 He preferred the explanation put forward by Xu to that by Wei.161 In the last decades of the dynasty, Qing scholars, having an even greater corpus of translated materials at their disposal, developed a more subtle and sophisticated knowledge of European history and geography. his led several scholars to revisit the consensus of Xu Jiyu and He Qiutao that Khungghar indicated Turkey. he new position was evidently irst staked out in the Shuofang beisheng zhiji of Li Wentian 李文田 (1834–1895). In the course of annotating He’s earlier compendium, Li observed, “his Khungghar is Hungary (Xiongyali guo 匈 牙利國 ). Further, it is the Kongga’er country in Qi-shi-yi’s record of things seen and heard.” As he explained, Xu Jiyu, Wei Yuan, and He Qiutao simply had not examined the afairs of Hungary in detail. he books of these three gentlemen all frequently mention maters pertaining to Hungary. Hungary (Xiongyali) is a phonetic representation (duiyin 對音) of Khungghar. . . . Xu Jiyu takes Khungghar to be the capital of Turkey. his is also wrong.162 his dissenting opinion was also held by another Han scholar of the frontier, the diplomat Hong Jun, who, as we have seen, included Khungghar among the researches into Mongol history he started while posted abroad.163 Following intricate kaozheng textual deliberations, Hong concluded that when the Otomans conquered Hungary their ruler must have begun to call himself “Khan of Hungary,” just as Victoria had become “Empress of India.” Earlier Qing authors like Tulišen had misunderstood this and begun to refer to Turkey as Khungghar— that is, “Hungary”—not realizing the origins of the term.164 A similar view was expounded by the late Qing scholar Ding Qian 丁謙 (1843– 1919) who printed an annotated edition of Tulišen’s work in 1915. To explain his reference to Khungghar khan, Ding made the following argument: 160 161 162 163 164 He Qiutao, Shuofang beisheng, XXSKQS, vol. 742, 57.3a–3b. He Qiutao, Shuofang beisheng, XXSKQS, vol. 742, 56.1a–1b. Li Wentian, Shuofang beisheng zhaji 朔方備乘札記 XXSKQS, vol. 742, 1.20a–20b. Hummel, ed., Eminent Chinese, pp. 360–361. Hong Jun, vol. 293, 27下 .3b–4a. 198 Matthew W. Mosca Khungghar was originally the name of a race of people. hey call themselves Magyars (Majia 馬加). In truth, this race of men is descended from the Xiongnu. Westerners call the Xiongnu “Huns” (Henni 狠尼), or Hungary (Xiongyali). . . . Khungghar is a phonetic rendering (zhuanyin 轉音) of this. All lands subject to Turkey . . . belonged to the Xiongnu in ancient times. herefore the Sultan of Turkey holds concurrently the title “Khungghar khan.” It is just like how at present the ruler of Austria concurrently holds the title King of Hungary [lit. “khan of the Magyars,” Majia han 馬加汗], which is an example of this sort of thing. Earlier, commentators were confused in their arguments and none of them made this point.165 hese examples show that at the turn of the twentieth century, despite (or perhaps because of) a greater familiarity with Western geographical works, the mistaken view that Khungghar derived from the word “Hungary” was gaining prominence. he textual turn in frontier research diminished the perceived necessity and usefulness of new geographic and political intelligence from Inner Asia. In efect, though for diferent reasons, the late Qing Han literati were almost as far removed from certain currents of frontier information as they had been in the Kangxi and Yongzheng periods. Increasingly great efort was put into translating and circulating European and American geographic knowledge, but no comparable efort was made to translate and circulate the knowledge and world views of Mongols, Tibetans, Eastern Turkestanis, and other residents of the Inner Asian frontier. Ater the Opium War, Khungghar was viewed as simply a variant way of referring to some foreign country more properly called by a standard name derived from a European source (whether that name was Prussia, Turkey, or Hungary). In court documents and translated Western sources, the Otoman Empire was almost always called “Turkey” ater 1840.166 For instance, the Grand Council remarked in an edict of 1875 to Zuo Zongtang 左宗棠 (1812– 1885) that they had learned that “the cities of Xinjiang . . . border Turkey (Tu’erqi) to the west . . . recently it has been heard that the Muslim 165 Ding Qian, Yiyu lu dili kaozheng 異域錄地理攷證 , in Penglaixuan dilixue congshu 蓬 萊軒地理學叢書 (Beijing tushuguan chubanshe, 2008), 4:20a–b. 166 For instance, in 1854 the Shanghai daotai Wu Jianzhang 吳健彰 transmited reports from foreign merchants explaining the outbreak of the Crimean War between Russia and Deji E� (i.e., Turkey). Other references to Turkish afairs can be found in the Index to Ch’ing Tai Ch’ou Pan I Wu Shih Mo (Hamden, Conn.: Shoe String Press, 1960), p. 716. Frontier Intelligence in the Qing 199 leader of Kashgar [Ya‘qūb Beg] has newly received the enfeofment of the Muslim tribe of Turkey.”167 Khungghar became literally a footnote to studies of history and current afairs, dominated by printed materials and analyzed by textual scholarship. Conclusion he circulation and interpretation of the term Khungghar show three layers of information networks that carried news, about the Otomans and their wars with neighboring countries, from the shores of the Caspian Sea to the core of China. First, there were the people—envoys, merchants, pilgrims, nomads, and missionaries—who carried political intelligence across Eurasia as a by-product of political, economic, and religious activities. Second, there was the information circuit of Qing Inner Asia—frontier residents, Qing administrators, and the court agencies that supervised them—who circulated knowledge via Manchu memorials, Mongol chronicles, Tibetan histories and geographies, and tales told orally. Finally, there was the scholarly world of Chineselanguage publications, ranging from comprehensive gazeteers to short jotings. Writings in Chinese formed the largest and most efective information-propagating tool within the empire, but their authors were largely dependent on non-Chinese informants for information about Inner Asia and more distant places. For information to transit these networks, it had to pass two critical botlenecks: the Inner Asian political boundary and the Chinese cultural boundary. he Qing government possessed limited knowledge about the outside world, especially non-tributary states, but when occasion demanded or permited, both the state and individual functionaries were skillful and tenacious in pursuing foreign intelligence. hrough Manchu and Mongol oicials, like Yin-xiang at court, Qi-shi-yi and Sungyūn on the borders, and Tulišen abroad, knowledge about the outside world entered the Qing empire. From the standpoint of information circulation, it proved more diicult for this knowledge to penetrate the boundary between the Chinese and non-Chinese segments of the empire than it had for it to low across the Inner Asian frontier into the Qing state apparatus. Writings about Khungghar in 167 Qingji waijiao shiliao 凊季外交史料 (Taibei: Wenhai chubanshe, 1985), 1:1.4a. 200 Matthew W. Mosca Manchu, Mongolian, and Tibetan had virtually no inluence on the Han literati, whose bureaucratic careers also largely excluded them from personally experiencing the frontier. Rather, the circulation of knowledge of the lands beyond the frontier depended on the relatively small cohort of Manchu and Mongol frontier administrators who happened to possess literary abilities and ambitions in Chinese. Analyzing the second of these botlenecks in light of the case of Khungghar highlights changes in the wider imperial information order. In the irst phase, before 1750, the indings of Manchu and Mongol frontier administrators were largely inaccessible to Han bureaucrats and literati. In the next phase, between 1750 and 1800, the low of frontier intelligence into Chinese reached an unprecedented volume, as Han literati at court were prompted to seek a deeper understanding of the frontier for both professional and personal reasons. During this period, a large volume of information about Khungghar passed through private writings based on information gained in “of-duty” encounters facilitated by oicial functions. In the Qianlong reign, to a degree never seen before or aterwards, Han literati became privy to current information circulating across Eurasia, with the result that the three information networks became more tightly integrated. he developing patern of communication among Han Chinese scholars in the Qing period resembles in several respects changes in the Song dynasty identiied by Hilde De Weerdt. In both periods there emerged “a literati network disseminating . . . oicial news and archival materials for literate elites regardless of their ranking or membership in the bureaucracy” via “parallel networks” closely linked to oicial channels. hese networks relied increasingly on publishing to reach literati without bureaucratic posts.168 However, the structure of the Qing government added a further dimension to the information order: the Han literati, whether at court or outside it, formed only one circuit of knowledge circulation and depended on the non-Han frontier elite for fresh information. herefore, even though information already available in Chinese came to be propagated more widely among Han literati circles over the course of the nineteenth century, fresh frontier knowledge was not necessarily more available to them. 168 Hilde De Weerdt, “Byways in the Imperial Chinese Information Order: he Dissemination and Commercial Publication of State Documents,” HJAS 66:1 (2006): 145–49. Frontier Intelligence in the Qing 201 Indeed, in the changing political and intellectual context of the early nineteenth century, the transfer of information that had been successful in the previous decades created conditions that disengaged Han scholars from Inner Asian information networks. In the third phase of the circulation of frontier knowledge, ater 1800, a new cohort of literati emerged that was eager to research the frontier even when no high-level court duties demanded it. Relying primarily on writings that were published ater 1790, they came to develop the ield largely independent of court and frontier intelligence. Dismissive of hearsay accounts, they sought to correct older sources using their own scholarly judgment, without the advantage of new information from the frontier. Following the advent of translated Western sources ater the Opium War, scholars like Wei Yuan, Yao Ying, and Xu Jiyu reconsidered further past claims of Khungghar’s prominence; their indings would inluence later experts in frontier geography, such as He Qiutao, Li Wentian, Hong Jun, and Ding Qian. he debates about Khungghar thus remained vital even in the early Republican period. Ater 1800, in addition to the methods of textual criticism adopted by Han Chinese scholars, other factors served to detach Han literati from the frontier information circuit: the return of the Torghud in the 1770s and the rebellions that troubled large tracts of Inner Asia ater the mid-nineteenth century doubtless diminished the low of information into the empire from Central Asia. Moreover, the demand for a standardized geographic vocabulary to coordinate global diplomacy in the late Qing sidelined geographic terms not used in translated European sources. Yet the decrease in new intelligence from the frontier is evident from very early in the nineteenth century, before most of these factors took hold. he impulse to verify knowledge through wide reading led Han scholars to form a closed, autonomous sphere of argument. Although they were receptive to European materials, which were in a systematic, writen form, they no longer wished to hear the oral testimony of Inner Asian informants or the views of Manchu and Mongol oicials who relied on those testimonies. he changing structure of the imperial information order is relevant to the interpretation of Chinese-language sources. On one hand, it must be remembered that, especially in periods when Han literati and oicials were more isolated from frontier intelligence, silence on a topic even in major Chinese reference works or documentary collections is 202 Matthew W. Mosca no guarantee that the topic was not familiar to high oicials or among frontier administrators; such was true of Khungghar in the Yongzheng period. On the other hand, when the Han literati were connected with sources of knowledge about the frontier, they produced writings of unique value. hey were much more active than their Manchu and Mongol counterparts in joting down and preserving intelligence. Similarly, Manchu and Mongol authors seem to have preferred to use Chinese when writing about the frontier, presumably because doing so would give them a wider audience. For this reason, Chinese sources are oten the only place where one can ind information that was obtained or circulated through oral inquiry but was irrelevant for governance. It seems safe to assume that every item of frontier news that made it through formidable barriers to receive even a single citation in Chinese can be taken to have enjoyed a much wider circulation in oral transmission than surviving records indicate. Even ater Manchu and Mongol documents about the frontier have been more fully exploited, then, works like the Xiyu wenjianlu and the biji of court oicials will preserve value not only as records of the circulation of knowledge, but also as repositories of information not found elsewhere. he advantage of studying the wider information order of the Qing empire through the case of Khungghar, as opposed to, say, travel accounts of cities in Xinjiang, is that interest in it was not limited to only one sector of the empire. Other topics, such as the diaries of exiles or poetry about the frontier, would give undue emphasis to the distinctive preoccupations of Han Chinese culture. By contrast, studying how information from Eurasia circulated within the Qing empire ofers a more comprehensive view of the Qing realm as a whole. Because of its empirewide relevance, conclusions concerning the case of Khungghar can form part of a new approach to the study of Qing intellectual and cultural history. When the Manchu Aisin Gioro ruling house brought diverse regions under its rule, the resulting empire was more than the sum of its parts. New paterns of interaction, exchange, and integration were made possible within the Pax Manjurica, reproducing on a smaller but more intensive scale the unprecedented Eurasian linkages permited centuries earlier by the Pax Mongolica. Any analysis of the regions involved must recognize the inluence of supraregional networks and inluences. In this regard, research into the political and cultural history of Qing Mongolia has been particularly Frontier Intelligence in the Qing 203 advanced.169 Scholars of commerce and political economy have also come to study the empire as an interconnected unit.170 A similar commerce in knowledge integrated the empire, blending perspectives and cultures, especially in the Qianlong period, the time of multilingual scholars Gomboǰab and Rasipungsuγ, whom I have discussed above. Undoubtedly the most emphatic proponent of the necessity of polyglot knowledge for scholarly research was the Qianlong emperor himself. A leitmotif in his copious scholarly output was demonstrating to his Chinese subjects that monoglots ignorant of the languages and histories of Inner Asia would profoundly misinterpret their own historical, and indeed Classical, heritage.171 Only a supra-regional analysis can show how individual Qing subjects interpreted the new constellations of information made available to them by Qing rule. Accordingly, to understand the term Khungghar and its inluence, one must examine every major region and language of the Qing realm from 1715 to 1915. One cannot hope fully to grasp Chinese, Manchu, Mongol, or Tibetan perspectives on the world in the Qing period by examining each independently. To understand why Turkey was called Khungghar in Chinese, it is necessary to take an intellectual journey through the major cultural, political, and linguistic spheres of the empire. Each subject’s perspective For instance, Christopher P. Atwood has noted that as early as the eighteenth century, Manchu rule had introduced Chinese tropes of imperial grace into Mongol conceptions of their relations with Qing rulers, creating a single language of loyalty throughout the empire; “‘Worshipping Grace’: he Language of Loyalty in Qing Mongolia,” LIC, 21.2 (2000): 86–139. Johan Elverskog in his Our Great Qing: he Mongols, Buddhism and the State in Late Imperial China (Honolulu: University of Hawai‘i Press, 2006) has also demonstrated the transformative role of Manchu rule in Mongol self-conceptions through “larger intellectual and cultural discourses” (p. 10). 170 For an early, Marxist-inspired view of changing economic relations between diferent parts of the empire, see M. Sanjdorj, Manchu Chinese Colonial Rule in Northern Mongolia (New York: St. Martin’s Press, 1980); for more recent work, see James A. Millward, Beyond the Pass: Economy, Ethnicity, and Empire in Qing Central Asia, 1759–1864 (Stanford: Stanford University Press, 1998); Perdue, especially pp. 303–406, Hua Li 華立 , “Qian-Jia shiqi Xinjiang nanbacheng de neidi shangmin” 乾嘉時期新疆南八城的內地商民 , in Xiyu kaocha yu yanjiu 西域考察與研究 (Urumqi: Xinjiang renmin chubanshe, 1994), pp. 373–90. A forceful recent argument of the role of commerce in binding the Qing empire is Kwangmin Kim. 171 For a representative essay of this type, see his “Wusi-Zang ji Wei-Zang shuo” (Discourse explaining that Wusi-Zang means Wei-Zang), in which he explains that Chinese references to Tibet in standard histories will be misunderstood unless one understands the underlying Tibetan language and its structure; Yuzhi wenji, vol. 1301, 5.1a–3a. Elsewhere, Qianlong argues that the Chinese Classics are not properly understood unless read in Manchu. 169 204 Matthew W. Mosca was shaped by the broader imperial environment in which they functioned. In reference to case of the Otoman Empire Cemal Kafadar argued that it is important to reconstruct the “Otoman point of view,” to preserve the “imperial character” of its history from an “ethnicization” that reduces individuals to narrow compartments.172 In the history of intelligence gathering it is likewise important to recognize a “Qing point of view” that recognizes the intersection between local sources of knowledge and the state structure of the Qing government. Tracing the genealogy of individual references to Khungghar uncovers the intellectual pedigree of an idea that deies classiication into simplistic Manchu-Han, center-periphery, or oicial-unoicial dichotomies. hrough the compounding of hybrid views, the Qing empire looked at the world. 172 Cemal Kafadar, “he Otomans and Europe, 1400–1600,” Handbook of European History, 1400–1600 (Leiden: E. J. Brill, 1994), 1:619–20. Frontier Intelligence in the Qing 205 Appendix Publishing History of Frontier-Related Private Writings Composed between 1644 and 1799 Title Author Date Printings Composed pre 1840s Pre-Qianlong-Reign Works Fang Xiangying 方象瑛 (l. 1667– 1679 1679) [H] 1688 Fengshi Eluosi xingcheng lu Zhang Penghe 張鵬翮 (1649–1725) [H] 奉使俄羅斯行程錄 Fang Gongqian 方拱乾 (1596– 1662 Jueyu jilüe 絶域紀略 / Ningguta zhi 寧古塔志 1666) [H] 1683 Saibei xiaochao 塞北小鈔 Gao Shiqi 高士奇 (1645–1704) [H] Xizheng jilüe 西征紀略 Yin Huaxing 殷化行 (1643–1710) [H] ca. 1695 Zangcheng jilüe 藏程紀略 Jiao Yingqi 焦應旗 (l. 1711–21) [H] 1721 Yiyu lu 異域錄 Tulišen 圖理琛 (1667–1741) [Ma.] 1720 Feng Changbaishan ji 封長白山記 SL, LW, ZD, XH SL, YH SL, ZD SL, KX, ZD SL, ZD 1721 1723, 1724, JY, ZG, ZD, ZH 1734 Xizheng jilüe 西征紀略 Wang Wanxiang 王萬祥 (d. 1702) [H] ? Longsha jilüe 龍沙紀略 Fang Shiji 方式濟 (d. 1717) [H] ca. 1713–17 1755, JY, ZG, ZD Xibeiyu ji 西北域記 / Mei- Xie Jishi 謝濟世 (1689–1765) [H] ca. 1726–30 LW, 1825 zhuang zazhu 梅莊雜著 Qian Liangze 錢良擇 (1645–1707) ca. 1688 JY, ZG, ZD, Chusai jilüe 出塞紀略 ZH [H] JY, ZG, ZH Feng Yipeng 馮一鵬 [H] ? Saiwai zaji 塞外雜紀 CY, ZD Ningguta jilüe 寧古塔紀略 Wu Zhenchen 吳振臣 (b. 1664) ca. 1664–81 [H] ZD Fan Zhaokui 范昭逵 (l. 1719) [H] ? Congxi jilüe 從西紀略 ZD Jin Zang jicheng 進藏紀程 Wang Shirui 王世睿 (l. 1732) [H] ca. 1732 ZD Saicheng bieji 塞程別記 Yu Cai 余寀 (l. 1697–1700) [H] ? ZD Waiguo ji 外國記 Zhang Yushu 張玉書 (1642–1711) ? [H] ca. 1720–21 ZD Zangxing jicheng 藏行紀程 Du Changding 杜昌丁 (l. 1720– 1721) [H] None Song Daye 宋大業 (jinshi 1685) [H] 1696 Beizheng riji 北征日記 Bukui fengtu ji卜魁風土紀 Fang Guancheng方觀承 (1698– ca. 1711–31 None 1768) [H] None Fang Guancheng [H] ? Congjun zaji 從軍雜記 Ding Zang jicheng Wu Tingwei 吳廷偉 (l. 1703–1721) ca. 1720–21 None [H] 定藏紀程 None Lin Benyu 林本裕 (l. 1690) [H] 1690 Liaozai qianji 遼載前集 206 Matthew W. Mosca Title Ningguta shanshui ji 寧古塔山水記 Saibei jicheng 塞北紀程 Saishang zaji 塞上雜記 Xizang kao 西藏考 Xizheng lu 西征錄 Zang jigai 藏紀概 Zang Lu zongji 藏鑪總記 Author Date Printings Composed pre 1840s Zhang Jinyan 張縉彥 (jinshi 1631) ca. 1660–68 [H] Ma-si-ha/ka 馬思哈 /喀 (d. 1704)a ca. 1690 [Ma.] Xu Lan 徐蘭 [H] ? Anon. YZ Wang Zhenxuan 王振翧 (l. 1728) ca. 1728 [H] Li Fengcai 李鳳彩 (YZ period) [H] YZ Wang Woshi 王我師 (l. 1719–1724) ca. 1719–23 [H] None None None None None None None Qianlong-Reign Works (to 1799) Hetao zhi 河套志 Xizang jishu 西藏紀述 Rusai shi 入塞詩 Xizang jianwen lu Chen Lüzhong 陳履中 [H] ca. 1739 Zhang Hai 張海 (l. 1731–1741) [H] ca. 1731–41 Fang Guancheng [H] ? Xiao Tenglin 蕭騰麟 [H] ca. 1737–41 1742 1749 1755 1759 Qi-shi-yi 七十一 (l. 1754–1777) 1777 [Ma.] Ma Jie 馬揭 and Sheng Shengzu 盛 1792 繩祖 [H] Anon. Before 1741 Anon. ? Song-yun 松筠 (1752–1835) [Mo.] 1795 Sai-er-deng 塞爾登 (l. 1736–1750) ? [Ma.] Ji Yun 紀昀 (1724–1805) [H] ca. 1770–71 1777, 1814, 1818, 1837 1792 西藏見聞錄 Xiyu wenjian lu 西域聞見錄 Wei Zang tuzhi 衛藏圖志 Xizang zhi 西藏志 b Xizang ji 西藏記 Xizhao tulüe 西招圖略 Saiwai fengfan cao 塞外封藩草 Wulumuqi zashi 烏魯木齊雜詩 Xizheng lu 西征錄 Wang Dashu 王大樞 [H] Yong-gui 永貴 [Ma.] Shen Zongyan 沈宗衍 (l. 1793– 1795) [H] 蒙古沿革志 San Zang zhilüe 三藏志略 d Shen Zongyan [H] Wei Zang tongzhi He-lin 和琳 (d. 1796) [Ma.] Huijiang zhi 回疆志 Menggu yange zhi 1792 LW 1798 QL JY ca. 1788–99 1814c None ca. 1763 None 1793 None 1795? None ca. 1792–94 衛藏通志 Wulumuqi zhenglüe Anon. ca. 1778 None Anon. ca. 1793 None Su-er-de 蘇爾德 [Ma.] 1772 None Chen Kesheng 陳克繩 (l. 1733) [H] Ge-beng-e 格琫額 (l. 1775–1777) [Ma.] QL 1775 None None 烏魯木齊政略 Xichui jishi benmo 西陲紀事本末 Xinjiang Huibu zhi 新疆回部志 Xiyu yiwen 西域遺聞 Yijiang huilan 伊江彙覽 Frontier Intelligence in the Qing 207 Table Notes: he table divides works into those writen before the Qianlong reign, and those written under his rule (to 1799). Within each of these categories, works are listed in order of printing. Unpublished works are listed alphabetically by title. his list does not claim to be exhaustive and is more thorough for prose works than for poetry collections concerning the frontier. he details provided in the table are derived almost wholly from the following sources: Zhongguo difangzhi zongmu tiyao; Hummel, ed., Eminent Chinese of the Ch’ing Period; Chūgoku shiseki kaidai jiten 中国史籍解題辞典 (Ryōgen shoten, 1989); HOLLIS catalog (hollisclassic.harvard.edu); Catalog of National Library of China (opac.nlc.gov.cn); National Bibliographic Information Network (nbinet.ncl.edu.tw/screens/opacmenu.html). Title column: a slash ( / ) indicates alternate titles. Author column: the author’s ethnicity is indicated within the square brackets with the abbreviations H = Han Chinese; Ma. = Manchu; and Mo. = Mongol. Dates-composed column: a question mark indicates that the date is unknown. When the precise date of composition is unknown, “ca.” indicates the approximate period when the author was on frontier. Here and in the next column, the following abbreviations are used: KX Kangxi reign (1661–1722) YZ Yongzheng reign (1723–1736) QL Qianlong reign (1736–1796) Printing column: For individual printings, the dates are given; where date is unknown, the reign period of printing is indicated. For works published as part of collectanea, the date of printing is indicated by a leter code corresponding to the following abbreviations: SL Shuoling 說鈴 , ed. Wu Zhenfang 吳震方 . Printed 1702–1705; rpt. 1799, 1825. LW Longwei mishu 龍威秘書 , ed. Ma Junliang 馬俊良 . Printed 1794; rpt. 1796. JY Jieyueshanfang huichao 借月山房彙鈔 , ed. Zhang Haipeng 張海鵬 . Printed 1807–1810. YH Yihai zhuchen 藝海珠塵 , ed. Wu Xinglan 吳省蘭 . JQ printing. ZG Zeguzhai chongchao 澤古齋重鈔 , ed. Chen Huang 陳璜 . An abridgement of JY. Printed 1823. CY Ciyantang congshu 賜硯堂叢書 , ed. Gu Yuan 顧沅 . Printed 1830. XH Xuehai leibian 學海類編 , ed. Cao Rong 曹溶 . Printed 1831. ZD Zhaodai congshu 昭代叢書 , ed. Zhang Chao 張潮 (in 1695), supplemented by Yang Fuji 楊復吉 (1776). he sections relevant to this article were irst printed in 1833. ZH Zhihai 指海 , ed. Qian Xizuo 錢熙祚 . Printing began in 1836. Speciic References: a Maska/Ma-si-ka 馬斯喀 appears to be the correct form of his name; Hummel, ed., Eminent Chinese, p. 568. However, several editions of the work render it Ma-si-ha 馬斯哈 . b he authorship of this work, which covers events up to 1741, is unclear. It was printed in 1792 by He-ning; see Zhongguo difangzhi zongmu tiyao 中國地方志總目提要 (Taibei: Sino-American Publishing Company, 1996), 3:24.8–9. c his work contains a preface by Wang Dashu himself entitled Ke Xizheng lu zhiyan 刻西征錄識言 (Note on printing the Xizheng lu), dated 1814. I have not, however, found reference to an actual printed edition. d his work bears a 1795 preface by Shen Zongyan; however, Shen states that it is substantially the work of Yue Zhongqi (1686–1754). Moreover, the manuscript also contains material from the later Jiaqing and Daoguang (1821–1850) periods. It is dated here according to the preface.