Five
Chuang Ts'un-yü and Kung-yang Confucianism

Traditions defy categorization. J. G. A. Pocock notes that a tradition may emphasize the continuous transmission of its teachings. The normative power of past practice can direct human conceptions of reality into acceptable social and political behavior. Traditions that stress continuous transmission become normative models for action and belief.^[1] But a tradition may also stress its creative and charismatic origins. The authority of the original charisma often merges with the mediating chain of transmission, and is thereby "routinized" or "abridged" into an ideology. Hence we can draw a conceptual distinction between traditions that stress the reenactment of the originating creative actions and those that represent themselves as the mere transmission of earlier creative events. But even this distinction is problematic, for transmission and charisma are usually found together. Traditions offer a variety of responses to a given situation. Every tradition is open to diverse public and private responses.

Changes in traditions closely follow the political, social, and economic changes to which they refer. The Confucian ordering of reality, like all traditions, was subject to the inherent tensions between reality and human conceptions of that reality, including perennial moral and political issues. Although governed by the politics of the Confucian state, the responses to these issues were not static. Confucian concepts, interpretations, and verbalizations—and the actions based on them—

[1] Pocock, Politics, Language, and Time , pp. 233-72. See also Edward Shils, Tradition , pp. 3, 10, 54-55, and Schwartz, "Limits of 'Tradition versus Modernity.'"

changed dramatically over time. Identical terms from the Classics could mean different things to Confucians faced with different problems.^[2]

As repositories of the wisdom of antiquity, the Classics of Change, Documents, Poetry, and Rites were regarded by Confucians as creations of the sage-kings. The Spring and Autumn Annals, accepted as the only Classic by Confucius himself, thus became the textual lens through which Confucius's vision of history was interpreted. Although all Classics eventually became embroiled in the New Text vs. Old Text controversy, the task of unraveling Confucius's legacy focused on the Annals.

Ch'ing dynasty New Text Confucians of the late eighteenth and early nineteenth centuries were faced with what they regarded as unprecedented problems. We can therefore understand why they would find the New Text portrait of Confucius painted by the Han Confucians Tung Chung-chu (179?-104? B.C. ) and Ho Hsiu (A.D. 129-182) more to their advantage. Rather than teaching an Old Text tradition based on a chain of classical transmission, followers of the Ch'ang-chou New Text tradition stressed creative and heroic acts.^[3]

As we saw in chapter 4, Chuang Ts'un-yü possessed a compelling, voluntaristic vision of human responsibility for upholding the universal standards of heaven (discovered by the sages). This vision carried over to his remarkable interpretation of the lessons Confucius had encoded in his chronicle of chaos—the Spring and Autumn Annals. Like the Ming dynasty Confucian Huang Tao-chou, who had been a Tung-lin partisan in the 1630s and a Ming loyalist in the 1640s, Chuang saw the Change and the Annals as complementary, embodying a unified political vision.

As with the Change Classic, Chuang Ts'un-yü's views on the Annals

were little known outside his lineage or outside of Ch'ang-chou until Liu Feng-lu's (1776-1829) influential Ch'un-ch'iu Kung-yang ching Ho-shih shih-li (Master Ho [Hsiu's] explication of precedents in the Kung-yang [Commentary ] on the Spring and Autumn Annals Classic ), published in 1805, made it widely known that Ts'un-yü had developed his own views concerning the Kung-yang Commentary.

[2] Pocock, Politics, Language, and Time , pp. 244-45. Shils, Tradition , p. 44, 258. For discussion, see my "Philosophy versus Philology," pp. 175-222, and "Criticism as Philosophy." See also the essays in Hobsbawm and Ranger, eds., Invention of Tradition .

[3] Ch'ien Mu, Liang-Han ching-hsueh chin-ku-wen p'ing-i , pp. 235-36, 264-65, and Lu Yuan-chün, "Ching-hsueh chih fa-chan yü chin-ku-wen chih fen-ho," p. 39. For a discussion see Ku Chieh-kang, "Ch'un-ch'iu te K'ung-tzu ho Han-tai te K'ung-tzu," vol. 2, pp. 130-39. Similar tensions are apparent in the life of Jesus and in Pauline Christianity. See Pelikan, Jesus through the Centuries .

Chuang Ts'un-yü's return to the Kung-yang Commentary must be understood in light of the vicissitudes in Ch'un-ch'iu-hsueh (studies of the Annals ) during earlier dynasties. I shall summarize these studies and then continue with our evaluation of the classical tradition in Ch'ang-chou. I shall also address Chuang Ts'un-yü's scholarly contributions to studies on the Change Classic and Annals. As in the case of I-hsueh, eighteenth-century studies of the Annals were influenced by the Han Learning research agenda and the application of k'ao-cheng methods to resolve the textual problems in interpreting it and its three orthodox commentaries.^[4]

Uses of the Annals

During the Former Han (206 B.C. —A.D. 8) heyday of New Text Con-fucianism, a consensus emerged that saw an implicit code of good and evil in the Spring and Autumn Annals, a code that Confucius had included in his chronicles of Lu. "Praise and blame" (pao-pien) histo-riography were stressed by both of the Former Han New Text Kung-yang and Ku-liang commentaries, although the former became the touchstone for New Text moral evaluation. With the eventual rise of the Old Text tradition of the Later Han (A.D. 25-220), this praise-and-blame tradition was incorporated into Old Text interpretations of the Classics as well. By the T'ang dynasty Confucians had forgotten that "praise and blame" historiography had been a New Text creation and was accepted for the most part with little questioning.

Later Confucians believed that Confucius, by assuming the historiographical prerogatives of the Chou king and speaking for him through the chronicles of Lu, had adapted each entry to express model judgments on every event and participant he recorded. The chronicles of Lu, which encompass the 254 years between 722 and 468 B.C. (when the Old Text Tso chuan halted), became the basis for the traditional histo-riography of ancient China. Tradition, classics, and historiography were indistinguishable.^[5]

Praise-and-blame doctrines carried over from the Annals to other Classics as well. From the Han until the Sung dynasties, for example,

[4] See Huang Tao-chou, "Fan-li" (statement of contents). Cf. Sato^[*] , "Shincho^[*]   Kuyogakuha^[*] ," pp- 20-25. For a discussion see my Philosophy to Philology , pp. 204-29.

[5] For Mencius's views of the Annals , see Meng-tzu yin-te , 25/3B/9. Cf. Legge in Four Books , pp. 676-77, and Lau, trans., Mencius , p. 114. Ssu-ma Ch'ien praised the Annals in his Shih-chi, 6/1943 (cbüan 47). Cf. Bodde's translation of Fung Yu-lan, History ofChinese Philosophy , vol. 1, pp. 45-46, 400-403. See also Chou Yü-t'ung, Ching chin-ku-wen hsueh , pp. 22-27, and Ku Chieh-kang, Han-tai hsueh-shu-shih lueh , pp. 62-72.

commentators read political intent into many of the poems collected in the Poetry Classic (Shih-ching ). Poetic omens were first articulated in the Former Han, and during the Later Hah Cheng Hsuan (127-200) drew explicit parallels between the rise and fall of a dynasty and the vicissitudes in its poetry. Society and literature reflected the same moral judgments. Constant relearning of these ancient patterns of behavior offered Confucians conceptions of the right order of society and appropriate rituals (li ) for proper conduct in both public and private life. Images of the Confucian past, accordingly, were stereotyped and transmitted through the moral prism used by Confucius to interpret the chronicles of Lu.^[6]

The bloody details of the Spring and Autumn Period were judged normatively as precedents. For both New Text and Old Text Confucians, tradition, as it was related to the venerable Confucius, became synonymous with correct historiography. Although tensions between the New Text and Old Text manipulations of the Spring and Autumn Annals never disappeared, the two traditions contributed in different proportions over time to political discourse in imperial China.

For instance, Tu Yü (222-284) wrote an influential work entitled Ch'un-ch'iu shih-li (Explications of precedents in the Spring and Autumn Annals ). A systematic list of historical precedents, the work was based on an Old Text reading of the Annals drawn from the Tso chuan. These precedents were devised to give the Annals the kind of political specificity that the Rituals of Chou spelled out for institutions. The classification of Confucius's moral censures of historical events in the Annals allowed Tu Yü and his Confucian successors to categorize historical events. Forty-two general categories for precedents (e.g., legitimate political succession, proper forms for meetings and alliances, appropriate military campaigns, etc.) were used to organize the multiplicity of names and events in the Annals . Executive policy (hsing-fa ) drawn on precedents in the Annals could now compete effectively, or be used in tandem, with the institutions of the Rituals of Chou . Old Text scholars by now had reinterpreted what was left of New Text Confucianism.^[7]

Prior to the T'ang dynasty (when Mencius was "rediscovered" by

[6] Shils, Tradition , pp. 54-55. See also Chia-ying Yeh Chao, "Ch'ang-chou School," p. 159, and James J. Y. Liu, Chinese Theories , pp. 64-65.

[7] See original preface, Tu Yü, Ch'un-ch'iu shih-li , by Liu Fen. Ch'ang-chou New Text scholars also made extensive use of the Chou-li . It was not until Sung Hsiang-feng (see chapter 6) that the Chou-li was categorically dismissed by New Text scholars.

Han Yü) Confucius and the Duke of Chou ("Chou-K'ung") were regarded as the axis of ancient Confucianism. During the T'ang, however, Confucius and Mencius ("K'ung-Meng") became the axis. The T'ang historian Liu Chih-chi even dismissed the Annals as little more than a cursory chronology. For Liu, Confucius had simply edited earlier historical records with no intention of developing a systematic framework for legal and moral precedents. He questioned the usefulness of the Kung-yang Commentary and dismissed the "praise and blame" tradition associated with the New Text Kung-yang and Ku-liang commentaries. Instead, he favored the Tso Commentary because of its detailed historical information.

In many ways, Liu's approach represented a clean break with the orthodox view of Confucius and the Annals. Enunciated with imperial support, this orthodox view was fashioned by K'ung Ying-ta (574-648) and his staff into T'ang commentaries on the Five Classics prepared as state guidelines for the civil service examinations. Later, Tan Chu (725-70) reduced the authority of all "three [Han] commentaries" to the status of mere amplifications of orally transmitted traditions that had lost touch with the original Annals Classic. More radical scholars such as Chao K'uang (fl. eighth century A.D. ) went further and, on the basis of technical and stylistic considerations, rejected Tso Ch'iu-ming as the author of the Tso chuan. Chao questioned the canonical status of the commentary.^[8]

During the Northern Sung dynasty (960-1126), when Wang An-shih sought precedents for his reform party, he appealed to the Rituals of Chou as the classical text that contained the actualized political ideals of the sagely Duke of Chou. To lend credence to this Old Text Classic, Wang belittled Confucius's Spring and Autumn Annals as a useless collection of dates and names. He eliminated them as a special field of preparation for the imperial examinations, purportedly claiming the Annals was "worthless fragments of a government bulletin." He favored the Rituals because its detailed plans for government institutions justified a more assertive government—which in turn justified the stricter control and regulation of society. As an institutional reformer, Wang An-shih placed little store by the moral principles read into the Annals.^[9]

[8] Liu Chih-chi, Shih-t'ung (shih-p'ing ) pp. 381-91 (wai-p'ien no. 5), and Inaba, "Chu -To^[*] , ni okeru shin Jugaku undo^[*]   no ichi kosatsu^[*] ," pp. 377-403. See also Pulley-blank, "Neo-Confucianism and Neo-Legalism," pp. 88-91, and P'i Hsi-jui, Ching-hsueh li-shih , p. 264.

[9] For Wang An-shih's remarks, see Chou Yü-t'ung's notes in P'i Hsi-jui's Chinghsueh li-shih , pp. 29-30, and P'i's own discussion on p. 250. Cf. James T. C. Liu, Reform in SungChina , pp. 30-33.

Conservatives led by Ssu-ma Kuang (1019-86) and Su Shih (1036-1101) rejected Wang An-shih's efforts to create new and expansive institutions as politically unfeasible and classically inaccurate. They maintained that Confucius's vision of social order and the conflict between good and evil, developed in his Annals , were the true lessons of the Classics. In addition to proposing the reform of bureaucratic procedures, they advocated a system of moral censure that would invigorate the government and complement political reform as the key element in effective government. Moreover, many Northern Sung scholars maintained that the Rituals of Chou itself was a questionable text, associated with Liu Hsin's (45 B.C. —A.D. 23) dubious efforts to legitimate Wang Mang's (r. 9-23) usurpation. Some claimed that the Chou-li had originally had little to do with the Duke of Chou and in fact may have been first composed during the Former Han dynasty.^[10]

In contrast to Wang An-shih's efforts to dismiss the Annals, the conservatives saw evidence in the Annals for classical models that lent dynastic prestige to the ruler. Following the lead of their T'ang precursors, Sung dynasty scholars of the Annals proposed that the "three commentaries" (Kung-yang, Ku-liang, and Tso ) be discarded in favor of following the text of the Classic itself (ch'i-chuan ts'ung-ching ). This approach freed T'ang and Sung scholars of the Annals from the classical authority of Han interpretations and permitted them to explicate the Annals ' "general meaning" (ta-i ). Ancient precedents were directly referred by analogy to contemporary political problems.

After the failure of the Wang An-shih's reform program, Chu Hsi and other Neo-Confucians subordinated the Rituals of Chou and Annals to the Analects and Mencius. Of the Four Books these two were said to provide a more intimate understanding of the sageliness of Confucius. Despite the Neo-Confucian stress on Confucius and Mencius, however, Chu Yuan-chang the founder of the Ming dynasty (r. 1368-98), utilized the Rituals to legitimate his triumph over the Mongol Yuan dynasty (1280-1368). In particular, he saw in the Rituals a classical basis for restoring pre-Yuan institutions, thereby affirming

[10] Uno, "Shurai^[*] Ryu^[*]   Kin gisaku setsu ni tsuite." See also Yang Hsiang-k'uei, Ching-shih-chai , pp. 267-74, and Yeh Kuo-liang, Sung-jen i-ching kai-ching k'ao , pp. 97-109. Cf. Yoshiwara, "Hoku-So^[*]   Shunju^[*] gaku no ichi sokumen," p. 633

through both ritualistic and political means the governmental structure of the Chou dynasty (1122?-221 B.C. ). Like Wang Mang and Wang An-shih before him, Chu Yuan-chang relied on Chou models to add classical legitimacy to his dynastic policies.^[11]

During the Northern and Southern (1127-1279) Sung dynasties, "honoring the ruler" (tsun-wang ) and "driving out the barbarians" (jang-i ) became particularly important concepts in contemporary explications of the Annals. Such themes were clearly a reflection of threats from northern "barbarians" that materialized with the collapse of the Northern Sung in 1126 and the formation of the Southern Sung. Until the Mongol triumph in 1280 Sung rulers and ministers remained preoccupied with the "barbarian problem."

For many Confucians, such as Ch'en Liang (1143-94), the need to drive the Jurchen from north China and to restore it to Sung dynasty rule became a political obsession during the Southern Sung. Hoyt Tillman has pointed out that traditional Confucianism played an important role in Ch'en Liang's patriotism. Ch'en invoked Confucius and his Annals as the proper guide for the "way of civilized people" (jen-tao ), in contrast to that of barbarians, who violated Chinese norms for ritual, social, and ethical behavior.^[12] Ch'en Liang's own sense of national peril led him to enunciate, in a particularistic fashion, the lessons Confucius had encoded in the Annals. According to Ch'en's reading of the Annals, one such lesson was that Chinese and barbarians had separate Ways (Tao ), which should not be mixed. Accordingly, he attributed the decline of the Chou dynasty to its failure to keep barbarians out of the Chinese heartland. In Ch'en Liang's hands, the Annals became a handbook on barbarians.^[13]

In an influential work entitled Ch'un-ch'iu tsun-wang fa-wei (Bringing to light the honoring of the ruler in the Spring and Autumn Annals ), Sun Fu (992-1057) had compared the barbarian threats posed to the Sung dynasty with the chaos in the time of the Annals. Just as the con-

[11] Ch'ien Mu, Liang-Han ching-hsueh chin-ku-wen p'ing-i , p. 265, and Yang Hsiang-k'uei, Ching-shih-chai ,pp. 149, 249-51. See Chü-chieh Huang, “Old Pursuits and New Knowledge," 211-12, and Huang’s "Mencian Morality." See also Kung-chuan Hsiao, Rural China , pp. 43-46, 569 n. 6, and Yun-yi Ho, Ministry of Rites , pp. 52, 63.

[12] For a discussion see Ch'en Ch'ing-hsin, "Sung Ju Ch'un-ch'iu tsun-wang yao-i te fa-wei yü ch'i cheng-chih ssu-hsiang." Cf. Tillman, Utilitarian Confucianism , pp. 31, 108, 166-67, and Tillman, "Proto-Nationalism in Twelfth-Century China?" 403, 423.

[13] Ch'en Liang, Lung-ch'uan wen-chi , 4.5b-6b. See Tillman, "Proto-Nationalism," pp. 410-11. See also Langlois, "Spring and Autumn Annals in Yuan Political Thought," pp. 124-25, and P'i Hsi-jui, Ching-hsueh li-shih , p. 250.

tentiousness of the feudal lords (chu-hou ) threatened, and eventually brought down, the Chou dynasty, so the barbarian threats of the eleventh century imperiled the Sung dynasty. What was required, Sun contended, was a strengthening of the imperial institution. "Honoring the ruler" was his proposed doctrinal remedy to end the dynasty's internal divisiveness and to unite the empire.

Sun Fu also called for "driving out the barbarians." It is interesting that Sung Confucians saw in the Annals principles for behavior that governed relations with barbarian dynasties. The stability of the Northern Sung dynasty clearly depended on successful relations with the Khi-tan and Jurchen tribes, which were slowly but surely encroaching on the north China plain. Sun Fu and his followers, however, took a hard line. Contacts and alliances with the barbarians were frowned upon. They called instead for military action to attack the intruding "barbarians" and to defend the dynasty.^[14]

"Imperial majesty" (wei ), designed to overawe foreigners, was now stressed, and protonationalism became a rallying point for antibarbarian sentiments. Chinese culturalism (based on Confucian culture rather than on particularistic appeals to the dynasty) did not automatically restrict the development of nationalism in the imperial state. The ambiguities of patriotism and nationalism during the Sung dynasties permitted interesting forms of protonationalism to develop. The Annals provided a convenient framework for the articulation of such sentiments and for the tributary principles they required.^[15]

Hu An-kuo also composed an influential commentary to the Annals that elaborated on Sun Fu's twin themes of "honoring the ruler" and "driving out the barbarians," which became slogans of the hard-liners on the "barbarian question." They tried to rally support around the Sung emperor and then to retake lands in north China already lost to the Khitan (Liao, 947-1125) and Jurchen (Chin, 1115-1234) dynasties. Like Sun Fu, Hu An-kuo thought the plight of the Sung dynasty was brought on by moral failure and the betrayal of values encoded in

[14] Sun Fu, Ch'un-ch'iu tsun-wang fa-wei , 1.1a-2a, 1.13a-13b, 1.16a-16b, 2.3b, 12.8a-8b. For a discussion see Mou Jun-sun, "Liang-Sung Ch'un-ch'iu-hsueh chih chu-liu, shang," 113-15. Cf. P'i Hsi-jui, Ching-hsueh li-shih , pp. 250-51.

[15] Tillman, "Proto-Nationalism in Twelfth-Century China?" See also Trauzettel, "Sung Patriotism." During the last years of the Tokugawa shogunate (1600-1867) in Japan, the rallying cry for dissident samurai who wished to overthrow the Tokugawa house was "sonno^[*]   joi^[*] " (to revere the emperor and expel the barbarians). This explosive combination in Japan drew from Chinese historical experience and indirectly from the Sung interpretation of the Annals . See Najita, Japan , pp. 43-68, and Earl, Emperor and Nation, pp. 82-210.

the Annals. After the Spring and Autumn period, according to Hu, Chinese and barbarians had culturally mingled to the detriment of both.

The broader question of the emperor's authority and power led Hu An-kuo and others to call for revenge (fu-ch'iu ) against the barbarians when north China fell in the twelfth century to the Jurchen, ancestors of the Manchus. Hu's patriotic reading of the Annals was so powerful that it inspired the young Lu Chiu-yuan (Hsiang-shan, 1139-92)—later Chu Hsi's main intellectual antagonist—to resolve at age sixteen to become a soldier and to help drive the barbarians out of the Chinese heartland in the north. Classical scholarship and protonationalism were interwoven to form an antibarbarian political position; it proved unsuccessful in the long run.

Hu An-kuo's commentary on the Annals became so prominent after the fall of the Sung dynasty that when Chinese Han rule was restored in 1368, it was informally known as one of the "four commentaries" (ssu-chuan ), standing on a par with the orthodox "three commentaries" dating from the Han dynasties. Leading Confucians perceived the need for imperial authority to ensure domestic solidarity and to rally against external threats. Sung Confucians were trying to exalt imperial power, but they were not advocating its unlimited power. But their intentions belied the Legalist consequences when Yuan and Ming emperors appropriated Sung political discourse^[16]

Uses of the Annals carried over in the notion of legitimate succession (cheng-t'ung ), developed during the Northern Sung. The criteria of legitimate succession began to shift from theories of yin-yang and the five evolutive phases (which Former Han New Text Confucians such as Tung Chung-shu had used in their political cosmology) to issues of moral right and political unification. Ou-yang Hsiu (1007-72) and Chu Hsi formulated a theory of legitimate succession that eventually became a prominent principle of Chinese historiography.

Ou-yang Hsiu's official histories of the T'ang dynasty and Five Dynasties period (907-60) left him sensitive to the ambiguities of writing on periods when political legitimacy was contested. He was also the first important historian of the post-T'ang period to reject earlier New Text theories that drew on the portents and the evolving configurations of the "five phases" in determining legitimate succession. Instead, Ou-

[16] On Hu An-kuo's commentary see Mou Jun-sun, "Liang-Sung Ch'un-ch'iu-hsueh chih chu-liu, hsia," 170-72. See also Sung Ting-tsung, “Sung-Ju Ch’un-ch’iu jang-I shuo"; Hervouet, ed., Sung Bibliography ,pp. 39-40; and Schirokauer, “Neo-Confucians under Attack," pp. 165-66.

yang measured degrees of political unification as the key determinant of legitimacy. His single political criterion provided a fixed and universal historiographical guideline for evaluating changes in dynasties.^[17]

Ou-yang's views were fleshed out by Chu Hsi in his Tzu-chih t'ung-chien kang-mu (Outline of the Comprehensive Mirror for Aid in Government), published in the late twelfth century. Although Chu Hsi thought the theme of kingship in the Annals was more important than that of relations with barbarians, he nevertheless admitted that the Annals favored keeping China and the barbarians in separate inner and outer zones. In contrast to Ch'en Liang, Chu Hsi distinguished between restoring north China to Sung control and long-term revenge against the barbarians. National unity was thus a more compelling reason for addressing the barbarian menace for Chu Hsi than it was for Ch'en Liang and others, who preferred to focus on the menace.

Chu Hsi's digest, which was later further amplified, classified China's imperial regimes as either legitimate or illegitimate. Ssu-ma Kuang had explicitly refused to do so in an earlier work (on which Chu based his digest), which tells us that Chu Hsi was facing very different political pressures. Chu's classification, based on criteria of moral right and political unification, also drew on Ou-yang Hsiu's earlier intepretations of the Kung-yang Commentary. Ou-yang had elaborated on the theme of "magnifying universal rule" (ta i-t'ung ), which Tung Chung-shu and Ho Hsiu had both stressed in their New Text interpretations of the Annals. We find here an intriguing undercurrent of normative theory in Sung political thought. It affirmed the Former Han praise-and-blame tradition of the Spring and Autumn Annals and employed, in circumscribed form, doctrines once associated with New Text Confucianism. Classicism was here joined with a method for using history as a policymaking guide.^[18]

Issues regarding legitimate succession continued unabated during the Mongol Yuan dynasty, when protonationalism was replaced by Sung loyalism. Confucian scholars charged by the Yuan court to compile his-

[17] Davis, "Historiography as Politics," 33-39. See also Hok-lam Chan, "Chinese Official Historiography," pp. 68-71.

[18] See Ssu-ma Kuang's "Hsu" (Preface) to the Tzu-chih t'ung-chien , vol. 1, pp. 33-34, and Hok-lam Chan, p. 96. Cf. Davis, "Historiography as Politics," pp. 40-42, and Tillman, Utilitarian Confucianism , pp. 33-34, and Tillman, "Proto-Nationalism," p. 413. See also Hartwell, "Historical-Analogism," 690-95. Chu Hsi's digest of China's imperial regimes was based on chronicles completed in 1084 and entitled Comprehensive Mirror for Aid inGovernment. In the chronicles, Ssu-ma Kuang had used Confucius's Annals as a model to cover the period from the Eastern Chou (when the Annals left off) to the Five Dynasties.

tories of the earlier Liao, Chin, and Sung dynasties became embroiled in heated discussions over their legitimacy. Confucius's Annals had ordained the historiographical prerogative of the sage-king; hence all subsequent emperors compiled records of their predecessors to justify their own legitimacy.

Historiography was thus a politically charged vocation. To interpret the past (Sung, Liao, Chin) was to affirm the present (Yuan). Efforts by Confucians under Mongol rule to formulate acceptable principles for legitimate dynastic succession were further complicated by the fact that the Liao and Chin were both alien conquerors. Further, Yuan rulers, sponsors of the history projects, were not only foreign conquerors but had also destroyed a legitimate Han Chinese dynasty, the Southern Sung.

For Han Chinese, it was imperative to defend the legitimacy of the Sung, even in the face of Mongol pressure to grant the Liao and Chin dynasties legitimacy according to Confucian historiography. For its part, the Yuan government refused to accede to Chinese scholar-official demands that the Sung be accorded priority in legitimacy simply because it was a native Chinese dynasty. Debate was so intense that the writing of official histories was paralyzed, and the Yuan History Bureau became little more than a storehouse for documents.

Composed sixty years after Sung rule had been erased, Yang Wei-chen's "Polemic on Legitmate Succession" (Cheng-t'ung pien ) reveals that the political legitimacy of pre-Yuan dynasties was very much in dispute during the Yuan. A historian with an ongoing interest in the Spring and Autumn Annals, Yang Wei-chen was an ardent spokesman for traditional Confucian values and defender of the Confucian his-toriographical tradition. His dilemma (which Richard Davis describes as the "dilemma of a Chinese proto-nationalist") centered around his efforts to demonstrate that Northern Sung legitimacy had been transmitted to the Southern Sung and then to the Yuan—a scheme that entirely bypassed the Liao and Chin as legitimate dynasties. Yang connected philosophical orthodoxy to political legitimacy through the Neo-Confucian concept of Tao-t'ung (legitimate succession of the Way), which had earlier been employed by Chu Hsi and his Northern Sung predecessors to affirm the true transmission of Confucian values from the time of Confucius and Mencius to the Sung dynasty.

According to this construct, China's spiritual center, and thus its political values, had followed the Northern Sung court south when the north fell to the Jurchen in 1126. Although Yang Wei-chen accepted the

Yuan as a Chinese dynasty, this flagrant display of Han chauvinism led the Yuan court to ban his essay because of its disparaging remarks about the Khitan and Jurchen "barbarians." Yang's ethnic bias aroused court indignation and the suspicion that he was implicitly criticizing the Mongols as well.^[19]

When the Ming dynasty restored Han Chinese rule, the stamp of political legitimacy was withdrawn from the alien Yuan dynasty. This is seen in the work of the historian Wang Chu (fl. ca. 1521), who refuted the legitimacy of the Mongol rulers in his influential Sung-shih chih (Verified history of the Sung Dynasty). He accomplished this by fabricating a chronology based on Chu Yuan-chang's ancestors (Chu founded the Ming dynasty), thereby placing Chu's line in direct succession to the Sung house. Accounts of the Liao and Chin were relegated to a section entitled the "monographs on foreign nations." Ming Confucians like Wang Chu used principles enunciated in the Spring and Autumn Annals to reassert the supremacy of Chinese rule over all alien conquerors in China's past. "Sung loyalism" was transformed into Ming "Han chauvinism."^[20]

Ming-Ch'Ing Studies of the Annals

Some of the New Text doctrines from the Kung-yang and Ku-liang commentaries on the Annals, which were prepared during the Former Han dynasty, remained perennially important. For the most part, however, the Tso chuan was preeminent after the T'ang dynasty in establishing the historical precedents (li) that later Confucians had documented in the Annals. Consequently, the Old Text interpretations of the Annals were the standard for the examination system during the Ming and Ch'ing dynasties. Tu Yü's Explication of Precedents in the Spring and Autumn Annals was the most influential pre-T'ang reconstruction of this Old Text reading of the Annals because the author had harvested Old Text interpretations of the Classics from the Later Hah. During the T'ang dynasty, however, Tu's book was divided up and included with other commentaries in K'ung Ying-ta's definitive Ch'un-ch'iu cheng-i (Orthodox meanings in the Annals), which was part of an extensive imperial program for classical orthodoxy in the seventh cen-

[19] Hok-lam Chan, "Chinese Official Historiography," pp 71-88 and Davis, "Historiography as Politics," pp. 45-51.

[20] Hok-lam Chan, "Chinese Official Historiography," pp. 95-105.

tury. In fact, Tu Yü's book was lost as an independent work (as were many of Cheng Hsuan's writings), although his views were incorporated into the T'ang academy. Ku Yen-wu (1613-82) and Hui Tung (1697-1758) later played important roles in reconstituting Tu Yü's original work on the Annals.^[21]

Although the Old Text views articulated by Tu Yü remained orthodox for later dynasties, there were important exceptions. Neither Wang An-shih nor Chu Hsi—major figures in Northern and Southern Sung Confucianism respectively—had placed much store by the Spring and Autumn Annals, believing instead that the Rituals of Chou was a more authoritative text. In his old age, Chu Hsi openly stated that the Annals was not worth studying because it offered little guidance for moral improvement, asking: "What possible relevance does it have for US?"^[22]

Moreover, Ch'eng 1 and Chu Hsi had both expressed their doubts that the Tso Commentary had indeed been composed by Confucius's disciple Tso Ch'iu-ming. Ch'eng I nonetheless maintained, unlike Chu Hsi, that the Annals was an important world-ordering text whose historical precedents revealed "laws for a hundred ages" (see chapter 8). Confucius's Annals survived its Sung dynasty critics. During the Ming dynasty, doubts about its provenance were replaced by questions concerning its earlier interpretation. As in the T'ang, the three Han commentaries were again subject to critical scrutiny, particularly the derivation of historical precedents from the Annals enunciated in the Tso chuan.^[23]

Yang Shen

An eccentric Ming Confucian, Yang Shen pioneered k'ao-cheng studies when he was exiled to Yun-nan during the sixteenth century. In an essay entitled "Ch'un-ch'iu li" (Precedents in the Spring and Autumn Annals ), he challenged Tu Yü's Explication of Precedents in the Annals. Reversing the traditional Old Text perspective, Yang contended that the Annals was not simply a record of historical precedents. "In the writings of the sage," he asked, "are we to assume that he first had precedents in mind and then later wrote the Spring and Autumn Annals ?" The Annals was first and foremost a chronicle of historical

[21] Ssu-k'u ch'üan-shu tsung-mu , 26.10b-14a, 28.16b-17b, 29.5a-6b, 29.31a-34a.

[22] Chu Hsi, Chu Wen-kung wen-chi, hsu-chi , 2.6b.

[23] Erh-Ch 'eng ch'üan-shu, Ho-nan Ch'eng-shih i-shu , 15.16a-17a, and esp. 20.1a.

events, argued Yang, and only secondarily a statement of historical precedents.^[24]

Appealing to a more activist and creative interpretation, Yang Shen rejected the Old Text precedents as a "fixation on the past" (ni-ku ), which confused the letter with the spirit of antiquity: "Consequently, the sages established ritual and music as institutions. They [themselves] were not institutionalized by ritual and music. They established the law as an institution and were not institutionalized by the law."

According to Yang, Confucius had likewise created the Annals to record the chaotic events following the collapse of the Western Chou in 770 B.C. Confucius thereby established precedents to be mastered, but precedents were but part of the Annals. The Old Text position on the Annals—institutionalized through hundreds of historical precedents— had misrepresented Confucius's intent in composing his chronicle. Yang saw instead an appeal to the priority of voluntarism over passivity.^[25]

Yang Shen's classical studies stressed the importance of recovering the Confucian legacy of the Han dynasty, which in scope and tenor was remarkably resonant with Ch'ing dynasty Han Learning:

Someone asked Mr. Yang [Shen]: "With regard to the Classics, you choose much from Han Confucians, but do not choose from Sung Confucians. Why is that?" He answered: "When have I not selected the refined theories of Sung Confucians? One sees that Sung Confucians were mistaken in discarding Han Confucians, and using only their own views. Let me ask you: The Six Classics were completed by Confucius. The Han era was not separated from [the time of] Confucius by very long. Although the [Han] transmitters had weaknesses, their theories still contained the truth. Sung Confucians were separated from Confucius by 1500 years. Although their intelligence surpassed ordinary folks, how could they in a single morning completely discard old [Han views] and independently become enlightened through their minds?

Hui Tung could not have put it any better. In fact, Tai Chen (1724-77) later posed this exact argument. When studying the Great Learning at age ten with his teacher, Tai asked:

How does one know in this case that these are the words of Confucius recorded by Tseng-tzu [the reputed author of the Great Learning]? Moreover, how does one know that Tseng-tzu's intentions were recorded by his followers?

[24] Yang Shen, T'ai-shih sheng.an wen-chi , 43.12b. On Yang Shen see Lin Ch'ing-chang, Ming-tai k'ao-cheng-hsueh yen-chiu , pp. 36-127.

[25] Yang Shen, "Su-Ju ni-ku," 68.2a-2b; see also 45.15a-17a.

The teacher replied: "This is what the earlier Confucian Chu Hsi said in his notes."

Tai Chen asked another question: "When did Chu Hsi live?" The teacher replied: "Southern Sung" [1127-1279]. Tai asked again: "When did Confucius and Tseng-tzu live?" The teacher replied: "Eastern Chou" [770-221 B.C. ]. Tai asked again: "How much time separates the Chou [dynasty] from the Sung?" Reply: "About two thousand years." Tai questioned again: "Then how could Chu Hsi know that it was so?" The teacher could not reply.

This reconsideration of the "precedent tradition" concerning the Annals became part of the agenda to recover Han versions of the Classics. As we have seen with T'ang Shun-chih (1507-60) and Kuei Yu-kuang (1507-71), Hah dynasty classicism was becoming an important undercurrent in the Neo-Confucian mainstream.^[26]

Chi Pen

Chi Pen, a Che-chiang native and student of Wang Yang-ming (1472-1529), also reconsidered Tu Yü's "precedent tradition" in his influential study of the Annals. Entitled Ch'un-ch'iu ssu-k'ao (Personal study of the Spring and Autumn Annals ), Chi's study was honored with a long and detailed preface composed by T'ang Shun-chih in 1550. According to T'ang, the strength of Chi's analysis lay in its articulation of the true standards of the sages: "The Annals is a work exhibiting the right and wrong of the sages and is not a work that slanders or praises. What the true Way affirms as right, the Annals also affirms as right. What the true Way rejects as false, the Annals also rejects as false."

T'ang Shun-chih observed that earlier Confucians had lost touch with Confucius's concrete message of ameliorating chaos and punishing criminals—the hallmark of the Annals —and instead had turned to an esoteric discussion of historical precedents for "praise and blame." Chi Pen set out to correct such errors.^[27]

Chi Pen's own discussion of the Annals stressed the Mencian tradition that Confucius's chronicles were "the key to understanding the Way of rulership." Chi was critical of the Old Text tradition of interpretation based on the Tso chuan because the latter was written much later than either the Kung-yang or Ku-liang commentaries, which had been accepted in the official academy during the the early years of the

[26] Ibid., 42.2a-3b. The story about Tai Chen is recounted first in Wang Ch'ang, Ch'un-jung-t' ang chi , 55.6b.

[27] T'ang Shun-chih, "Hsu" (Preface) to Chi Pen, Ch'un-ch'iu ssu-k'ao .

Former Han. Chi also noted that the phraseology of the Tso chuan was suspiciously unlike other works predating the Warring States period (403-221 B.C. ). In fact, the text of the Tso chuan showed a remarkable resemblance to writings from the southern feudal state of Ch'u; he concluded that Confucius's disciple Tso Ch'iu-ming, who like Confucius was from the northern state of Lu, could not have produced it. In this way, Chi Pen undercut the Old Text link between the Tso chuan and the Annals.^[28]

To elucidate the "basic intent" (pen-i ) of the Annals, Chi Pen followed the T'ang interpretation tradition by emphasizing the priority of the Annals as a Classic over its three traditional commentaries. In so doing, Chi hoped to avoid the contentious debate that had pitted the Kung-yang and Ku-liang against the Tso chuan as the legitimate arbiter of the Annals' meaning. The distinction of right from wrong, according to Chi, should proceed directly from the Classic itself—thus avoiding futile efforts to master the hundreds of praise-and-blame precedents that had been read into the Annals on the basis of the commentaries.^[29]

In his 1557 afterword to Chi Pen's study, Wang Chiao clarified the reason for Chi's dissatisfaction with the Old Text "precedent tradition." He explained that the effort to read elaborate historical precedents (fan-li ) into the Annals on the basis of Confucius's encoding "praise and blame" into the chronicles of Lu had produced a tradition classifying the Annals as "Confucius's book of punishments." Rather than stressing the centrality of ritual precedents, as in the other Classics, Confucians had misappropriated the Annals as a legal handbook of historical precedents (see chapter 8). Consequently, the Annals' stress on the ritual aspects of social order had been overlooked and the true mind-set of the sages (sheng-jen chih hsin ) overturned.^[30]

In his own writings on the Spring and Autumn Annals, T'ang Shun-chih also maintained that the tedious elucidation of precedents in the Annals detracted from Confucius's central concern, which was ending the chaos of his age. Mencius had been right, T'ang thought: "The Annals recorded the affairs of the Son of Heaven." In line with his own more practical concerns, he called the Annals a "book of statecraft, whose statecraft [content] revealed how to rectify chaos and crime." It

[28] Chi Pen, "Hsu" (Preface) to ibid., pp. 4a-7a. See also Meng-tzu yin-te , 25/3B/9. Chi Pen pointed out that Chu Hsi also noted that the Tso chuan was a work of history linked to the state of Ch'u.

[29] Chi Pen, "Hsu" (Preface), Ch'un-ch'iu ssu-k'ao , pp. 7a-b.

[30] Wang Chiao, "Hou-hsu" (Afterword), to Chi Pen, Ch'un-ch'iu ssu-k'ao .

contained concrete lessons for a time of chaos and accorded with the requirements of change (pien )^[31]

Hao Ching

Many of the issues explored by Chi Pen were continued by the Hu-kuang literatus Hao Ching. As we note earlier, Hao was an opponent of official corruption in the late sixteenth century and had become involved with the Tung-lin partisans while serving in Ch'ang-chou Prefecture as magistrate in Chiang-yin County. Hao Ching's classical scholarship is most noteworthy for its assault on the authenticity of the Old Text portions of the Documents. He was even critical of Chu Hsi for his superficial analysis of the problem, distraught that Chu Hsi, despite his doubts about their authenticity, could have instructed his student Ts'ai Shen to annotate the Old Text chapters of the Documents Classic. Hao was also angry that Chu Hsi had told his students that much of the New Text version was indecipherable, viewing this an affront to the authentic words of the sages. Through an analysis of the stylistic inconsistencies in the Old Text chapters, Hao Ching hammered yet another nail in the coffin of the Old Text Documents, which Yen Jo-chü acknowledged in his own analysis of the issues five decades later.^[32]

Questions concerning the authenticity of the Tso chuan as an orthodox commentary on the Spring and Autumn Annals reached an interesting climax in the late Ming. Hao Ching turned his attention to this issue in a work entitled Ch'un-ch'iu fei Tso (The Annals is not linked to the Tso chuan ), also challenging the Old Text position on the Annals. Only the exposes of Liu Feng-lu and K'ang Yu-wei in the nineteenth-century would surpass Hao's acerbic assault on one of the pillars of the Old Text classical tradition.

In his preface to Hao Ching's study of the Tso chuan, An Chieh-ch'uan noted that Hao's account represented the (then) final step in a long line of scholars dating back to the T'ang who suspected the claim that Tso Ch'iu-ming had compiled a commentary on the Annals. In his own 1610 preface, Hao Ching took the bull by the horns. He explained his intent was to demonstrate that Tso Ch'iu-ming had nothing to do not only with the Tso chuan but also with the Annals. Contending that

[31] T'ang Shun-chih, Ching-ch'uan hsien-sheng wen-chi (1573), 7.2a-3a, 17.4a-6a, 17.12b.

[32] Yen Jo-chü, Shang-shu ku-wen shu-cheng , 8.14a-18b. See also Hao Ching, "Tu-shu," p. 3b.

the Tso chuan had achieved such exalted status that no one dared criticize its links to Confucius's chronicles of Lu, Hao Ching explained it was his duty to defy convention in order to right the classical record.^[33]

Hao traced mistaken views on the Tso chuan back to Ssu-ma Ch'ien (145-86? B.C. ), who as grand historian in the Former Han had been the first to identify Tso Ch'iu-ming as the author of the Tso chuan. Pan Ku (32-92), author of the History of the Former Han Dynasty, and Tu Yü had carried on this tradition, and even accepted Liu Hsin's earlier claim during the Wang Mang interregnum that Tso Ch'iu-ming personally knew Confucius and had thus recorded his true intentions for the Annals. Hao saw them as the culprits who had "slandered the Annals and deluded scholars" to an unprecedented degree.

Hao wondered why, if Tso Ch'iu-ming wrote the Tso chuan as a result of his personal relation to Confucius, he wasn't even mentioned among the seventy disciples? Surely a disciple who had the inside story on the Annals —Confucius's own historical record—would have been worthy of mention. That Tso Ch'iu-ming had no relation to Confucius or the Annals, according to Hao Ching, was confirmed by the fact that Confucius had terminated the Annals with the capture of the marvelous lin (a one-horned doe-like animal, perhaps similar to a unicorn) in 481 B.C. (see chapter 7), whereas the Tso chuan went on for another thirteen years, not ending its account of the Spring and Autumn era until 468 B.C. If the Tso chuan were actually a commentary on the Annals, why this discrepancy?

In his analysis, Hao Ching said he accepted the position of New Text Confucians. During the last years of the Former Han they had opposed Liu Hsin's efforts to advance the Old Text Classics, particularly the Tso chuan, into the Imperial Academy. Moreover, Hao explained that his efforts were not new. Appealing to the authority of illustrious T'ang Confucians such as Han Yü, Hao concluded: "Therefore, the fact that the Tso [Commentary ] is not reliable was already realized by former people. It is not just due to me." Hao was seeking to reestablish the New Text version of Confucius as an "uncrowned king" (su-wang ), which had been "unclear" (pu-ming ) for several dynasties because of the influence of the Tso chuan in Ch'un-ch'iu-hsueh.^[34]

[33] Hao Ching, "Hsu" (Preface) to his Ch'un-ch'iu fei-Tso . See also An Chieh-ch'uan, "Hsu" (Preface) to ibid.

[34] Hao Ching, Ch'un-ch'iu fei-Tso , 2.29b-30b, 2.32b-33a. See also Hao, "Tu ch'un-ch'iu," pp. 2b-4b.

It would be interesting to speculate about Hao Ching's radical attack on Old Text classicism. How did the Documents and Tso Commentary relate to his checkered official career and largely unsuccessful reform efforts? Hao was not a New Text scholar, but one senses in his efforts to gainsay the orthodox Old Text position and in his anti-Sung Learning stance an inchoate appeal to a Confucianism unsullied by the values of his day. Again, classicism was not an escape from political reality but a search for the way back to ancient ideals betrayed in the present.

Yao Chi-Heng

Questions about the Old Text tradition of "precedents" in the Spring and Autumn Annals carried over into the Ch'ing dynasty, when the turn to Han Learning became more pronounced. Yao Chi-heng, the distinguished evidential research scholar and antiquarian, who in addition to his eradication of spurius elements from the Confucian Canon (including the Old Text Documents ) wrote the Ch'un-ch'iu t'ung-lun (Penetrating discussions of the Annals ). This work called for an impartial (that is, k'ao-cheng ) reevaluation of the "precedent tradition" associated with the Annals. Citing both Chi Pen and Hao Ching as his immediate predecessors on the matter, Yao Chi-heng reopened the problem of the provenance of the Tso chuan as a commentary that legitimated the historical precedents encoded in the Annals .^[35] Yao reaffirmed the priority of the Annals over its commentaries and changed the epistemological context in which the "right and wrong" taught by the Annals would be corroborated:

Everyone has a mind that discerns right from wrong. If something is right, then the Classic makes it right. If something is wrong, then the Classic makes it wrong. There have never been two branches with regard to the scope of right and wrong. One should examine a work for its reasonableness. If the human mind is of itself the same, then there is no need to speak [of the commentary]….This is why scholars should appropriately discard the commentary and follow the Classic. One cannot discard the Classic and follow the commentary.

Analysis of the Classic should be corroborated by the reasonableness of claims made for right and wrong. The commentaries, Yao maintained, had strayed from the practical concerns of right and wrong that in-

[35] Yao Chi-heng, "Hsu" (Preface) to his Ch'un-ch'iu t'ung-lun. See also Yao "Ch'un-ch'iu lun-chih," in ibid., A.6b.

formed the Annals and become lost in esoteric discussions of tedious points based on its phraseology.^[36]

Chief among the faults in Ch'un-ch'iu-hsueh , according to Yao Chiheng, was its long-established "precedent tradition," which was purported to hold the key to the meaning of the Annals:

The single graph for "precedent" [li ] did not exist in antiquity. Later some vulgar [that is, noncanonical] graphs were converted to this graph [of li ] to discuss the Annals . From the beginning this was a mistake. Moreover, how could one say that Confucius had used this single graph to compose the Annals ? If one had forced Confucius to compose the Annals on the basis of a single graph [of li ], then Confucius would not have been as ignorant as this. Accordingly, to search out the precedents he purportedly created does not agree very much with Confucius.

T'ang and Sung Confucians had erred by accepting a bogus "precedent tradition." By turning the Annals into a work that recorded "extraordinary affairs" (fei-ch'ang chih shih ), Confucians had converted it into a casebook for ritual and legal transgressions (fei-li fei-fa ), which in the end misrepresented Confucius as a Legalist (fa-chia ) and the Annals as a book of punishments (hsing-shu ) (see chapter 8).^[37]

The culpit for this misguided line of interpretation, according to Yao, was Tu Yü, whose Explication of Precedents was the first work to invoke the Tso chuan as the repository of historical precedents in the Annals. Moreover, Tu had prepared a commentary on the Chin dynasty legal code of A.D. 268, which had overlapped with his work on the Annals. At about the time Yao Chi-heng was writing, Ku Yen-wu and others were rescuing Tu Yü's work from the T'ang dynasty scissors of K'ung Ying-ta (see above). For Yao Chi-heng, unlike Ku Yen-wu, the Old Text tradition associated with the Tso chuan had clouded the Annals' true meaning: "Alas, since the single graph of li [precedent] appeared, the meaning of the Annals from the beginning has not been illuminated in the empire. The sage, basing himself on the history of Lu, composed the Annals. How could he have first affirmed its precedents?"

Yao demonstrated Tu Yü's errors by appending a detailed case-by-case rejection of the precedent tradition entitled "Ch'un-ch'iu wu-li hsiang-shuo" (Detailed discussion on the lack of precedents in the Annals ) to his study of the Annals. In addition, Yao claimed that the effort to systematize the events in the Annals into precedents was untrue

[36] Yao Chi-heng, "Hsu," p. 2a.

[37] Ibid., pp. 2b-3a.

to history (shih ): "History is used to record events. Events have ten thousand changes. How can precedents even things out? If it matches with this event, then it betrays that one; if it is the same as this event, then it differs from that one." History was too complicated to be reduced to a single system of precedents. Tu Yü had recognized this problem and created a special category called "changing precedents" (pien-li ) in an effort to address it—although pien-li is a contradiction in terms, if a precedent really represents an unchanging historical lesson.^[38]

According to Yao Chi-heng, the precedent tradition provided the theoretical basis for the praise-and-blame (pao-pien ) tradition that had grown up around the Annals. Not only did Confucius have no intent to praise and blame (wu pao-pien chih hsin ), but also the Annals itself included no formalized and consistent pattern for praise or blame. The Old Text precedent-based interpretation placed all the various items in the Annals into a straitjacketed account of good and evil that had forced history into an unchanging pattern of events.^[39]

Yao exhibited a view of history that would later be championed by eighteenth-century k'ao-cheng historians such as Ch'ien Ta-hsin and Wang Ming-sheng. Like them, Yao appealed to the integrity of history as the complicated record of human events, not a simplistic acting out of precedents in a praise-and-blame drama: "History is used to record events. Consequently, normal and abnormal events are both recorded. If one selects only abnormal events to record, and does not record normal events, how can it be history?" In essence, Yao Chi-heng was calling for a demystification of the Annals, hoping to restore it to its legitimate position as a historical chronicle of the state of Lu during the decline of the Eastern Chou dynasty.^[40]

To correct the prevailing precedent tradition, which had portrayed Confucius as a Legalist, Yao proposed that the Annals be evaluated in light of its "essential points" (yao-chih ). Yao's use of "point" (chih ) to replace "precedents" (li ) represented an effort to recast the interpretation of the Annals. It may also represent the impact of the New Text tradition of Ch'un-ch'iu-hsueh, which stressed the "three classifications and nine points" (san-k'e chiu-chih ) to clarify the Annals (see chapter 7).

But Yao Chi-heng discarded both the praise-and-blame tradition

[38] Yao Chi-heng, Ch'un-ch'iu t'ung-lun , A.la-b and "Fu" (Appendix). See also Ssu-k'u cb'üan-shu tsung-mu , 26.Sa-6b.

[39] Yao Chi-heng, Ch'un-ch'iu t'ung-lun , A.2a-3a.

[40] Ibid., A.3a-b. On k'ao-cheng historiography see my Philosophy to Philology , pp. 70-76.

associated with the New Text commentaries (Kung-yang and Ku-liang ) and the precedent tradition associated with the Old Text Tso chuan:

Everyone knows that the Kung-yang and Ku-liang [commentaries] are perversions. For the moment there is no need to mention them. Few people, however, are aware of the impractical errors and perversions of Master Tso's [commentary], and thus they remain deluded. Why is this? The Kung-yang and Ku-liang [commentaries] both discuss meanings [i ]; hence their perversions can be brought to view. Master Tso's [commentary] discusses events [shih ]; hence its mistakes and perversions are difficult to encompass.

The locus for the interpretation of the Annals was in flux. It is interesting that Chuang Ts'un-yü would later entitle one of his essays Ch'un-ch'iu yao-chih (Essential points in the Annals ), suggesting the influence of Yao's attack on the Old Text precedent tradition.^[41]

Yao singled out the Tso chuan for criticism because it was the basis for the dominant precedent tradition. In partial agreement with Hao Ching's stinging rejection of the Tso chuan, Yao Chi-heng noted the many factual inconsistencies between the commentary and the Classic. Yao also rejected the tradition that Tso Ch'iu-ming, Confucius's reputed disciple, had composed the commentary. Although he did not go as far as Hao Ching in linking the Tso chuan to Liu Hsin, Yao still undercut one of the essential texts in the Old Text classical canon. Liu Feng-lu would later weave the various strands in Hao Ching's and Yao Chi-heng's criticism of the Tso chuan into a coherent attack on the Old Text Classics as a whole.^[42]

Han Learning and the Annals

Efforts to properly interpret the Annals became caught up in Han Learning currents through Hui Shih-ch'i's research on Confucius's chronicles of Lu, published in 1749 by his celebrated son Hui Tung. As we note in chapter 1, Hui family traditions were decisive in determining the prominence of Hah Learning in eighteenth-century Su-chou. Hui Shih-ch'i's attempt to restore the Annals to its Han dynasty form in many ways served as the model for Hui Tung's later efforts to reconstruct "Han Learning on the Change Classic."^[43]

Because the Han Learning movement in Su-chou affirmed Later Han Confucians such as Cheng Hsuan, Hsu Shen (58-147), Ma Jung (76-

[41] Yao Chi-heng, Ch'un-ch'iu t'ung-lun , A.4a-5b, 6a-6b.

[42] Ibid., A.6b-7b.

[43] See Yang Ch'ao-tseng, "Pei" (Epitaph) for Hui Shih-ch'i.

166), as well as Tu Yü (men who were directly and indirectly tied to the triumph of Old Text over New Text classicism), Hui Tung and his followers viewed earlier efforts to expunge the Tso chuan from the classical record with disdain. In fact, the editors of the Ssu-k'u ch'üan-shu project in the 1780s—most of whom were mainstream Han Learning advocates—were extremely critical of Chi Pen and Hao Ching for their temerity to question the Tso chuan. The editors went so far as to argue that the Ming dynasty represented the lowest point in the long history of Ch'un-ch'iu-hsueh. At least Chi Pen's and Hao Ching's works on the Annals were considered important enough to be included in the Imperial Library. Yao Chi-heng's work on the matter was not even mentioned in the catalog of works on the Annals. From the outset of their account of the development of Ch'un-ch'iu studies, the Ssu-k'u ch'üan-shu editors made their allegiance clear: (1) Tso Ch'iu-ming had indeed compiled the Tso chuan, and (2) the Tso chuan was not only a legitimate commentary on the Annals but was also far superior to both the Kung-yang and Ku-liang alternatives. Hah Learning had pitched its tent in Old Text classicism.^[44]

A champion of Later Hah sources, Hui Shih-ch'i was critical of Tang and Sung Confucians for ignoring Hah dynasty commentaries in their efforts to elucidate the Annals. Hui singled out Tan Chu and Chao K'uang in particular for their pernicious influence on Annals scholarship. What troubled him was the inability of post-T'ang Confucians to see that the Annals and Rituals of Chou were complementary Classics that together completed the classical paradigms for political and social order. Chao K'uang, for example, argued that the Chou-li was a later forgery; Wang An-shih asserted the Annals was worthless. In many ways, Shih-ch'i sought to reconstruct the Later Hah version of the Annals while reaffirming the Old Text emphasis on the Rituals of Chou .^[45]

Unlike Yao Chi-heng, Hui Shih-ch'i continued to accent the precedent tradition associated with the Annals. He continued to regard Tu Yü's Explication of Precedents as authoritative because it was a post-Hah dynasty work chronologically close to the Han "schools system" of classical transmission. In addition, Hui's commitment to Later Hah classicism led him to reaffirm the priority of the Tso chuan over the Kung-yang and Ku-liang commentaries for understanding the historical precedents in the Annals.^[46]

[44] Ssu-k'u ch'üan-shu tsung-mu , 30.11a-b, 17a-b.

[45] Hui Shih-ch'i, Pan-nung hsien-sheng ch'un-ch'iu shuo , 7.18a-19b.

[46] Ibid., 5.10a, 1.22a-22b.

According to Hui Shih-ch'i, the Rituals of Chou was a work of the Western Chou dynasty, while the Annals exemplified Eastern Chou rituals and institutions. Together they formed the basis, Hui thought, for understanding the Chou ritual traditions that Hah Confucians had drawn on: "The Duke of Chou accordingly created the six records [that is, the Rituals of Chou ], and Confucius composed the Annals. Both were means [to understand] completely the nature of man and things, and how to oppose a chaotic age and restore it to order."

Sung Confucians had questioned the authenticity of the Chou-li but, aside from Wang An-shih, had never wavered on the provenance of the Annals. In Hui Shih-ch'i's Hah Learning study of the Annals, the latter was secondary to the Rituals of Chou in importance. The Chou-li was drawn from early antiquity (that is, the Western Chou), while the Annals came much later (that is, the Eastern Chou). The roles of the Annals and Rituals in classical discourse were thus more properly understood as complementary rather than antagonistic. Consequently, it is misleading to assume, as earlier studies have done, that the New Text school regarded Confucius alone as its model and that only the Duke of Chou was the ideal of the Old Text school.^[47]

Ch'Un-Ch'Iu-Hsueh in Ch'Ang-Chou

In the middle of the eighteenth century, research on the Spring and Autumn Annals showed little unanimity on the problems of the precedent tradition and the relationship between the Tso chuan and the Annals. But by the time Chuang Ts'un-yü formulated his ideas concerning the Kung-yang Commentary, an influential consensus had been reached. Hah Learning scholars of the Su-chou school became wedded to the Old Text traditions of the Later Han and accepted without question the precedent tradition associated with the Annals since Tu Yü. Moreover, they stressed the legitimacy of the Chou-li as an orthodox Classic. This consensus penetrated the highest levels of Confucian scholarly institutions and became the official view of the Imperial Library on Ch'un-ch'iu-hsueh. Both Hui Shih-ch'i and Hui Tung were praised in no uncertain terms by the editors of the Ssu-k'u ch'üan-shu project. Shih-ch'i was lauded for his use of the Tso chuan as a source; Hui Tung's efforts to reconstruct Tu Yü's annotation of the Tso chuan was likewise commended.^[48]

[47] Ibid., 8.18b-19a, 9.26a-27b. See also Karlgren, "Chou-li and Tso chuan Texts."

[48] Ssu-k'u ch'üan-shu tsung-mu , 26.27b-28b, 26.31a-34a.

This Han Learning consensus was also influential in local scholarship. In Ch'ang-chou both Ku Tung-kao and Yang Ch'un reaffirmed the Old Text position. In a prodigious work of scholarship entitled Ch'un-ch'iu ta-shih piao (Table of major events in the Spring and Autumn Annals ), the Wu-hsi scholar Ku Tung-kao brought together chronological, geographical, genealogical, and economic information, using the Annals as a historical source on the Eastern Chou and adding all relevant material he could find in related sources—in effect a tour de force in k'ao-cheng historical studies.^[49]

In his preface to Ku Tung-kao's study, Yang Ch'un traced the evolution of the precedent tradition from the Kung-yang and Ku-liang commentaries to the Tso chuan. He claimed that the Tso chuan provided the most precise understanding of precedents. Tu Yü's Explication of Precedents, according to Yang Ch'un, represented a distillation of views that could be traced back to both the New and Old Text schools. For Yang, the strength of Ku Tung-kao's account was its success in drawing on the virtues of each of the three Han commentaries and weeding out doctrines and facts that betrayed Confucius's intent. For example, Ku's account emphasized both precedents (li ) and meanings (i), suggesting to Yang Ch'un an effort to synthesize the Former Han and Later Han commentaries.^[50]

In his own discussion, however, Ku Tung-kao maintained that although he used the Former Han commentaries, he had gainsaid their message. On the question of why Confucius made the capture of the mythical lin the final entry of the Annals, Ku rejected Ho Hsiu's messianic New Text theories as "lies and deceptions" (tan-wang ). Ho Hsiu's fanciful interpretation of the Kung-yang chuan made Ku conclude that Confucians had misunderstood Confucius's true intent in recording the capture of the mythical lin (unicorn).

According to Ku Tung-kao, the sage had ended his account in 481 B.C. because in that year a vassal had not been punished for assassinating his lord. By this time, Confucius was seventy-one, and he saw no hope for ending the chaos of the age. The capture of the unicorn came at precisely this time, confirming the irrevocable decline of the Eastern Chou. Therefore, Confucius ended his chronicles. The capture of the lin had historical significance but, according to Ku, no transhistorical import.^[51]

[49] Ibid., 26.28b-30a. For a discussion see my Philosophy to Philology ,p. 188.

[50] Yang Ch’un, “Hsu" (Preface) to Ku Tung-kao, Ch’un-ch’iu ta-shih-piao .For Yang's own views on the Annals see his Meng-lin-t'ang chi , 5.13b-17b.

[51] Ku Tung-kao, Ch'un-ch'iu ta-shih-piao , 42.18b-21b.

Although Yang Ch'un vigorously attacked the Old Text Rituals of Chou as a Legalist text in Confucian garb, he still upheld the Old Text position on the Annals. As yet, there was no inconsistency in such a stand. The Old Text-New Text debate had not yet been fully reconstructed, and the classical consensus seemed impervious to the full implications of growing interest in New Text sources from the Former Hah dynasty. Like Yao Chi-heng, however, Yang Ch'un rejected efforts to turn the Annals into a simple list of historical precedents. The moral judgments encoded in the chronicles by Confucius did not have independent status: Confucius "drew on the history of Lu for meanings; he did not create the meanings all by himself."^[52]

Links between Ku Tung-kao and Yang Ch'un indicate that Han Learning and its Old Text biases had powerful protagonists in Ch'ang-chou Prefecture when Chuang Ts'un-yü began to turn to New Text studies. But Chuang's new interest was rooted in the kao-cheng movement. Important scholars such as Yao Chi-heng—less wedded to Hah Learning and more committed to impartial research methods—initiated questions about the Old Text precedents tradition by calling the Tso chuan into doubt as a legitimate source for interpreting the Annals' meaning (i ). Even the editors of the Ssu-k'u ch'üan-shu admitted the chief weakness of Hui Tung's Hah Learning agenda was its unquestioning adoration of antiquity (ni-ku )^[53]

In fact, Chuang Ts'un-yü's New Text interests had been anticipated in Ch'ang-chou by Yang Fang-ta. Yang's Ch'un-ch'iu i pu-chu (Appended notes to the Meanings in the Annals ) furthered the work of his mentor, the distinguished Hanlin academician Sun Chia-kan, who had contended that "the Change, Poetry, and Annals constituted the complete Classics of the sages." For his part, Yang Fang-ta turned his attention to the long-neglected Kung-yang and Ku-liang commentaries. In addition, he began to rely for some of his interpretations of the Annals on Tung Chung-shu's Ch'un-ch'iu fan-lu (Spring and Autumn Annals ' radiant dew), the definitive Former Han New Text interpretation of Confucius's chronicles.^[54]

Chuang Ts’un-yü, followed by his grandson Liu Feng-lu, would further clarify and enlarge on Sun’s and Yang’s writings. Han Learning

[52] Yang Ch'un, "Ch'un-ch'iu k'ao hou-hsu" (Afterword to a study of the Annals), 5.15b-17b.

[53] Ssu-k'u ch'üan-shu tsung-mu , 29.34a. Cf. my Philosophy to Philology , pp. 59-60.

[54] Ssu-k'u ch'üan-shu tsung-mu , 31.36a-38a. The editors describe Yang as "secretly honoring Han Confucians" because he did not cite his borrowings from Tung Chung-shu. See also Lu Wen-ch'ao, Pao-ching-t'ang wen-chi , pp. 361-66 (chüan 27).

scholars, in their attempts to recover and reconstruct the original teachings of the sages, were employing the Old Text exegetical tradition of the Later Han. Continuing this scholarly enterprise, Ch'ang-chou scholars associated with the Chuangs and Lius in the late eighteenth and early nineteenth centuries began to push back the frontiers of classical learning and to focus on the Former Han dynasty as a better source for the wisdom of the sage-kings.

Chuang Ts'Un-Yü and the Annals

In a series of works on the Rituals of Chou initiated in 1783, Chuang Ts'un-yü sought to restore that text to its original form by adding quotations and materials from other works that cited the Rituals of Chou. The version with which Chuang was working was severely limited as a repository of references to classical institutions. Although he made no claim that the Rituals was a forgery, Ts'un-yü was critical of the surviving version. In this he was corroborating suspicions dating from the Northern Sung, and most recently reiterated in Ch'ang-chou by Yang Ch'un.^[55]

In keeping with his ecumenical position on the Classics, Chuang Ts'un-yü affirmed as much as possible of the classical legacy that under` girded official ideology. As with the Old Text chapters of the Documents Classic, Chuang saw no benefit to be had from purist efforts to rid the Confucian Canon of suspicious Classics such as the Chou-li. Instead, he feared the k'ao-cheng assault on Confucian orthodoxy might spill over into the political arena. His position on the Annals was syncretic, and he incorporated both New and Old Text Classics, believing them to be the core of a vision of antiquity bequeathed by the sages to the present. To verify certain points in the Annals, Chuang used both the Ku-liang and Tso commentaries. If he affirmed the Kung-yang over the Tso commentary, for example, it was not because he attempted to restore the New Text Classics as a whole to their Former Hah position of preeminence.^[56]

Kung-Yang Confucianism

Chuang's work on the Spring and Autumn Annals did, however, break new ground. Although he focused on the historical "precedents" (li )

[55] Chuang Ts'un-yü, Chou-kuan chi , 161.1a. See also Chuang Shou-chia, "Chou-kuan chi pa," pp. 36a-37a.

[56] Li Hsin-lin, "Ch'ing-tai ching chin-wen-hsueh shu," 174.

and “meanings" (i ) that Confucius had encoded in his chronicles of Lu from 722 to 468 B.C., Chuang followed the New Text alternative and rejected the Old Text position. According to Ts’un-yü, the Tso chuan was primarily a historical tecord. It stressed historical affairs (shih ) in the Annals but went no further than elaborating the facts. To overcome the limitations of this Old Text commentary, Ts’un-yü turned to the Kung-yan chuan , which contained historical interpretations that, in Ts’un-yü’s mind, revealed the true design for Confucius’s chronicle of events.^[57]

Ho Hsiu’s Later Han explication of the Kung-yan chuan was the chief source for Chuang Ts’un-yü. Long ridiculed for his “preposterous theories" (we-li chih shuo ), Ho Hsiu had been Cheng Hsuan’s archrival for the articulation of the Classics during the last century of the Later Han. Unlike his more famous rival, Ho Hsiu had been a staunch defender of the Kung-yan Commentary and New Text interpretations of the Classics. Besides composing an influential commentary to the Kung-yan text, HO also penned a vitriolic trilogy attacking Cheng Hsuan’s position on the Annals. Cheng Hsuan, an advocate of synthesizing Old Text and New Text views whenever possible, defended his position on the Annals in his equally stinging replies. The bone of contention in the Later Han New Text vs. Old Text controversy was the Kung-yan chuan.^[58]

By stressing Ho Hsiu and the Kung-yan chuan , Chuang Ts’un-yü was unequivocally controverting his Han Learning contemporaries who had installed Cheng Hsuan as the patron saint of Han Learning. Moreover, the influential compilers of the Ssu-k’u ch’üan-shu had, as we have seen, established the Old Text position associated with the Tso chuan as the authentic representative of the Han Learning version of the Annals. Initially championed in Su-chou by Hui Tung in the 1752s, the Han Learning wave crested in the 1780s with the imperially sponsored enshrinement in Peking of Later Han Confucianism in the Imperical Library.^[59]

To be sure, Chuang Ts’un-yü was still appealing to the authority of Han Learning with his emphasis on the Kung-yan chuan. But his form of Han Learning was in many ways inimical to Old Text versions domi-

[57] Ssu-k'u ch'üan-shu tsung-mu , 26.4b-6b.

[58] Ho Hsiu's trilogy is entitled Tso-shih kao.huang (Incurability of Master Tso), Ku-liang fei-chi (Disabling diseases of Ku-liang), and Kung-yang mo-sbou (Stalwart defense of Kung-yang). For a discussion see Dull "Apocryphal Texts of the Hah Dynasty," pp. 388-400.

[59] Ssu-k'u ch'üan-shu tsung-mu , 26.1a-4b.

nant in Su-chou and Yang-chou. Because it was a Han dynasty source, Ho Hsiu's Kung-yang chieh-ku (Explication of the Kung-yang Commentary ) could not be summarily dismissed by those who championed Han Learning.

If we cannot call Chuang Ts'un-yü a New Text scholar, then we can at least refer to his position on the Annals as the turning point in the revival of the Kung-yang Commentary as the key to Han Learning. Chuang's overall classical vision (which encompassed the Change, Chou-li, and Annals ) will therefore be described as " Kung-yang Confucianism." Later advocates of "New Text Confucianism" such as Wei Yuan (1794-1856) maintained that Chuang Ts'un-yü was the "true Han Learning scholar" (chen Han-hsueh che ), in contrast to his "false" contemporaries.^[60]

Historical "Precedents" vs. "Guiding Points"

Chuang Ts'un-yü's principal work on the Annals was entitled Ch'un-ch'iu cheng-tz'u (Correcting terms in the Annals ), to which were appended two shorter works: Ch'un-ch'iu chü-li (Examples of precedents in the Annals ) and Ch'un-ch'iu yao-chih (Essential points in the Annals ). As the titles suggest, Chuang affirmed the legitimacy of the "precedent tradition" while stressing the phraseology of the Annals as the key to its interpretation.

Although they acknowledged the praise-and-blame legacy of the Annals , Chuang's works also revealed the influence of Ming-Ch'ing Confucians who had questioned Tu Yü's reading of the Tso chuan, which we describe above. As a result, Chuang turned away from Tu Yü's detailed list of precedents in the Annals and emphasized a simpler and more general list drawn from the Kung-yang Commentary. In addition, Ts'un-yü discussed "guiding points" (chih, lit., "directives") in the Annals . Yao Chi-heng had earlier proposed such "directives" as an appropriate alternative to "precedents" in elucidating the significance of the Annals .^[61]

During the Former Han dynasty Tung Chung-shu enunciated "Ten Guiding Points" (shih-chih ), upon which Yao Chi-heng and Chuang Ts'un-yü both drew. Tung wrote:

[60] Wei Yuan, Wei Yuan chi , p. 238, includes Wei's "Hsu" (Preface) to the collected works of Chuang Ts'un-yü.

[61] All these works are included in Chuang Ts'un-yü's Wei-ching-chai i-shu . The character that Chuang used for chih includes the "hand" (shou ) radical, while the character for chih in Yao's "points" does not. The meanings are homologous, however.

The Annals is a text covering 242 years, in which the great outline of the world and the broad changes of human events are all fully included. In summary, however, it may be reduced to Ten Guiding Points, by which [all] the events [it narrates] may be linked, and from which [all] the transforming influences [in the rule] of kings may be derived. (1) To describe the changes in human events and show what is important in them is one guiding rule. (2) To show what these changes lead to is another. (3) To utilize what leads and controls them is another. (4) To strengthen the trunk, weaken the branches, stress what is primary, and minimize what is secondary, is another. (5) To discriminate among uncertainties and differentiate what [seemingly] belong to similar categories is another. (6) To discuss the appropriate [use] of the good and talented, and differentiate them according to the abilities in which they are preeminent, is another, (7) [To show how a ruler should] cherish persons who are close to him, induce those who are distant to come near, and identify himself with the desires of his people, is another. (8) [To show how], having inherited the refinement of Chou, he is to revert to simplicity, is another. (9) [To make clear that] Heaven's starting point lies in the fact that the wood [phase] produces the fire [phase], which constitutes summer, is another. (10) And to analyze how those whom it criticizes are punished, and examine how prodigies are accordingly applied in [compliance with] Heaven's principle, is another.^[62]

By stressing the "key points" of the Annals, Chuang was emptying the Annals of Old Text historical "precedents" and replacing them with New Text "meanings."

The immediate thread of Chuang Ts'un-yü's position went back further than Ming dynasty questions regarding Tu Yü's position on precedents, however. Chuang appealed to the authority of pre-Ming scholarship in an effort to turn Ch'un-ch'iu-hsueh in a direction he thought would get at the heart of what Confucius had encoded in the chronicles of Lu.

Chao Fang

Chuang Ts'un-yü's Correcting Terms in the Annals drew heavily on the work of Chao Fang, the Yuan dynasty scholar. Chao's Ch'un-ch'iu shu-tz'u (Comparative phraseology of the Annals ) had tried to delineate Confucius's systematic linguistic framework for compiling the Annals. According to Chao Fang, the key to unlocking the secrets of the Annals lay in grasping the interrelation of Confucius's use of the words and his

[62] See Su Yü's (d. 1914) edition of Tung's work entitled Ch'un-ch'iu fan.lu i-cheng , 5.9a-9b, for Tung Chung-shu's account of the "Ten Guiding Points." Chuang and Tung use the same character for "points." Cf. Fung, History of Chinese Philosophy , vol. 2, p. 76.

comparative evaluation of historical events (shu-tz'u pi-shib ), which during the Han dynasty had enabled Confucians to use the Annals as a legal casebook (see chapter 8).

In his preface to Chao's study, the distinguished Confucian scholar Sung Lien (1310-81) noted that Chao Fang had reopened the path to understanding the "law-models of the Annals " (Ch'un-ch'iu chih fa ). In addition, by applying historical methods (shih-fa ) to the text of the Classic (ching-wen ), Chao Fang had recaptured the links between language and moral vision Confucius had used in his chronicles.^[63]

Chao Fang was a follower of the Sung-Yuan Neo-Confucian Wu Ch'eng (1247-1331), who had attacked the authenticity of the Old Text chapters of the Documents Classic. Chao also studied the Annals under Huang Tse, who stressed the Old Text tradition of precedents based on the Tso chuan as interpreted by Tu Yü. In his Ch'un-ch'iu Tso-shih chuan pu-chu (Additional annotations to Master Tso's commentary for the Annals ), Chao Feng pioneered efforts to restore Tu Yü's commentary to the Tso chuan to its original form, for it had been cut up and included piecemeal in K'ung Ying-ta's T'ang dynasty versions of the Classics and their orthodox commentaries. Ku Yen-wu and Hui Tung would later bring to completion such reconstructions of Tu Yü's commentary.^[64]

The Han Learning editors of the Imperial Library catalog (completed in the 1780s) praised Chao's research as a model for k'ao-cheng scholarship on the Annals. They noted that Ming scholars had been unable to maintain Chao's high standards of research:

His clarifications and explanations are all correct. Not only has he added to Tu [Yü's] commentary, he has also contributed to the study of the Tso chuan . So much so that the points left unsaid by the sage are brilliantly illuminated. Consequently, Chao Fang [has presented] impartially the theories of the schools [of thought] regarding the Annals .

Despite his predilection for the Tso chuan and Tu Yü's commentary, Chao Fang was not averse to using the Kung-yang and Ku-liang New Text commentaries to correct errors in the Tso chuan. By drawing on both the Old Text and New Text commentaries, Chao Fang, according

[63] Sung Lien, "Hsu" (Preface) to Chao Fang, Ch'un-ch'iu shu-tz'u . See also Chao Fang, "Hsu" (Preface) to his Ch'un-ch'iu shu-tz'u, and Hsu-hsiu Ssu-k'u ch'üan-shu t'i-yao , vol. 2, pp. 736-37.

[64] See my "Philosophy versus Philology," pp. 186-88. See also Ssu-k'u ch'gan-shu tsung-mu , 28.13b-18a, and Goodrich et al., eds., Dictionary of Ming Biography , pp. 125-27.

to the editors, had "gotten the salient points of each" and encompassed both traditions on the Annals.^[65]

Chao Fang had made some interesting discoveries in his analysis of comparative phraseology in the Annals. Although he did not reject the "precedent tradition" of the Annals, Chao contended that earlier commentators had gotten bogged down in idle discussion of praise-and-blame and thereby missed the Annals' raison d'être (shih-chung ). Even Sung Neo-Confucians, according to Chao, had failed to grasp the "esoteric points of the Five Classics" (wu-ching wei-chih ): "Consequently, I say: If the meaning of the Annals is unclear, then the knowledge of scholars will be insufficient to understand the sage. As a result, they will not follow the teachings of the Annals ." For Chao Fang, it was important to grasp correctly the teachings of the Annals in order "to be able to talk about the statecraft meanings of the later sage [that is, Confucius]."^[66]

Interestingly, Chao was critical of the Tso chuan because its records of historical precedents (shih-li ) were divorced from the meanings (i ) around which Confucius had organized the Annals. Although he recognized that the New Text commentaries were strong precisely because they elucidated these "meanings," Chao was critical of the Kung-yang and Ku-hang commentaries because they failed to take into account the historicity of Confucius's chronicles. Chao's attempts to explicate the eight "general precedents" (fan-li ) in the Annals, which Tu Yü had drawn from the Tso chuan, represented an interesting compromise position.

Some earlier commentators had rejected the notion that there were any general precedents at all. Others argued that "the Annals originally had no precedents." What scholars like Tu Yü had done was "to create precedents based on the traces of events" described in the Annals. To deflect this powerful line of criticism (which Yao Chi-heng and others would later develop further), Chao's teacher Huang Tse said: "The history of [the state of] Lu has precedents. The sacred Classics have no precedents. It is not that they lack precedents, but only that they use meanings as precedents."

Chao Fang similarly tried to escape from the polarized positions on the Annals. Seeing a larger pattern that encompassed both the "meanings" and "precedents" of the Annals, Chao insisted that the precedent

[65] Ssu-k'u ch'üan-shu tsung-mu , 28.15a-16b.

[66] Chao Fang, "Hsu" (Preface) to his Ch'un-ch'iu shu-tz'u .

tradition was not able to grasp the totality of Confucius's teachings. Accordingly, Chao recommended his shu-tz'u pi-shih approach to the Annals as a way to cut through the impasse while affirming the "uniform chronicles" (t'ung-chi ) of Lu.^[67]

Chao Fang's ability to transcend the debates surrounding the precedents in the Annals and to focus instead on the Annals ' overall framework first impressed Mao Ch'i-ling (1623-1716). A leading protagonist in the defense of the Old Text chapters of the Documents Classic, Mao Ch'i-ling also compiled a work on the Spring and Autumn Annals , building on Chao Fang's earlier discovery of the "interrelation between Confucius' use of words and his comparative evaluation of historical events" (shu-tz'u pi-shih ).

Using the study of ritual to decipher the Annals , Mao Ch'i-ling created twenty-two divisions of "meanings and precedents" (i-li ) drawn from the comparative phraseology pioneered by Chao Fang. The editors of the Imperial Library catalog noted in their account of Mao Ch'i-ling's study that Mao had differentiated between the Rituals of Chou as an example of references to early Chou imperial governmental structures and the Tso chuan as an account of events dealing with imperial vassals (chu-hou ). This distinction allowed Mao Ch'i-ling to affirm that the Chou-li provided a model of ideal government and that the Tso chuan was a record of institutional reform.

Witness to the travail that had befallen China after the Ming debacle, Mao Ch'i-ling feared that attacks on the Classics by Yen Jo-chü and others, if left unchallenged, would cut literati off from their classical heritage. Interestingly, his efforts to read reformist themes into the Annals smacked of New Text commentaries. In effect, Mao was interpreting the Tso chuan as a record of change, unlike the Rituals of Chou, which he interpreted as the unchanging essence of statecraft.^[68]

A subtle shift in Ch'un-ch'iu-hsueh was nonetheless evident. Like Mao Ch'i-ling, Chuang Ts'un-yü would find Chao Fang's approach to the Annals a useful way to balance classical orthodoxy against contemporary reformism. Evading fixations on the past was no easy task. The Classics had to be freed from their post-Han shrouds. An alternative vision of the classical ordering of the world would preserve the Classics and affirm the need for change and adaptation.

[67] Ssu-k'u ch'üan-shu tsung-mu , 28,15a-b. See also Goodrich et al., eds., Dictionary of Ming Biography , p. 127.

[68] On Mao Ch'i-ling's position see Ssu-k'u cb'üan-shu tsung-mu , 29.17a-b.

Guiding Principles in the Annals

Chao Fang's study of the Annals appealed to a scholar-official like Chuang Ts'un-yü because he, like Chao in the fourteenth century, was trying to avoid a divisive insistence on one commentary to the Annals to the exclusion of the others. When it involved the Annals, the acrimony of the Han vs. Sung Learning debate was for the most part played out during the Ch'ing dynasty as another version of the long-standing fissure between the repressed Kung-yang and triumphant Tso commentaries. Chuang, like Chao, was seeking consensus.

In addition, Chao Fang's analysis of the Annals' phraseology was in tune with the k'ao-cheng temper of the eighteenth century. The compilers of the Ssu-k'u ch'üan-shu noted this with pleasure in their summaries of Chao Fang's works on the Ch'un-ch'iu. Such agreement was no doubt one reason for Juan Yuan's inclusion of Chuang Ts'un-yü's Correcting Terms in the Annals in the Ch'ing Exegesis of the Classics. Ts'un-yü's study could be seen as a continuation of Chao Fang's and Mao Ch'i-ling's pioneering application of linguistic criteria to Confucius's chronicles of Lu.^[69]

Although Chao Fang had been ecumenical in his selection from the Old and New Text commentaries to the Annals, his focus had remained on Tu Yü's annotations to the Tso chuan. Chuang Ts'un-yü, however, focused on the New Text Kung-yang Commentary to clarify the phraseology of the Annals. In his "Examples of Precedents in the Annals," for instance, Chuang chose his examples exclusively from the Kung-yang Commentary to reconstruct the code of precedents Confucius had entered in his chronicles.

Despite his affirmation of precedents in the Annals, Chuang Ts'unyü's discussion implicitly rejected the complex framework of detailed precedents that Tu Yü had "discovered" in the Tso chuan. In place of Tu's tedious listing of hundreds of historical events organized according to forty-two general precedents, Chuang (following the lead of Tung Chung-shu's New Text interpretation based on "Ten Guiding Points") substituted a simpler list of organizational schemes, which Confucius had used to construct the praise-and-blame phraseology of the Annals.^[70]

Seeing a "system of meanings" (i-fa) in the Annals, which ancient-prose stylists such as Fang Pao (1668-1749) and Yao Nai (1732-1815)

[69] Ssu-k'u ch'üan-shu tsung-mu . 28.15a-b. See also Hsu-hsiu Ssu-k'u ch'üan-shu t'i-yao , vol. 2, p. 736.

[70] Chuang Ts'un-yü, "Ch'un-ch'iu chü-li," pp. 1a-4b.

also discerned, Ts'un-yü articulated precedents in the Annals in light of "Ten Guiding Principles":^[71]

1. When there is no false resemblance between the noble and mean, the Annals uses the same title in both cases. When there is no false resemblance between good and evil, the Annals uses the same phrase in both cases.

2. When the Annals gives a detailed account without omissions, this indicates approval of correctness [of the action].

3. When one event appears twice, the first is specified in detail, while the second is generalized.

4. Once an event appears [in the record], it does not have to be repeated.

5. The Annals does not delay apportioning blame or criticism to condemn evil. When an incident occurs, the Annals does not have to first blame or renounce it in order to reveal its condemnation of evil.

6. The Annals blames and renounces and then condemns evil. When an incident occurs it can blame and renounce it in order to reveal the condemnation of evil.

7. The Annals selects the incident that warrants the most severe criticism and applies its criticisms there.

8. Blame must be placed on the most severe incidence [of evil].

9. The Annals criticizes the first occurrence of an incident. The Annals condemns the first incidence [of evil].

10. When the Annals repeats a key phrase, then one must scrutinize its context, because there must be something deserving praise within.

Compared with the Old Text historical precedents enumerated by Tu Yü, the precedents Chuang culled from the Kung-yang Commentary were general and flexible. Tu Yü's system placed the Annals within a tightly knit construct of specific historical classifications. Following Chao Fang, Chuang Ts'un-yü preferred a focus on judgmental language in the Annals rather than on the events themselves. In effect, Chuang was returning to Tung Chung-shu's interpretation of the Annals . Tung studied the Annals ' "latent language" in an effort to unveil the intent

[71] Ibid. See also Malmqvist, "Gongyang and Guuliang Commentaries 1," pp. 73, 77, 82, 123, 163, 168.

(chih ) behind Confucius's listing of events in Lu from 722 to 481 B.C.

By examining the phraseology of the Annals , Chuang Ts'un-yü was able to reconstruct a more flexible web of "words and meanings" drawn from the more voluntarist New Text tradition. He was making the "precedent tradition"—the dominant feature of the Tso chuan- based Old Text position since Tu Yü—secondary to an affirmation of "essential points" (yao-chih ) drawn from the New Text tradition of the Former Han.^[72]

The "Essentials Points" in the Annals

Chuang Ts'un-yü's use of the term "points" (chih) reveals a shift away from "precedents" (li) toward earlier traditions associated with the Annals. The "three classifications and nine points" (san-k'e chiu-chih) (which we shall discuss when we turn to Liu Feng-lu's role in the emergence of New Text Confucianism in Ch'ang-chou) had been the cornerstone of Ho Hsiu's New Text studies in the Later Han. Moreover, the word chih resonated powerfully with the Han legal term for "rescripts" (chih), used for imperial responses to memorials. In chapter 8 we shall describe how legal decisions during the Former Han dynasty had frequently been based on the Annals and the Kung-yang chuan. Ho Hsiu reaffirmed this link between legal and classical studies in his own New Text studies.^[73]

In his analysis of the "essential points" (yao-chih) in the Spring and Autumn Annals, Chuang explained: "The reason why people value the Annals is not because it is a history that records affairs. What is not included exceeds what is. On the basis of what is not included, we know [the meaning of] what is included. On the basis of what is included, we know [the meaning of] what is not included." Hence the significance of the Spring and Autumn Annals lay not in the events recorded, but rather in their encoding:

The Annals uses words [tz'u ] to form images [hsiang ]. It uses images to spell out laws [fa ]. It speaks to later generations in the world through the sublime mind-set [hsin ] of the sages. When looking at its words, one must use the mind-set of the sages to preserve [its images and laws]. . . . Therefore, those who are proficient in the Annals stop at the laws [taught by] the sages, and that's all.

[72] Chuang Ts'un-yü, "Ch'un-ch'iu yao-chih," pp. 1a-11b.

[73] See Vandermeersch, "Chinese Conception of the Law," p. 8.

The will (chih ) of the sages was expressed through the chronicle of events. The events themselves were secondary, however, to their pattern. Chuang turned the Han Learning preoccupation with facts, institutions, and "names and their referents" (ming-wu ) inside out. Han Learning scholars were, he thought, missing the forest for the trees.^[74]

Confucius's holistic vision of the Annals, thought Chuang Ts'un-yü, could be glimpsed in Chao Fang's efforts to grasp the interrelation of the Annals ' phraseology and its comparative evaluation of historical events (shu-tz'u pi-shih ). Chuang wished thereby to escape from the dustbin of historical precedents that had limited Confucius's message to idle scholarly classification and tedious historical detail, which influential Confucians such as Wang An-shih and Chu Hsi had both scorned.

History was released from its fixation on "events" (shih ) and reattached to the perennial rediscovery of "meaning" (i ). Confucius's words (tz'u ) were for Chuang Ts'un-yü the correct (cheng ) route to recover the true spirit of the past. Events described in the Annals were in and of themselves a dead end, a representation of chaos, death, and destruction. The events pointed instead to a higher order of meaning: the vision of the sage-kings that Confucius encoded in the Classics.

In a remarkable conversion (not unlike Tai Chen's turn late in his life from precise philology to abstract theory) Chuang Ts'un-yü, who was a friend and colleague of Tai Chen, stood Han Learning and k'ao-cheng on its head. As in k'ao-cheng, "words" (tz'u ) remained for Chuang the key to "meaning." A major research strategy of evidential research had been turned against its practitioners. Tai Chen contended in the 1770s that "through language we can penetrate the mind and will of the ancient sages and worthies." With regard to the mind and will of Confucius, Chuang Ts'un-yü argued that the Kung-yang chuan, not the Tso chuan, was the key to decoding the Annals.^[75]

In some ways, Chuang's discussion of twenty-two "essential points" in the Annals overlapped with his earlier account of "precedents." His comments on each "essential point," however, clarified what Chuang considered at the heart of the Annals. According to Chuang, the Annals had been composed as a lesson for an age of chaos. As we note in chapter 4, the Change Classic was a record for a time of order. It represented the fountain of ancient wisdom. The Annals recorded the demise of the classical world. It was a call to reform the present in the name of

[74] Chuang Ts'un-yü, "Ch'un-ch'iu yao-chih," pp. 1a-2a.

[75] Ibid., pp. 1a-1b. See also Tai Chen, Tai Chen wen-chi , p. 146. Cf. my Philosophy to Philology , pp. 27-29.

the past (t'o-ku kai-chih ): "The Annals orders chaos. It must express this [intent] through subtlety. This is what [Ssu-ma Ch'ien] meant by 'rituals prevent [improprieties] before they appear.' All entries [in the Annals ] express [this subtlety]. That is why the Annals has no meaningless entries [k'ung-yen ]." For Chuang Ts'un-yü, the Annals , a record of chaos, provided a schema for order. In essence, the Annals was united with the Change Classic in outlining the cosmos: "The Annals chronicles the affairs of Heaven and man, of inner and outer. It focuses on recording [events] in order to establish [sacred] teachings [li-chiao ]. Then it makes many connections [between events] and broadly finds a thread [of meaning] to encompass them. Consequently, the kingly way is complete within." A unified vision underlay the historical events chronicled in the Annals.^[76]

As a record of the state of Lu during the decline of the Chou dynasty, Confucius's Annals, according to Chuang, was a "reliable history" (hsin-shih ) whose choice of entries was intended "to be preserved and transmitted for ten thousand generations so that chaos would not arise again." Old Text historiography had lost the essential message of the Annals. Chronology was not the central issue. Rather, the Annals "subordinated the consecutive recording of events to their [overall] meaning." Instead of a list of historical precedents, the Annals was a work "inspired with dealing with Heaven's affairs" (chih t'ien-shih ). Within it, Confucius's "statecraft intent" (chien ching-shih chih chih ) was disclosed.^[77]

In the phraseology of the Annals, Chuang thought he had rediscovered the "points" (chih ) around which the Annals had been constructed. For example, Chuang summarized the intent behind Confucius's numerous references to burial practices as a general statement of the classical ideal of loyalty and filial piety:

The Annals gives meaning to death and burial. The sages regarded seeing off the dead as an important matter. When speaking of a ruler or a father, the reason why individual life was praised was the same reason why individual death was praised. When speaking of an official or a son, the burial of his immediate [the ruler or father], therefore, is the means by which one completed oneself. And one had to complete oneself through ritual. In this way, one reached the height of loyalty and filial piety.

[76] Chuang Ts'un-yü, "Ch'un-ch'iu yao-chih," p. 2a.

[77] Ibid., pp. 2b-3a.

On the frequent mention of the walling of cities and the formation of military alliances in the Annals, Chuang explained such entries as the "perversion of what was right" (hsieh-cheng ). He went on:

Each event has its counterpart. If a single meaning and a single pattern are sufficient to judge the general framework, then when the general framework cannot be ascertained, the Annals omits and does not record [the event]. The Annals is not a history of recorded events. It controls the narrative [yuehwen ] and enunciates meanings [shih-i ].

Chuang, in essence, agreed with Tung Chung-shu's Former Han estimation of the Annals. Tung had concluded: "Such is the way in which the Annals respects the good and emphasizes the people. Thus, although there are several hundred instances of warfare and aggression [in its 242 years of chronicles], it records them all one by one, thereby to express sorrow at the heavy extent of their harm."^[78] The applied moral message of the Annals differentiated it from other histories that were not Classics in the strict sense.

For Chuang Ts'un-yü the Former Han Kung-yang chuan and Ho Hsiu's Later Han annotations were the chief tools left from the early empire that could be used to reconstruct the Han "meaning" of the Annals. Chuang not only championed the Kung-yang chuan but also began to use the New Text commentary to explicate the other four Classics. Through these New Text accounts, the Old Text fixation on the Tso chuan and Tu Yü's tedious classification of precedents in the Annals could be superseded.^[79]

Chuang left unconsidered, however, how Kung-yang Confucianism would later apply to New Text controversies that he himself only vaguely perceived. The conservative intent behind his emphasis in the late eighteenth century on the Kung-yang chuan to counter the deleterious effects of the Ho-shen era is analytically distinct from the more radical consequences of full-blown New Text Confucianism in the early nineteenth century. Chuang Ts'un-yü's intellectual formulations were designed to counter what he perceived to be the radical political implications of Han Learning, which since the 1740s seemed to threaten the orthodox underpinnings of the state.

However, Chuang was not an orthodox Sung Learning advocate

[78] Ibid., pp. 4a-4b. See Su Yü, Ch'un-ch'iu fan-lu i-cheng , 2.2a, for Tung's remarks, and Fung, History of Chinese Philosophy , vol. 2, p. 77.

[79] Chuang Ts'un-yü, "Ch'un-ch'iu yao-chih," p. 7b. See also Li Hsin-lin, "Ch'ing-tai ching chin-wen-hsueh shu," p. 63.

either. Chu Hsi, for example, had never used the Spring and Autumn Annals to advance his classical teachings and, in fact, had relegated it to secondary importance. By revealing the Annals as the key to Confucius's classical vision, Ts'un-yü was making an end run around Chu Hsi and reclaiming the New Text portrait of Confucius. By expressing his views through a text as unorthodox as the Kung-yang chuan, Chuang Ts'un-yü was effectively countering not only Han Learning but also orthodox Sung Learning as well. "True Han Learning" cut both ways. Reconstruction of Former Han New Text Confucianism would inspire Chuang's followers to challenge all post-Han political discourse based on the Old Text Classics.

Although Chuang Ts'un-yü relied on Confucian theory, the textual basis for his theory was circumscribed by Han dynasty sources. Despite his mastery as a young scholar of the Ch'eng-Chu Neo-Confucian tradition and his subsequent success in and service for the civil service examination system, Chuang recognized late in his life that a new theoretical framework was required to legitimate his opposition to Ho-shen. Chuang Ts'un-yü's importance in the reemergence of New Text Confucianism should therefore not be underestimated.

Chuang could not have foreseen that his bold initiative would eventually lead to a radical reformulation of the classical language of politics and in turn affect the legitimation of the Confucian state he was trying to preserve. But he could see that new ideas were required to defend the classical legacy. The k'ao-cheng dismantling of Sung Learning would, he correctly perceived, leave the state with a bankrupt ideology. This was why he defended the Old Text Documents and affirmed the Rituals of Chou. Responding to what he viewed as a moral breakdown in imperial politics, Chuang sought to authorize Confucian activism and, so, turned to the Kung-yang tradition for his personal moral high ground.

Ts'un-yü's conservative posture, as we discussed earlier, represented both local and national interests. The interstices of classical theory and political discourse, interesting for the history of ideas, take on added meaning when viewed in light of (1) the long-term Chuang lineage traditions in Ch'ang-chou and (2) the threat Ho-shen posed to the Chuangs' continued status as a "professional elite" in national affairs.

Ts'un-yü's turn to Kung-yang Confucianism is evidence of the deep inroads Han Learning had made in Ch'ang-chou society. New scholarly strategies were afoot, and the transmission of Chuang lineage traditions from Chuang Ts'un-yü to Liu Feng-lu took place within the larger con-

text of a lineage adjusting its cultural resources and scholarly agenda from Sung Learning to Han Learning. No "secret transmission" of New Text doctrines was at work here. Chuang made a conscious decision largely based on political convictions, to turn to Kung-yang Confucianism. In so doing, the grand secretary transformed and redirected the scholarly traditions of his distinguished and powerful lineage.

Two grandsons, Chuang Shou-chia and Liu Feng-lu, carried on his teachings. Along with Sung Hsiang-feng, Chuang Ts'un-yü's nephew, they all studied under the auspices of the Chuang lineage, encouraged by Ts'un-yü. As we have seen, Liu Feng-lu's mother brought her son to her father for instruction at an early age. Liu studied the Classics and other ancient texts under him before turning to Tung Chung-shu's Ch'un-ch'iu fan-lu (The Spring and Autumn Annals ' radiant dew) and the Kung-yang chuan based on Ho Hsiu's Kung-yang chieh-ku.

Liu Feng-lu absorbed the "esoteric words containing great principles" (wei-yen ta-i ) that his grandfather Chuang Ts'un-yü had stressed in his classical writings and teachings. Ts'un-yü died when Liu Feng-lu was only twelve. The remainder of his education was guided by younger, but still senior, members of the Chuang lineage, whose links to Han Learning and evidential research ran deeper than they did for Ts'un-yü. Chief among Liu's mentors was Chuang Shu-tsu (1751-1816). Denied official appointment because of Ho-shen's intervention, Shu-tsu devoted himself to classical philology and New Text studies. In the hands of Liu Feng-lu, New Text Confucianism would represent the wedding of Chuang Shu-tsu's Han Learning philology and the Kung-yang studies of Chuang Ts'un-yü.^[80]

[80] Chu-chi Chuang-shih tsung-p'u (1883), 8.30b-31a, 8.36a. See also Wang Niensun, "Shen-shou fu-chün hsing-shu" (1876), 12.46b, and Liu's own account in his Liu Li-pu chi (1830 ed.), 10.25a-25b. Sung Hsiang-feng was the son of Chuang P'ei-yin's (1723-59) third daughter, who married into the Su-chou Sung family.