Three
Government, Society, and State: On the Political Visions of Ssu-ma Kuang and Wang An-shih

Peter K. Bol

Contrasts and Contexts

As scholars and political leaders Ssu-ma Kuang (1019-1086) and Wang An-shih (1021-1086) articulated fundamentally different political visions. Their views presented the literati with a classic choice between an activist government, which sought to manage social and economic developments in the interest of all, and a more limited government, which sought to maintain necessary public institutions at minimum expense to private interests. They found, once Wang gained power, that their differences were irreconcilable. Ssu-ma wrote Wang, in a letter that criticized all his major policies: "What I have said here runs exactly counter to your ideas; I know you will not agree. But although our directions are different, our ultimate goal is the same. You wish to gain power to pursue your way and benefit all the people, and I wish to leave office to pursue my purpose and save all the people." Even if they shared the common goal of serving the general welfare, their courses diverged so greatly that it is hard to believe that their ends could have been the same. Wang An-shih responded that they had differed in every instance on policy, in spite of having long been friendly, "because the methods we have adopted are for the most part different." Their "methods" (shu ) had to do with bureaucratic process and policy (cheng ), but they were also the methods of scholarship and learning (hsueh ) that each relied upon to justify his ideas about what government ought to do. Wang explained in 1069 to the emperor: "I certainly wish to aid Your Majesty in accomplishing something, but today customs and institutions are all in ruin. . . . If Your Majesty truly wishes to use me . . . we should first discuss learning, so that you are convinced of the necessary connections in what I have

learned." And, in response to the emperor's comment that some believed Wang's learning had not prepared him for practical leadership: "Methods from the Classics (ching shu ) are the means with which to manage the problems of the age (ching shih-wu ). . . . The priority of the moment is to change customs (pien feng-su ) and establish institutions (li fa-tu )." Ssu-ma Kuang was just as confident that his historical studies, which he was presenting to the same emperor, fully supported his views. Their visions were political and institutional, but also cultural and moral. Both men thought they knew how men ought to learn and where their learning should come from.^[1]

This chapter asks how Wang and Ssu-ma reached such different conclusions about the correct functioning of government. Rather than take Wang's advocacy of political reform and Ssu-ma's conservative opposition as the starting point of analysis and so depict them as polar opposites, it is more useful, I think, to view them as men who, responding to a shared set of questions, exhibited generational similarities as well as crucial differences. We can then see that the political stands they took after 1068 are related to the visions they developed earlier before they became political leaders. It is particularly misleading to see Ssu-ma only as an opponent of "change" and to ignore his own positive vision. Although past studies have sometimes inclined to polar analysis based on attitudes toward the New Policies, they have established many of the differences between the two men. They may serve here to remind us of whom we are speaking.

Contrasts

Higashi Ichio's study provides a point of departure for reviewing some of the differences between Wang and Ssu-ma.

Regional Origin . Wang came from the south (Lin-ch'uan county in Fuchou in Chiang-nan-hsi circuit, south of Lake P'o-yang). His region had been growing dramatically in population and wealth. Ssu-ma came from the north (Hsia county in Shan-chou in Yung-hsing-chün circuit, west of Lo-yang). His home had seen little dramatic growth and produced far fewer officials through the examination system. Wang served as a local official only in southern areas (Huai-nan, Liang-che, and Chiang-nan circuits) and

[1] Ssu-ma Kuang, Ssu-ma Wen-cheng kung ch'üan-chia chi (Wan-yu wen-k'u ed.) 60:725 (cited hereafter as SMK ); Wang An-shih, Lin-ch'uan hsien-sheng wen-chi (Peking: Chung-hua, 1959) 73:773 (cited as LCC ); Hsu tzu-chih ch'ang-pien shih-i 4:3b, in Li T'ao, Hsu tzu-chih t'ung-chien ch'ang-pien (rpt. Taipei: Shih-chieh, 1964; cited as HCP ). As Chang Hao has argued, Sung ideas about statecraft generally entailed a larger social-moral vision, rather than focusing narrowly on "bureaucratic statecraft" in the manner of the late Ch'ing; see his "Sung Ming i-lai ju-chia ching-shih ssu-hsiang shih-i," in Chin-shih Chung-kuo ching-shih ssu-hsiang yen-chiu t'ao-lun wen-chi (Taipei, 1984), pp. 3-19.

retired to Nanking. Ssu-ma served only in the north (Ching-hsi, Chingtung, and Ho-tung circuits) and retired to Lo-yang. Thus they experienced two very different kinds of local society and economy.^[2]

Social Background and State Service . Both Ssu-ma and Wang came from landowning, literati (shih ) families that had been successful in the examinations from their grandfathers' generation on. Wang was the first official in his family to rise higher than a prefectural post. He claimed that he depended on his salary to support his family and became a strong advocate of increasing official salaries. Ssu-ma Kuang's father, Ssu-ma Ch'ih (980-1041), reached high office, however, and was able to bring relatives into the civil service through yin privilege. Ssu-ma passed the chin-shih examination in 1038, but he had been granted official rank five years before.^[3]

Umehara Kaoru has argued that the official careers of the two men reflected their family traditions. Wang was seriously concerned with local government, remaining in local posts by choice in some instances. Some see his years as administrator of Yin county (1047-1050) as the origin of the New Policies. Ssu-ma, however, left local service as quickly as possible, using his father's connections to gain a capital post. In this view Wang's commitment to local government led him to envision a government that would "include the people," whereas Ssu-ma's desire to join the "bureaucratic aristocrats" led him to see government as by and for the scholar-officials. This may overstate the case. But from family tradition and personal experience Wang had reason to believe that officials could take the lead in further developing society and economy at the local level. Ssu-ma's background gave him reason to believe that local affairs were best managed by old, locally prominent families, who could also provide conservative and experienced men to guide government and preserve stability. There is little solid evidence for the claim that Ssu-ma was a "representative" of the "class" of large landlords and aristocrats or that Wang represented medium and small landlords. However, we shall see that Wang wished to break the power of the wealthy to make others dependent upon

[2] Higashi Ichio, OAnseki jiten (Tokyo: Kokusho kankokai, 1980), pp. 236-37. Higashi is the author of a popular study of Wang and Ssu-ma, OAnseki to Shiba Ko (Tokyo: Chusekisha, 1980). This work is in fact largely devoted to Wang. He has also written the most extensive account of the New Policies, OAnseki shimpono kenkyu (Tokyo: Kazama shobo, 1970). Local histories record 179 degree holders for Wang's native Fu-chou in Northern Sung but only two for Ssu-ma's Shan-chou. See John W. Chaffee, The Thorny Gates of Learning in Sung China: A Social History of the Examinations (Cambridge: Cambridge University Press, 1985), pp. 197, 200. The figure for Shan-chou may be affected by the loss of records in the north. The primary biographical sources for the present study are, for Wang, Ts'ai Shang-hsiang, Wang Wen-kung nien-p'u k'ao-lueh (Shanghai: Jen-min, 1959), and Higashi Ichio, OAnseki jiten , and, for Ssu-ma, Ku Tung-kao, Ssu-ma t'ai-shih Wen-kuo Wen-cbeng kung nien-p'u (1917).

[3] Ku Tung-kao, 1.4a-4b.

them, while Ssu-ma thought government should not interfere with the existing social order.^[4]

Scholarship . According to Higashi, as well as James T. C. Liu and Ch'ien Mu, this difference can be described in terms of texts: Wang was a Rites of Chou (Chou li ) scholar and Ssu-ma a Spring and Autumn Annals ( Ch'un-ch'iu ) scholar. The two texts provided alternative models for thinking about the role of the state in society. Higashi notes that the Spring and Autumn Annals , although usually taken as a model for moral judgments in history, is also about political hierarchy, the issue with which Ssu-ma Kuang begins his Comprehensive Mirror for Aid of Government .^[5] The Rites of Chou , in his view, directs attention at preserving and benefiting the livelihood of the people.^[6]

Wang, author of the official New Policies commentary on the Rites , did write, in about 1070: "Managing wealth takes up the half of this one work the Rites of Chou ." But the Rites more obviously depicts an elaborate system of state institutions as the basis for political unity. In Sung and later periods the Annals was probably used less to justify moral judgments than to argue both for the centralization of authority and for the existence of limits on that authority. Wang discounted the Annals , but Ssu-ma defended the Rites against the charge that it was a later forgery. It is not hard to find congruence between their ideas and their understanding of these two texts. Wang did advocate the creation of institutions that extended the scope of state activity. Ssu-ma was more concerned with the stability and survival of the dynastic system.^[7]

Robert Hartwell has described the difference between Wang and Ssu-ma

[4] Umehara Kaoru, "Shiba Ko to O Anseki," Rekishi to chiri , 1973, no. 11:106-19. The association of Wang and Ssu-ma with different classes of landlords, the orthodox position in Chinese Marxist historiography, has recently been criticized. See, for example, Ku Ch'an-fang, "Ssu-ma Kuang yü Wang An-shih pien-fa," Chin-yang hsueh-k'an , 1984, no. 2:67-74.

[5] Ssu-ma Kuang, Tzu chih t'ung-chien (Peking: Ku-chi ch'u-pan she, 1956) 1:2-6.

[6] Higashi, OAnseki jiten . For a general discussion of how the difference was perceived, see James T. C. Liu, Reform in Sung China: Wang An-shih (1021-1086) and His New Policies (Cambridge: Harvard University Press, 1959), pp. 30-33. See also Ch'ien Mu, Sung Ming li-hsueh kai shu (Taipei: Hsueh-sheng, 1977).

[7] LCC 73.773. For a study of the Chou li 's system and its historical background, see Sven Broman, "Studies on the Chou Li ," Bulletin of the Museum of Far Eastern Antiquities 33 (1961): 1-88. For the Annals in Sung, see Alan Thomas Wood, "Politics and Morality in Northern Sung China: Early Neo-Confucian Views on Obedience to Authority" (Ph.D. diss., University of Washington, 1981), in particular his account of Sun Fu. For Yuan, see John D. Langlois, Jr., "Law, Statecraft, and the Spring and Autumn Annals in Yuan Political Thought," in Yuan Thought , ed. Hok-lam Chan and Wm. Theodore de Bary (New York: Columbia University Press, 1982), pp. 89-152. These two studies show how "praise and blame" could be used to address political problems rather than merely being a vehicle for moral didacticism. For Ssu-ma's defense of the Rites , see SMK , 42.538-39. However, Higashi claims that Ssu-ma thought that the Chou li was indeed a later forgery; OAnseki jiten , p. 237.

in terms of their attitudes toward history while noting their attachment to the two texts. In his view Wang was a "classicist"; he believed that the Classics "depicted the ideal society and provided an absolute standard for judging current policies." Ssu-ma Kuang practiced "historical analogism," the "use of principles abstracted from historical models" to judge policy. (Thus neither was merely a "moral didact" concerning himself with apportioning "praise and blame to the participants in historical events as a means of encouraging the good and warning the evil.") The distinction is useful. Wang held that universally valid ideas could be inferred from particular authoritative texts (thus his official commentaries on the Rites of Chou, Odes , and Documents ). Ssu-ma believed that the necessary principles of government could be inferred from the past conduct of affairs. I would not say, however, that texts determined their visions, but suggest that one kind of text did lend itself more readily to one kind of vision than another. However, their intellectual interests were broad. Wang did not limit himself to the Classics, nor Ssu-ma to the histories. Both thought the examination system ought to test knowledge of the Classics rather than literary skill and memorization, yet both wrote in the various genres of literary prose and poetry. They also wrote on the Book of Change and the Classic of Filial Piety .^[8]

We should also note that both Wang and Ssu-ma claimed to speak for the Ju ("Confucian," or "scholarly," or "classicist") tradition. Yet they differed greatly in their appreciation of "earlier Ju." Wang championed Mencius, while Ssu-ma promoted Yang Hsiung of Han and developed an interest in numerology. Ssu-ma's later critique of Mencius is well known, but Wang did not reject Yang Hsiung. They also differed in their attitudes toward non-Ju intellectual traditions. Wang was known to be eclectic, finding room for the Lao tzu and Chuang tzu in his scheme of things. Wang also had an active interest in Buddhism and wrote commentaries on various sutras. Ssu-ma criticized eclecticism, but he too wrote a commentary on the Lao tzu . It is fair to say, I think, that both men eventually came to be fairly inclusive in their scope, laying claim to the great books of the past, thinking about the processes of heaven-and-earth, and concerning themselves with human nature and the mind. Their efforts represented, I believe, an attempt to show that the principles they espoused were truly universal.^[9]

[8] Robert Hartwell, "Historical Analogism, Public Policy, and Social Science in Eleventh-and Twelfth-Century China." American Historical Review 76, no. 3 (1971): 690-727. For Wang An-shih's writings, see Yü Ta-ch'eng, "Wang An-shih chu-shu k'ao," Kuo-li chung-yang t'u-shu-kuan kuan-k'an , n.s. 1, no. 3 (1968): 42-46. For Ssu-ma's historical scholarship, see Ts'ui Wan-ch'iu, T'ung chien yen-chiu (Shanghai: Commercial Press, 1934), pp. 32-39; for his other writings, see Fumoto Yasutaka, "Shiba Onko no gakugyo ni tsuite," Boeidaigakkokiyo 11 (1965): 1-79. Wang's commentaries on the Change and the Classic of Filial Piety are now lost, although his collected writings contain essays on the Change . Ssu-ma's commentary on the Change (SKCS ) is unfinished; his Classic of Filial Piety commentary is extant.

[9] Ssu-ma wrote commentaries on Yang's T'ai hsuan and Fa yen as well as a minor work entitled "Hidden Vacuity" (Ch'ien hsu ), modeled on the T'ai hsuan . Known editions of the latter are elaborations on an original unfinished work, making it difficult to know what remains of Ssu-ma in the present text; see Ssu-k'u ch'üan-shu tsung-mu t'i-yao , vol. 3, pp. 2230-31. Nevertheless, this work, which treats cosmological and numerological issues in relation to political institutions, has repeatedly been used as the basis for discussing Ssu-ma's philosophy. Wang's commentaries on the Analects and the Mencius are now lost. Remnants of Wang's Lao tzu commentary have been collected in Jung Chao-tsu, Wang An-shih Lao tzu chu chi-pen (Peking: Chung-hua, 1979). Wang's commentary on the Surangama Sutra, now lost, is well known, but he also wrote commentaries on the Diamond Sutra and Vimalakirti Sutra; see Chikusa Masaaki, "Ssu-ma Kuang Wang An-shih yü fo-chiao," in Chi-nien Ssu-ma Kuang Wang An-shih shih-shih chiu-pai chou-nien hsueh-shu yen-t'ao lun-wen chi (Taipei: Wen shih che, 1986), pp. 477-87. Wang's decision to convert his estate into a monastery in 1084 suggests his devoutness. However, Ando Tomonobu notes that no compelling explanation exists for a presumed conversion; see Ando, "O Anseki to Bukkyo—Shozan yinseiki o chushin to shite," TohoShukyo 28 (1966): 20-34. A convincing case has yet to be made for Buddhist inspiration in Wang's political vision. Chikusa argues that Wang was not a Buddhist layman, that his creation of a monastery was a common act of filial piety among well-to-do literati, and that his interest in Buddhist texts proceeded from an interest in supporting his Confucian views. Ssu-ma's Lao tzu commentary survived in the Taoist canon under the title Tao te chen ching lun , 4 ch., with the author given as a "Mr. Ssu-ma." However, the introductory comments and the first section correspond to the 2 ch. Lao tzu tao te lun , mentioned in the "Record of Conduct," as that text is described by Ch'ao Kung-wu in his entry for Wen-kung tao te lun shu-yao of 2 ch.; see Ch'ao's Chün-chai tu shu chih (rpt. Taipei: Commercial Press, 1968), 4.824.

Modern scholars usually overlook a contrast of considerable importance to Wang and Ssu-ma. Wang An-shih was an active literary man for much of his career—a noted poet and anthologist of T'ang poetry, a calligrapher, and a persuasive essayist in the "ancient style" (ku wen ) mode. We know, for example, that Wang's Miscellaneous Ideas from South of the Huai (Huai-nan tsa-shuo ) and his Commentary on the Great Plan (Hung fan chuan ) were circulating in the early 1060s and did much to establish his reputation as an intellectual leader. Wang An-shih the literary man, classicist, and eclectic thinker came together in a work from the 1070s entitled Explanations for Characters (Tzu shuo ). Through an analysis of the way in which the constituent elements in a written character formed a whole, Wang sought to establish the true values or purposes of the things and affairs to which the character referred. Yet this work, which claimed that the sages had created the characters on the basis of what was "so-of-itself" (tzu-jan ) or natural, in the end reminds us that Wang's roots were in the literary tradition of culture, even as he railed against those who were "merely" literary.^[10]

[10] Hou Wai-lu and his collaborators believe that the essays in Wang's collection include those of the Huai-nan collection; see Chung-kuo ssu-hsiang t'ung-shih , vol. 4, pt. 1 (Peking: Jenmin, 1959), pp. 421-22. Hou et al. note that Lu Tien saw these works before 1065 in the south. Lu's testimony shows that, irrespective of when Wang wrote his essays, it was still possible for literati to read them as new and startling works in the 1060s; see Lu, T'ao shan chi (TSCC ), 15.164-65. Liu An-shih (1048-1125) later said that in the early 1060s, when both Wang and Ssu-ma were in Kaifeng and had become friends, Wang's essay was circulating and "all admired him and compared him to Mencius." See Ma Yung-ch'ing, Yuan-ch'eng yü-lu (TSCC ed.), 1.6. In 1070 some examination candidates were already citing Wang's commentary on the Great Plan ; see Li T'ao, Hsu tzu-chih t'ung-chien ch'ang-pien (Peking: Chung-hua, 1986), 215.5246. Wang's Tzu shuo is now lost. Fifty-two entries (some of which may be incomplete) have been collected in K'o Chang-i, Wang An-shih p'ing chuan (Shanghai: Commercial Press, 1933), pp. 242-47. Further quotations will be found in the reconstructed commentary on the Lao tzu and Wang's Chou kuan hsin i (SKCS ). Winston Lo was one of the first modern scholars to point to the importance of this work; see his "Wang An-shih and the Confucian Ideal of Inner Sageliness," Philosophy East and West 26, no. 1 (1976): 41-53. Wang's introduction to the Tzu shuo is in LCC 84.879.

Ssu-ma Kuang, who began by making himself known through "ancient style" writing, came to deny he was a literary man and refused promotions to positions he considered "literary," while Wang accepted the same positions. Ssu-ma eventually concluded that literary ability did not justify a claim to political authority or wisdom. He accused Wang not only of staffing the Finance Planning Commission, the office that drafted the New Policies, with "gentlemen of literary talent" as well as fiscal experts, but also of being inspired by the thoroughly literary desire to "change all the old models and make everything new and unusual (hsin ch'i )." When the Comprehensive Mirror lambastes the "persuaders" who used rhetoric to guide political action, we may suspect that Ssu-ma has in mind all those who used skill with words to persuade others. However, Ssu-ma himself was not averse to rhetorical devices. He was a brilliant memorialist with a gift for descriptive narrative. Nor was he uninterested in cultural accomplishment. Wang could imagine inspiring ideas hidden in the form of a character; Ssu-ma sought to fix the proper forms for conducting affairs and using words. He revised ceremonial rituals from the I li for modern use, wrote rules for organizing the family, and prepared models for all manner of public and private correspondence.^[11]

The distinction I would make here pertains to the manner in which Wang and Ssu-ma understood texts, for both men did claim to know the "Way" for government and society from past texts, in contrast to those who believed that men should try to know it directly with their minds. Wang, who came from an area where learning was especially marked by literary interests, was convinced that the way texts were organized revealed the principles that informed the authors' thinking or the actions depicted by the texts. Ssu-ma, who came from a region known for sober scholarship rather than literary composition, granted that texts gave an account of how men might have acted but held that principles were to be inferred from the rela-

[11] For an account of Ssu-ma's unease with the literary, see Ho Chi-p'eng, "Ssu-ma Kuang te wen-hsueh kuan chi ch'i hsiang-kuan wen-t'i," in Chi-nien Ssu-ma Kuang Wang An-shih , pp. 339-60. For his charges against Wang, see SMK 60:720, 726, first and third letter to Wang. For attacks on "persuaders," see Tzu-chih t'ung-chien 3:100 and 6:221-22. Ssu-ma's works on family, ritual forms, and written formats are collected in Ssu-ma shih shu i (TSCC ).

tion between actions and their historical consequences. Wang saw history in the context of literature; Ssu-ma saw literature as part of history.

Political Attitudes . The simplest contrast, proposed by Higashi and many others, is that Wang was a (progressive) reformer who sought to defend the interests of the people and make China strong, while Ssu-ma, a (reactionary) conservative, opposed Wang's reforms because they threatened the privileged elite. Wang saw the need for change, Ssu-ma defended tradition. Wang wanted to increase wealth and power, Ssu-ma did not. Wang was willing to defend the borders, Ssu-ma preferred appeasement. Wang was good, Ssu-ma was not. Until recently such a categorical assessment was insisted upon by historians in the People's Republic of China, but since 1980 Wang and the New Policies have been condemned by some while Ssu-ma's opposition has been defended. Both have been depicted as defenders of the feudal order, with Wang furthering the interests of the feudal state at the expense of the people, and Ssu-ma defending the feudal ruling class against the state. In some eyes, Ssu-ma has gained the upper hand, and he is said to be neither reactionary nor conservative.^[12]

There have been more nuanced and complex views. Hsiao Kung-ch'üan calls Wang a "Confucian activist (yu wei che )" and notes Ssu-ma's lack of activism. Hsiao argues that Wang's activism took form in the New Policies, which were meant to bring about reforms in a planned and realistic manner. Wang stressed institutions but saw the need for talented men and envisioned an educational system to provide them. He aimed to "enrich the nation and strengthen the military" by increasing production and reducing the burdens on the people. Ssu-ma, however, in Hsiao's view had no systematic political vision but was bent on securing the autocratic power of the ruler and the subordination of his ministers. He opposed government activism on the grounds that the wealth of the state was fixed; any increase in the government's share led to a decrease in the amount in private hands. Yamashita Ryuji takes a view similar to Hsiao's. Anthony Sariti has effectively challenged the view that Ssu-ma was furthering autocracy, suggesting instead that he was defending the bureaucracy.^[13] Yet Hsiao and Sariti both

[12] The opening shot in the reevaluation of Wang and the New Policies appeared in 1980; see Wang Tseng-yü, "Wang An-shih pien-fa chien-lun," Chung-kuo she-hui k'o-hsueh , 1980, no. 3:141-54. For a defense, see Chou Liang-hsiao, "Wang An-shih pien-fa tsung-t'an," Shih-hsueh chi-k'an , 1985, no. 1:19-37, and 1985, no. 2:9-17. For an account of the reevaluation of Ssu-ma, see Huo Ch'ün-ying, "Chin-nien lai Ssu-ma Kuang yen-chiu chien-lun," Chin-yang hsueh-k'an , 1986, no. 3:81-83. A more detailed account of this debate will appear in Bol, "1086 and 1986: Reversing the Verdict?" Journal of Sung and Yuan Studies (forthcoming).

[13] Hsiao Kung-ch'üan, Chung-kuo cheng-chih ssu-hsiang shih (Taipei: Chung-hua wen-hua, 1954), 456-61, 482-84. Yamashita Ryuji, "O Anseki to Shiba Ko," Tokyo . Shina Gakuho 13 (1967): 135-50. For a similar view; see Yeh T'an, "I-fa tou-cheng chung te Ssu-ma Kuang," Hsi-nan shih-fan hsueh-yuan hsueh-pao , 1985, no. 4. Anthony Sariti, "Monarchy, Bureaucracy, and Absolutism in the Political Thought of Ssu-ma Kuang," Journal of Asian Studies 32, no. 1 (1972): 53-76.

show that Ssu-ma was particularly concerned with the proper organization of relations within the imperial system. In contrast to Wang, I would note, Ssu-ma argued for a sharp divide between public and private interests.

Wang has received far more study than Ssu-ma. More impartial historians such as Saeki Tomi, Wang Yü-ch'üan, and James T. C. Liu seem to me to represent a consensus holding that Wang and the New Policies above all served the interests of the centralized state. Professor Liu shows Wang "as a bureaucratic idealist who upheld the ideal of a professionally well-trained and administratively well-controlled bureaucracy as the principal instrument for the realization of a Confucian moral society." Although I suspect Ssu-ma Kuang would have thought this described himself, not Wang, it is prudent to agree that the historical effect of the New Policies was to "put the interests of the state. . . above everything else."^[14] In chapter 2 in this volume Paul Smith argues that the New Policies represented an attempt to reassert the state's control over a burgeoning private economy.

Philosophical Orientation . A few modern scholars have tried to account for Wang's desire for reform and Ssu-ma's opposition on the basis of their respective ideas about the Way (of heaven-and-earth; i.e., Tao ) and human nature (hsing ). Their conclusions are remarkably similar. Ch'eng Yang-chih develops a nurture-versus-nature dichotomy, in which Wang believes that men are formed by society (thus one must change society to transform men), whereas Ssu-ma believes that each man receives a fixed nature that determines what he can be (thus one must accept one's role). Others have now argued that Ssu-ma Kuang was really a materialist. Teraji Jun's Wang is a scientist, his Ssu-ma a moralist. In this view Wang, who does not believe in any resonance between heaven and men, must learn from the world how to make it serve man's interest, while Ssu-ma, who believes in resonance (a conclusion I do not fully accept), asks only that men try to be "good" by accepting the lot heaven has given them and obeying their superiors. Eleventh-century thinkers were divided over the nature of the relationship between heaven and man, or the natural and the cultural. For some—and, like Teraji, I would include Wang An-shih and Ou-yang Hsiu in this group—the true values for human society were to be inferred from human history rather than from a particular view of cosmic process or human nature. This contrasted with the view that norms for human conduct were established by heaven and that government should make men fit those

[14] Liu, Reform in Sung China , pp. 114, 115. See also Wang Yü-ch'üan, "Pei-Sung she-hui yü ching-chi cheng-chih," Shih huo yueh-k'an , 3, no. 11 (1936): 23-34, and 3, no. 12, 21-43, and Saeki Tomi, OAnseki (1929).

norms.^[15] But I do not think Ssu-ma Kuang believed that men should act in a certain way because heaven had decreed that they should or because it was natural for men to do so.

In sum, these scholars argue that Ssu-ma's belief structure led him to conclude that it was impossible for men truly to change the given situation; if they wished to survive, they had to learn to accept their predetermined roles. Wang An-shih, however, saw that men could be formed by their environment. He set out to change that environment for the common good. Clearly, evidence has been adduced for these views, which I have only caricatured here. But I think it can be shown that the political-social visions of the two men preceded their philosophical speculation, and that such natural-philosophical ideas as they espoused were intended to confirm further what they already knew. In contrast to those who would later be seen as forerunners of Tao-hsueh , Chang Tsai and the Ch'eng brothers in particular, they did not make ideas about heaven-and-earth the foundation of their thinking about human values. Rather both, in my view, illustrate Ou-yang Hsiu's advice that scholars should first seek to know human affairs; they would find that heaven-and-earth fit the same principles.^[16]

I would redescribe the contrast the earlier studies suggest in this manner: both Wang and Ssu-ma believed that most people were guided by material desires. Wang believed that as long as government policy offered most men an opportunity to satisfy their desires, policymakers could establish institutions as they saw fit and men would become accustomed to them. Ssu-ma Kuang believed that such efforts must be undertaken with the utmost caution; that certain social relationships, in particular that between superior and inferior, were necessary to hold society together; and that sudden attempts to change the ways in which men satisfied their desires could destroy those fragile relations and, as a consequence, the state. We can begin to understand these positions by turning to their intellectual context, that is, to the kinds of questions that confronted Wang An-shih and Ssu-ma Kuang when, as thoughtful and ambitious young men, they began their careers and formed their ideas.

[15] Ch'eng Yang-chih, "Wang An-shih yü Ssu-ma Kuang," Wen shih tsa-chih 2, no. 1 (1942): 1-17. Hou Wai-lu and his collaborators predictably treat Wang as a materialist (for the most part) and Ssu-ma as a pure idealist. Wang knows society can be restructured to change the way men act and to benefit all. Ssu-ma supposes that some are endowed with heaven-given knowledge and thus demands that all others accept the lot heaven assigns them. Hou Wai-lu, pp. 449-69, 511-21. For Wang as scientist, Ssu-ma as moralist, see Teraji Jun, "Tenjin sokansetsu yori mita Shiba Ko to O Anseki," Shigaku zasshi 76, no. 10 (1967): 34-62. For Ou-yang Hsiu, see Teraji Jun, "Oyo Shu ni okeru tenjin sokansetsu e no kaigi," Hiroshima Daigaku bungakubu kiyo 28, no. (1968): 161-87. For Ssu-ma as materialist, see for example Chao Chi-hui, "P'ing Ssu-ma Kuang ti che-hsueh ssu-hsiang," Chin-yang hsueh-k'an , 1986, no. 4:56-59.

[16] Ou-yang Hsiu, Hsin wu tai shih (Peking: Chung-hua, 1974), 59:705-6.

Contexts

If we ask why Wang and Ssu-ma defined the "problems" the way they did and how they justified their "solutions," we must inquire into the visions they brought to bear on the times. For this our primary context is the intellectual world in which they participated. There are important similarities in their intellectual ambitions. First, they followed the previous generation, men born approximately fifteen years earlier (ca. 1005), in believing that the literati needed to unite behind an ideal purpose, and that this had to do with politics and institutions. (Later there would be men who, like Chu Hsi [1130-1200], believed that individual moral regeneration was prior to institutional action.) Second, both Wang and Ssu-ma held that the goal of learning was to identify the principles upon which a coherent set of policies could be based. This contrasts, I think, with the previous generation's view that the goal of learning was to cultivate the correct attitude toward politics, and with its reliance upon the power of opinion.^[17] (Su Hsun [1007-1066], discussed by George Hatch in chapter 1 in this volume, is representative of this earlier view.) Third, Wang and Ssu-ma assumed that once the government adopted a program based on right principles the "affairs of government" (cheng shih ) would be perfected and the common interest realized. Their confidence in the perfectability of institutional action would not be shared by leading thinkers in the next generation (men born ca. 1035). Ch'eng Yi (1033-1107), for example, gave precedence to "moral conduct" (te hsing ) as individual moral self-cultivation, and Su Shih (1037-1101) saw "literary learning" (wen hsueh ) as a means of cultivating and expressing an individual identity. Su and Ch'eng were concerned with how literati should act independently of the political system. But Wang and Ssu-ma defined intellectual and moral values for the literati in terms of institutional ideals.

We may begin with the expansion of the examination system into an institution for selecting large numbers of officials from among the shih ("lite-rati"). Until the advent of the New Policies the exams tested the "civil arts": literary composition for the prestigious chin-shih degree, memorization of various Classics, histories, or ritual texts for the lesser degrees. Five Dynasties shih had included military power-holders, officials, and generally those families with traditions of state service. But the new exam system, by rewarding with office those who had mastered civil-literary-cultural (wen ) learning, shifted the definition of what it meant to be a shih away from family tradition and the possession of power toward education. Those who

[17] See James T. C. Liu's discussion of the role of opinion and the Ch'ing-li reforms; "An Early Sung Reformer: Fan Chung-yen," Chinese Thought and Institutions (Chicago: University of Chicago Press, 1957), pp. 105-31.

pursued literary learning in hopes of gaining office and defending the civil order, whether or not they came from families with traditions of service, were now licensed to think of themselves as shih . In effect the shih , the "elite" as opposed to the "commoners" (shu ) or "people" (min ), became literati.

Mastering culture and being civil was a shared purpose and practice that "worked" as an idea and as a reality. Perhaps it was too successful. By the 1020s and 1030s being cultured and civil had lost its patina of idealism, as skill in the civil arts came to be appreciated above moral commitment. A new turn in the intellectual history of Sung literati began when some literati objected that, as men who justified their participation in politics on the basis of their schooling in the civil-cultural-literary tradition, literati should aspire to the highest ideals of that tradition. One immediate focus of their criticism was the examination system itself They did not object to examinations: they objected to testing literary skill alone, on the grounds that "good" wen (writing; literature; cultural expression) should show the way to realizing the ideals of the civil-cultural-literary tradition.

Because early-eleventh-century shih defined their shared values in terms of wen , attempts to redefine literati values easily took the form of redefining the nature of "good" wen . "Ancient style" writing, inspired by T'ang models, became the vehicle for this movement. Advocates of the new style claimed that wen would be good if the composer sought guidance from the Way of antiquity, which had guided the sages in establishing the cultural tradition and civilization. He revealed what he had learned through his writing. In effect, "antiquity" was held up as a source of values, and "good" wen promoted those values. The examination system, some held, should favor prose over poetry to give men with ideas a chance to express themselves and be judged accordingly. "Ancient style" was both a literary and an intellectual movement; it supposed that a change of style was also a change of heart and that the values that informed the mind, evident in what a man said and how he said it, would also inform his conduct. Ou-yang Hsiu (1007-1072) became the leading figure in this movement. He was involved in affairs, promoted antiquity and the broad study of the cultural tradition, and called for universal standards of moral judgment; but he also created a persona of ironic, self-conscious detachment and disinterestedness, the appropriate attitude for one who aimed to transcend selfish interest. Men of Ou-yang's generation shared an extraordinary confidence in their ability to know what was right and practice what they had learned. Representative of this spirit is Ou-yang's three-part essay "On the Basis" (Pen lun ), written in 1042 as part of a campaign to persuade the court to turn to Fan Chung-yen (989-1052) and his allies for political leadership. Here he made the essential claim of the movement: that truly ordering or

governing the state (kuo-chia ) requires that government guide society by serving the interests of society.^[18]

We shall see that Ssu-ma Kuang and Wang An-shih had very different ideas about what was necessary to accomplish this. In examining their views I shall try to maintain a distinction among three notions: first, "state" (kuo-chia or kuo ), referring to the dynasty as a larger entity that includes both the people (min ) on the one hand and the government on the other (although it is sometimes used to mean the government itself); second, "government" (cheng ), referring to official institutions and their administration; and third, "society," the realm in which people pursue their private interests (sometimes called "all under Heaven," t'ien-hsia ). Both Ssu-ma and Wang view "society" as part of the "state," but they have very different views on how "government" and "society" should be related within it.

Making Reputations: Intellectual Careers,, 1038-1057

Ssu-ma Kuang (chin-shih 1038) and Wang An-shih (chin-shih 1042) began their careers just when an upwelling of idealism, a desire for political change, and a border crisis were pushing the court into giving men of the new movement, led by Fan Chung-yen (989-1052), a brief trial in the aborted Ch'ing-li "reform" of 1043-1044. Their two careers were advancing rapidly when, between 1055 and 1067, old allies of Fan's, including Ou-yang Hsiu, gained positions on the Council of State. In the 1040s it was easy for men seeking reputations for learning both to believe that literati ought to have a common purpose and to accept that the "ancient style" was the right vehicle for expressing their views on what that purpose was and how it could be achieved. But Ssu-ma and Wang responded to the "ancient style" call in different ways.

From common beginnings in the late 1030s to early 1040s, the two men took significantly different career paths. Ssu-ma early began a long period of service at the center, in the capital; Wang held back, remaining deliberately in local posts. But for both men the period 1057-1058, which saw Ssu-ma's return to the capital after a brief absence and Wang's submission of

[18] For the first side of Ou-yang Hsiu, see Liu Tzuochien, Ou-yang Hsiu ti chih-hsueh yü ts'ung-cheng (Hong Kong: Hsin-ya, 1963). For the second, see Ronald Egan, The Literary Works of Ou-yang Hsiu (1007-72) (Cambridge: Cambridge University Press, 1984). "On the Basis" appears in Ou-yang Hsiu, Ou-yang Hsiu ch'üan chi (rpt. Taipei: Shih-chieh, 1971), Wai chi 9:411-13 (part 1 of the essay) and Chi 17:121-24 (parts 2 and 3). The second part of the essay in translated in Wm. Theodore de Bary et al., Sources of Chinese Tradition (New York: Columbia University Press, 1960). Ou-yang excised the first part of the essay in his later years, presumably because it could by then be read as justifying the New Policies approach to government. See the fuller discussion of this essay in the Introduction.

his famed "Ten-thousand Word Memorial," marked a new phase both politically and intellectually.^[19] It is fair to say that beginning in 1057-1058 Ssu-ma and Wang were being treated as prospective candidates for leading central positions. Although their places in the factional alignments of these decades are not entirely clear, it seems that Wang's sympathies lay with those who had been associated with Fan Chung-yen's reforms of 1043-1044, while Ssu-ma's patrons were not known as supporters of the reforms. The departure from office of his backer P'ang Chi in 1054, for example, was followed by the return of Fan's allies.^[20]

At the very beginning of their careers Wang and Ssu-ma adopted the prevalent method of making themselves known: they sent their compositions to high-ranking officials. For this purpose they composed ancient-style writings and showed that they were, as Ssu-ma explained to a patron, "superior men, concerned with knowing the larger and more far-reaching."^[21]

[19] Ssu-ma took the chin-shih degree in 1038, Wang in 1042. But because Ssu-ma then went into extended mourning for both his parents, he and Wang both received their first appointments in 1042. These were the usual local-government positions. In 1046, however, Ssu-ma took a post at the State Academy and proceeded to hold capital posts for the next twenty-five years, with the exception of 1054-1056 when he took local posts under his foremost patron, P'ang Chi, who had been forced out of his position as chief councillor. 1057 thus marks the beginning of Ssu-ma's second career as a capital official; the direction of his intellectual efforts shifts at this point as well. Wang, on the other hand, had remained in local posts, refusing an invitation to test for an academic position at court in 1051. Yet he was not completely successful in maintaining the role of outsider, he came to the capital for a nonacademic post in 1054-1056, and in 1058 he took a post in the financial administration. Although Wang left the capital in mourning in 1063 and remained out of office until 1067, the year 1058, when he submitted his memorial, does clearly mark a new stage in his career.

[20] P'ang Chi (980-1063) served on the Council of State from 1045 to 1053, rising from assistant commissioner of military affairs to chief councillor (1051/10-1053/7). For membership on the Council I have followed the chart for 1041-1125 in Higashi, O Anseki jiten , pp. 211-20. He was forced out of office when he was unable to quell the charge, brought by Han Chiang (who would later become an enthusiastic supporter of the first New Policies), that he had arranged the death of a man who had been found guilty of trying to bribe him; see Sung shih 311:10201. I have not yet been able to determine P'ang's attitude toward the reforms of 1043-1044. However, he may well have been allied with another faction. Although together with Fan and Han Ch'i he had been rewarded for solving the border problems in Shensi in 1043/2, he was not appointed to the Bureau of Military Affairs with them in 1043/4. Moreover, Tu Yen, who was to be allied with Fan, replaced Hsia Sung (an ally of P'ang's) as commissioner of the bureau at the same time. P'ang only joined the bureau in 1045/1, immediately after Fan's group was replaced. Ssu-ma also sought the patronage of Sung Hsiang, who was at odds with Fan's arch rival, Lü I-chien, but was brought into the Council of State to replace Fan in 1045/1. See these dates in Sodai shi nenpo (Hokuso ) (Toyo bunko, 1967).

[21] SMK 58.696; LCC 77.810; SMK 60.727-28. We have Ssu-ma's own testimony from 1062 that he began his career with the aim of making a name for ancient-style writing; see SMK 59.712-713. There is little question about Wang's reputation as an ancient-style writer. Wang Ch'eng saw him, with the Sus and the Tsengs (Wang's cousins), as the successor to Ou-yang Hsiu; see Tung-tu shih-lueh (rpt. Taipei: Wen-hai, n.d.) 115.1. For Wang's decision to make a name in this way, see Ts'ai Shang-hsiang's account of relations between Ou-yang, Tseng Kung and Wang in Nien-p'u k'ao-lueh 2.47-48, 3.52, 3.56, and 3.83. Ssu-ma eventually concluded that the ancient style offered little of value. Wang, although he denied that literary skill was an adequate standard by which to evaluate literati, continued to value ancient-style writing in his later career; see LCC 75.798-99 and 76.802.

Wang An-shih and Ku Wen: Integrating Government and Society

In 1044, the year Wang's cousin Tseng Kung (1019-1083) recommended him to Ou-yang Hsiu, Wang wrote Tseng: "We learn from the sages and nothing more. When we learn from the sages our friends and teachers will necessarily be men who also learn from the sages. The words and actions of the sages are uniform; of course [our friends and teachers] will resemble each other."^[22] Learning from the sages led to uniformity because the values that guided the sages were uniform and consistent. The sages had acted according to a single Way. In 1045 Wang claimed to have understood this Way by reading the texts of the sages (i.e., the Classics) and seeing their unity:

My ignorance is such that I do not understand changes in affairs and priorities. I only trust in the sages. When I heard that in antiquity there were a Yao and a Shun and that their Way was the great middle, perfectly correct, and constantly practiced Way, I got their books, shut the door, and read them. I lost all sense of anxiety or joy. I threaded together top and bottom; I immersed myself in their midst. The smallest, seamless; the largest, boundless; I sought only to fathom them as one.^[23]

Wang explained in 1046 why such a unity could be found in the Classics, although they were not the work of a single hand or period:

The teachings [that led to] order and the commands [that effected] policy were what the sages called wen . When written on slips of wood and when applied to all the people they were one. As for the sages in regard to Way, their minds must have apprehended it. When they acted and made teachings and commands, there were root and branch and what came first and last. They instituted principles according to the situation but unified them at the ultimate point. When they wrote them down on slips of wood they were simply recounting how it was.^[24]

Wang's sages acted in full accord with the Way as something that exists of itself. But Wang does not conclude from this that he too ought to apprehend the Way directly with his own mind. Instead, he points to several mediating levels between the Way itself and himself in the present. There are the sages' mental apprehension of the Way, the sages' policies, the writ-

[22] LCC 71.755.

[23] LCC 77.810.

[24] LCC 77.812.

ten record of policy (which he assumes to be equivalent to the society that those policies created), and the Classics as texts that carry those written records. But far from being an obstacle, these mediating levels make it possible to accord with the Way in the present. Wang makes three claims: the sages' apprehension, application, and written records are consistent with each other and realize the Way; these visible levels can be analyzed to reveal a coherent system; and any apparent contradictions in the records are merely situational variations that are ultimately in perfect agreement. This allows Wang to conclude that understanding the unity or coherent system of the Classics is equivalent to understanding the Way.

Literati have the material from which they can know the Way of the sages, both the Way the sages grasped and the way the sages acted. Wang can also tell whether they have understood those ideas correctly. A lack of unity, integration, and coherence in writing shows that the writer has missed what is fundamental; this comes of picking and choosing what one likes in the Classics rather than seeking to understand their systematic unity. True wen , the kind that can be applied in the present, must be based on an understanding of the coherence of the sages. Wang makes this point and then writes, in the letter just quoted:

Therefore if one writes it on slips and it is good, then when it is applied to all the people it will always be good. The Two Emperors and Three Kings [of antiquity] were the ones who applied it to all the people and it was good; Confucius and Mencius were the ones who wrote it down on slips and it was good. They were all sages. Under different circumstances it will also be good.^[25]

Wang has now reached a conclusion that he will continue to hold as chief councillor. Unlike Ou-yang Hsiu, he believes that a policy that is in accord with the principles that made the Way of the sages coherent is necessarily correct, and that its consequences will necessarily be good.

Wang envisioned the ancient world, at least during times of "perfect order" (chi chih ), as a place where political authority took responsibility for ensuring the welfare of all the people. As administrator of Yin county, he writes in 1047:

I have heard that among the ancients, in times of perfect order, ruler and ministers applied the Way to provide employment for all the people. If there was a single common man or woman who did not partake of the benefit, they were most ashamed and worried about them. The blind and the deaf, the pygmy and the dwarf, each was able to use his talent to draw support from the appropriate office. The transformations brought on by their sincere minds were such that the hooves of the oxen and sheep could not bear being un-

[25] LCC 77.812. For a similar statement on composing wen that can be applied in the present, see LCC 77.811.

benevolent to the grasses and shrubs, as the poem [from the Odes] "Traveling over the Reeds" shows. How much greater [the effect] on their gentlemen and great officers (shih tai-fu ). This is why they achieved harmony between superior and inferior and called it a time of perfect order.^[26]

The achievement of "perfect order" in antiquity reveals the fundamental and primary task to be integration and unification, economic, physical, social, and moral.

The details of [the sages'] carrying out of policy and instruction and all that they shared with others, especially what they enjoined [upon men] as priorities, are clearly easy to know. They connected the roads and rivers. They organized paddy-fields and mulberry [plantings]. They erected dikes, dug irrigation canals, and dredged out rivers to prepare against flood and drought. They established schools and assembled the people to practice rituals and music in them, thus bringing them to submit through transformation. These they enjoined [upon men] as priorities and clearly easy to know.^[27]

As county administrator Wang did make these his priorities, and he took the lead in organizing irrigation projects and establishing a local school. He sharply criticized policies that the common people could obey only at the risk of losing their land, and made this a matter of principle. Yet Wang also held that leading the people to greater prosperity could require coercing them to share the burden. "You can enjoy the final result with lesser men," he wrote about his irrigation projects, "but it is hard to think about the beginning with them. Even when there truly is great benefit you still must force them."^[28]

Wang's conviction that government should take responsibility for organizing society and furthering the general welfare is apparent in a telling comment from 1047. The current deficit, he writes, "is not only due to unrestrained expenditures, it is also the result of our having lost the way to create wealth (sheng ts'ai ). To enrich the family [the emperor's], you draw on the state (kuo ), and to enrich the state, you draw on society (t'ien-hsia ); but if you wish to enrich society you draw on heaven-and-earth." Increasing prosperity was only possible through increased production; otherwise wealth would simply be flowing from one party to another, without general benefit. (The appointment of commissioners to oversee increasing agricultural production was one of the first New Policies.)^[29] The same conviction is clear in the position Wang took on state education during these years:

[26] LCC 74.780.

[27] LCC 82.866. In this passage Wang is approvingly quoting a colleague.

[28] LCC 75.794. For a detailed account of Wang's activities in Yin, see Ch'eng Kuang-yü, "Wang An-shih chih Yin shih chih chih-chi yü fo-chiao," in Chi-nien Ssu-ma Kuang Wang An-shih , pp. 141-66.

[29] LCC 75.795.

The world under heaven cannot do without policy and teaching [cheng and chiao : compare Ou-yang Hsiu] even for a single day; thus the world cannot do without schools even for a single day. The ancients organized all fields into well-fields, and the various levels of local and national schools were established in their midst. All policy came from the schools—the archery contests and wine-drinking ceremonies, the gatherings for music in spring and fall, the care for the aged and agriculture, the honoring of worthies and employment of the able, the examination of arts and selections of sayings, and even planning military campaigns, presenting the ears of the slain, and the interrogation of captives. . . . Day and night, all the students (shih ) saw and heard was the Way with which one orders the world and the state.^[30]

This was not how schools functioned any longer, but it was how they should function. (Establishing state schools, examination reform, and the creation of a new curriculum were to be part of the New Policies.)

The world of perfect order Wang proposed saw no distinction between the institutional concerns of government and the moral and economic life of society. As Wang wrote in 1053, "For the people of the Three Eras, wealth was not divided between public and private; . . . and engrossing was criminal."^[31] But Wang had to recognize that in the present there was private property and private trade; "engrossing" (chien-ping ), whether of land or of capital, was perfectly legal. Similarly, he knew that the unity of values of antiquity did not exist in the present. He wrote in 1058, after being criticized for his actions in a judicial matter:

The ancients "unified values (tao te ) to make customs the same [for all]." Therefore when literati looked to what the ancients had accomplished for standards there was no difference of opinion. Today families hold to separate ways and men to different virtues. Moreover, [the values] they hold are pressed by the force of degenerate customs and they are unable to be like the ancients in every case; how can difference of opinion be fully repressed?^[32]

Wang had to rely on persuading the literati to share his vision.

As an "ancient style" essayist during this period, Wang set out to show literati that by turning to the Classics and seeking the Way of the sages they would see the true purpose of government—the transformation of society into an integrated order—and would learn that there were sure ways of achieving that end. Several of the essays justifying this view will be cited later. Here I only note some of his "methods" in writing. First, he writes largely about the Three Eras of antiquity. Second, he gives particular attention to why it was possible to establish and maintain an integrated order.

[30] LCC 83.870.

[31] LCC 4.114; cf. 12.177.

[32] LCC 72.768, "Yü Wang Shen-fu shu" (second letter), gives the 1058 case and response. I have translated a similar but more elaborate passage found in LCC 75. 794.

Third, in discussing the Classics as texts, as seen in his writings on the Book of Change, Odes , and Documents , he explains why the particular arrangement of parts within a text forms a coherent system, and he uses this explanation to reach conclusions about the "root and branch" of the sages' Way, thus to define the sequence of policies and the structure of ideas necessary for achieving an integrated order in the present.^[33]

Ssu-ma Kuang and the Ancient Style: Ending the Cycle of Order and Chaos

It is not coincidental that Ssu-ma Kuang's extant "ancient style" essays date from 1042 and 1045, just before and after the Ch'ing-li reform, and 1056-1057, when the ancient style was finally triumphing in the examinations under Ou-yang Hsiu's leadership. For these essays challenge the assumptions of the movement whose mode of expression Ssu-ma had adopted. It is not surprising that at the end of this period he discounts the possibility that the ancient style had a special claim to intellectual value. But rejecting the claim that literati should be guided by an ancient ideal order impelled him to provide an alternative. The difference is already apparent in a 1039 piece, which contrasts those whose concern with ideals is merely literary with Ssu-ma's true Ju:

He read the books of the Former Kings, not to master the commentaries but to seek their patterns (li ). Once he had apprehended their patterns he did not simply recite them to deceive others, he would necessarily practice them in his personal life and in his community. If he had had nothing to spare, then [these patterns] would not have shined outside of these [settings]. Had they not shined it would have been as if the Way of the Former Kings was screened. Thus he looked for cases in the politics and customs of the state where [these patterns] had been realized or lost and composed poems and prose to make them known.^[34]

The Former Kings of antiquity had a Way. This Way can be grasped by seeing the patterns of their conduct in the Classics.^[35] One acts out these

[33] For examples see his "Explanation of the Sequence of Chou-nan Poems," LCC 66.701. Shimizu Kiyoshi has made this point in his discussion of the "Explanation"; see his "O Anseki no shu-nan shi ji-kai ni tsuite," in Toyogaku ronso (Tokyo: Uno Tetsuto kinenkai, 1969), pp. 491-510. See also "Explanation of the Images in the Change," LCC 65.697-700, which presents the 64 hexagrams as a coherent sequence teaching the "way of the superior man." The introductory section of the Great Plan commentary is another example; see LCC 65.685.

[34] SMK 69.851. This passage plays on Analects I.6, which says to study wen only if one has strength to spare after learning to behave ethically. In an article that traces Ssu-ma Kuang's acquaintance with Ou-yang Hsiu and the intellectual similarities and differences between them Ch'en Kuang-ch'ung points out that, while the two had many friends in common, to which both wrote poems, there are no poems of friendship between the two; see his "Ssu-ma Kuang yü Ou-yang Hsiu," Shih-hsueh chi-k'an , 1985, no. 1:11-18. I think this personal distance corresponds to the intellectual differences noted here.

[35] The idea that one should read the Classics for models for the present, bypassing the commentaries, is not at odds with the general ancient-style view. But in 1069, having seen the consequences of reading the Classics without the commentaries, Ssu-ma rejects this approach; see SMK 42.538-39.

patterns in one's own private life before applying them to the judgment of political actions in writing for the present. Thus the test of value is not whether the ideas of the sages form a coherent system, but whether or not the individual can practice what he learns. Wen has moral validity only if the author is promoting standards he has realized in his own conduct. The unity of word and deed is more important than the unity of literary coherence.

Ssu-ma's early essays take aim at what he regards as the merely "literary" attempt to promote certain values. "On the Ten Worthies," written in 1042, argues for the priority of "ethical conduct" (te hsing ) over the "affairs of government," "literary learning," and "speaking." Here Ssu-ma interprets ethical conduct as a firm commitment to the "full realization of the good," in contrast to seeking fame by relying on a mere "talent" (administration, writing, speaking) that might be used for any end. As he argues in "On Virtue and Talent," talent is the particular skill a man is born with, while virtue is the acquired knowledge of the good ends talent should serve. The claim that literary talent does not imply a knowledge of good ends was hardly exceptional, but Ssu-ma goes on to attack the idealizing of antiquity and sagely order as merely another form of literary talent, one that is particularly harmful because it claims knowledge of good ends. Because Chia I, for example, idealized antiquity, he failed to understand that establishing ritual and righteousness and securing the dynastic succession were the basis of political order; his advice thus threatened the survival of Han.^[36]

The danger of the clever persuader is that he leads others to think that rhetoric can be more vital to the state's survival than real achievements in national defense and administration. "Virtue" involves a commitment to doing the best one can under present circumstances rather than trying to be a "hero" but failing to accomplish anything. Warring States figures who failed to restore "ritual and righteousness," "perfect impartiality," "policy and instruction," and the "great Way" should not be praised for trying or blamed for failing, Ssu-ma argues. They should be judged by how they served their rulers, benefited their states, and nurtured their people. Grandiose plans entail great disturbance and risk, as the Han emperor Wu-ti's

[36] SMK 65.302; 64.797-98, and cf. 72.882; 65.806-7. The four categories, "ethical conduct," "affairs of government," "literary learning," and "speaking," are found in Analects XI.3. I have argued elsewhere that Sung literati used the first three categories (thus excluding speaking) to define the various fields of endeavor in which they made reputations and the responsibilities they claimed for themselves as the national elite (i.e., for government, perpetuation of the cultural tradition, and promotion of ethical values); see Bol, "Culture and the Way in Eleventh Century China" (Ph.D. diss., Princeton University, 1982), pp. 79-84.

search for immortals, expensive tastes, military campaigns, and oppressive taxation show. Ssu-ma objects also to those who use the model of the sages to argue that political leaders can fully control the direction of events. In "On the Spring and the Balance" (or "On the Timely and Irregular Use of Power") from 1045 he contends that the sages did take exceptional actions to redirect the course of events. But they did so to enable "state and family to be at peace and benevolence and righteousness to be established," while his contemporaries are merely advocating expediency and promoting themselves.^[37]

In Ssu-ma's view it is profoundly irresponsible to advocate policies to re-fashion the world according to an ideal model. First, doing so turns attention away from the immediate tasks of government; second, it fails to judge policy by its likely consequences. It is evident that Ssu-ma believes policies that greatly alter the government's role in society will threaten the survival of the state as a political unity.^[38] In 1045, after the defeat of the Ch'ing-li reformers, Ssu-ma's concern is not to explain his views but to find an alternative purpose for the literati, one that will lead them away from the model of antiquity. His options are limited. Antiquity and, to a considerable degree, the sages have been captured by men like Wang An-shih, whose vision of social-political integration offers literati in government a lofty purpose and a goal never achieved since the Chou dynasty. Ssu-ma finds his own issue. He begins to write about the historical cycle of order and chaos (chih luan ), and he begins to contend that when men understand why "order" failed to survive in the past they can act to preserve it in the present. Sung can accomplish what all previous ages had failed to do: the permanent preservation of political unity.

We can see this view emerging in a series of eighteen historical critiques. Ssu-ma makes his point, first, by discussing the unification and dissolution of political authority, in this case from the rise of Ch'in through the end of Han, and considers the values of "the superior man who brings order to the age." Second, he argues that political outcomes were determined by the choices of men with political authority, rulers and ministers. Their actions account for success and failure in establishing and maintaining the political order, yet these rulers and ministers were normal, even average, men. Third, he uses political practice rather than an idea of sagehood to define the content of traditional ethical terms. His attitude toward sages is

[37] SMK 65.805-7; 65.802-4; 66.825; 64.790.

[38] As we shall see below, later writings were to supply grounds for this conviction: the survival of political unity depends upon the government's ability to harmonize existing interests in a manner that makes it possible for the government to act to maintain itself and protect the state as a whole (i.e., a balanced structure of competing interests); policies that threaten existing interests substantially will deprive the government of their support, thus making it difficult for the state to rebalance and defend the structure.

ambivalent at this point. They should not serve as models of rulership, because sage rulers were the exception even in antiquity; hereditary succession was the norm. But if sages are to be taken as models, they should be seen merely as men who were "good at correcting their mistakes," rather than as men incapable of error. Virtues are not innate human qualities but names for calculated modes of political behavior. Thus the judgment that Ch'in failed through lack of good faith, benevolence, and wisdom should be taken to mean, Ssu-ma claims, that Ch'in failed by betraying the Ch'u king, massacring the surrendered army of Chao, and failing to understand how these acts would affect the attitudes toward Ch'in of the feudal lords and masses.^[39] Historically, Ssu-ma is claiming, the creation and survival of political unity has depended on how those with political authority chose to behave. It follows that literati can take the survival of the state as their common purpose. To achieve this end they must judge actions by their effect on political unity and seek out the principles that ought to guide political conduct. Literati in government can ensure political stability, Ssu-ma concludes in 1052, and political stability is the basis for ensuring the welfare of "all the people."^[40]

Beginning in 1050 Ssu-ma identifies one such principle of political unity. The institutions of government must be kept "public" (kung ). That is, they must always function in the interest of the survival of the state's political integrity, never in the "private" (ssu ) interest of individuals in government. He sees his own task in the early 1050s as the defense of the public character of the state's institutions.^[41] Understood correctly, he contends, such terms as righteousness, merit, and even talent refer to individual achievements that serve to maintain the government as the single public and impartial institution. "Being principled" (i ) does not mean aiding the helpless or dying in defense of others, but "clarifying the great roles of ruler and minister, understanding the great principles of the world, and defending them to the death without changing."^[42] Ensuring that government is a "public" institution requires that men serve their roles, not take advantage of them.

We should note that Ssu-ma is not arguing that the government should suppress or challenge private interests. He is arguing that there must be a clear distinction between the government, which must be kept fair and impartial, and private interests, which guide men outside of government.

[39] The specific points noted here are found in SMK 67.827 and 828. The eighteen critiques as a whole occupy SMK 67.827-32.

[40] This is the theme of Ssu-ma's question for the decree exam of 1052; see SMK 75.920-21.

[41] See for example his criticism of attempts to use honorary titles to show the emperor's personal favor in 1050-1051, SMK 18:275-80.

[42] SMK 67.832.

When individuals serve in government they enter the sphere of the "public" and must subordinate their personal ambitions to the demands of their offices.

Ssu-ma's essays from the mid-1050s restate a number of ancient-style-movement themes in his terms.^[43] "On Factions" contends that a "good" faction is one that defends the public quality of the state (as opposed to the party of idealists of Ou-yang Hsiu's famous essay from the 1040s). "Knowing Men" argues that the ruler's task is to know which men will maintain a unified hierarchy of political authority. "On Merit and Fame" urges the ruler to find men truly committed to the survival of the state (not men who have achieved fame in literati opinion) and to employ them fully and trust them completely, defending them against criticism, so that they will not be afraid to do what is necessary to preserve the state. At the same time, in a rare discussion of a favorite topic of "ancient-style" writers, the creation of civilization in antiquity, Ssu-ma tries to define the role of the government in society. The present is better than "high antiquity," he argues, though the people have not changed. It is better because the sages, seeing how needs and emotions had led men to exploit their environment to the point of scarcity and how scarcity led men to self-destructive competition, created civilization. They selected the wise to be rulers and superiors, they divided the land and forced men to respect boundaries, they defined social relationships, they established rites and music and government orders, they made clear the virtues, and they punished and fought those who did not obey. Thus the people were able to live in security and sufficiency. Government is responsible for maintaining this "teaching" (chiao ); indeed, the survival of the state as a unity that includes the people depends upon its doing so. But this also suggests the limit of government's interference in society: government defends the historically proven rules of social harmony and economic sufficiency but leaves the rest up to private interests.^[44]

Ssu-ma's willingness to claim the sages for his own cause suggests that he has turned from criticism of others to his own conclusions. His "Impractical Writings" of 1057 holds that the literatus who follows the "Way of the sages" will not try to change what he cannot control—his intelligence and courage, rank and wealth—but will simply try to do his best in his given role. By accepting his lot he will escape the sense of affliction that comes from unfulfilled desires, whether idealistic or selfish.^[45] For a literatus to be

[43] For a discussion of why kung is the most basic value, see SMK 69.856-57, "Chang Kung tzu Ta-ch'eng hsu" (1056). Ssu-ma marked his return to the capital in 1057 with several ancient-style essays. These can be read both as a defense of his patron, the former chief councillor P'ang Chi, and as support for the new regime of former Ch'ing-li reformers and ancient prose thinkers (but on his own grounds).

[44] SMK 64.793; 65.799-801; 65.787-90; 67.827.

[45] SMK 74:905-7, "Yü shu" reading through "T'ien jen" (1057-1058). My comments are an interpretation of the first ten sections. Ssu-ma Kuang does qualify this. On the one hand, he recognizes that the world is made predictable in large part because men continue to follow desire and are usually incapable of following good advice, if it requires a subordination of private to public. On the other hand, he points out that accepting one's role does not prevent him from continuing to learn, and learning leads to an ever-broader perspective (even to the point where one may realize his past has been a mistake).

a Ju he need only put aside desire for fame and profit and seek it.^[46] The answers need not be sought in the realm of heaven-and-earth. "The teaching of the sages," he writes, is "order man, not heaven; know man, not heaven."^[47] It is not necessary to seek guidance through introspection. Nor is it correct to choose particular texts as authoritative expressions of the "Way of antiquity" and promote them by imitating the "wen of antiquity." Indeed, in this 1057 letter, Ssu-ma declares that contrary to his correspondent's impression, he does not practice the "ancient style." Now that he knows the true Way he can devote himself to advancing it and defining it; he does not need to persuade men through literary argument.^[48]

Ssu-ma now was ready to present his views systematically. In about 1060 he decided to make the writing of a great chronological history of China from 403 B.C. to A.D. 959 his scholarly goal.^[49] And in office he began to propose policies to ensure that even a government staffed by men with private interests could maintain its public function.

Programs for the Present, 1058-1067

By the end of the 1050s Ssu-ma Kuang and Wang An-shih were ready to announce programs. Wang's "Ten-thousand Word Memorial" of 1058 was presented on his appointment as supervisor of funds in the Finance Commission (tu-chih p'an-kuan ), a post he held until his promotion to special drafting official (chih chih-kao ) in 1061. In 1063 Wang left to mourn his mother's death, declining appointments until late in 1067 when it became

[46] SMK 59.707.

[47] SMK 67.833.

[48] In this letter Ssu-ma argues that one can be sure he is on the right track in understanding the Way if he can "base it in heaven-and-earth, check it against the Former Kings, support it with Confucius, and verify it in the present." Ssu-ma may claim that his own ideas satisfy these four tests, but he does not appeal to them consistently (particularly not the first until the 1080s). Nor does a historical account of his thought suggest that he sought understanding from these sources in this manner. Nevertheless, the four tests indicate the standards that claims to universality at the time were expected to meet. On the rejection of the ancient style, see also SMK 59.712-13, letter to P'ang Chi. Ssu-ma's growing antipathy toward the idea that wen hsueh could function as a method of both understanding the Way and choosing officials is evident in SMK 60.718-19.

[49] The dating of his decision to write a history relies on Liu Shu's recollection; see Wang Te-i, "Ssu-ma Kuang te shih-hsueh," in Chi-nien Ssu-ma Kuang Wang An-shih , p. 32-33.

clear that the new emperor, Shen-tsung, actually wanted his advice. Ssu-ma Kuang began to spell out his program on his appointment as coadministrator of the Bureau of Policy Criticism (t'ung-chih chien-yuan ) in 1061. He served as a policy critic or censor until his appointment as a Hanlin academician in 1067.

Ssu-ma Kuang's Program

Ssu-ma had an unusually long tenure as a policy critic; he said he preferred a position that allowed him to "define right and wrong" to one that merely required literary skill.^[50] On taking up this post in 1061 he presented his program for ensuring the survival of political unity. He did this, first, through a series of memorials addressed to the emperor on political principles and specific policies and, second, by writing the Chronological Charts , his first major historical work, the forerunner of his Comprehensive Mirror .

"Five Guidelines," one of his first submissions, sets out a broad historical vision and the conclusions he draws from it. He begins with a review of the rise and fall of dynasties and concludes that dynastic houses have succeeded in maintaining the unity of the "world under heaven" for only five hundred out of the past seventeen hundred years because of the adventurism and negligence of rulers.^[51] Whether the Chinese world (Ssu-ma does not include the surrounding barbarians) remains unified depends on whether the "state" (kuo ) that unifies it survives. There is nothing natural about the rise and fall of states. The state is a man-made structure; it is possible to preserve it forever. Ssu-ma's analogy for the state is a building. The people are its foundation, rules and rituals its pillars, the high ministers its beams, the rest of officialdom its roof, the generals its walls, and the soldiers its latch. Rulers who "continue the structure and maintain the finished models of their ancestors" can pass it on to their descendants. But to do so they must keep this building in good repair. They must see to timely preparations against foreign invasion and natural disasters by selecting military and civil officials well, training the soldiers, storing up grain, and ensuring effective local administration. Selecting officials well requires noticing faults of character before they have a political effect. Finally, they must take measures to ensure that the functions of government are accomplished in substance, not merely in appearance.^[52]

Order exists, and the state as an integrated structure of functions sur-

[50] See Ssu-ma's nine refusals of his appointment as special drafting official in 1062, SMK 24.339-46. I have cited the sixth. See also his explanation for this behavior in his letter to P'ang Chi, SMK 59.712-13. For an account of why he accepted and tried to decline positions through his career to 1070, see SMK 43.555-56.

[51] SMK 21.307-14; 27.307-9.

[52] SMK 21:309-14.

vives, in this view, when all the groups that constitute the state are kept in their proper relation and fulfill their partial roles in the whole. The historical task of the ruler is to maintain the structure once it has been put together by the founder of the dynasty. That founder is the king with "heaven's decree," that is the one whose intelligence and strength outlast those of his rivals.^[53] Rulers may be seen as "owners" of the building, but Ssu-ma treats them more as caretakers responsible for its upkeep. If they have done their job poorly in the past, it is because they have lacked the judgment necessary to make political decisions and choices, as another memorial explains:

I observe that there are but three great virtues for the ruler: benevolence, knowledge, and militancy. Benevolence does not mean genial indulgence. Establish transformation through education. Improve administration. Nurture the folk. Benefit all things. This is the benevolence of the ruler. Knowledge does not mean petty spying. Understand the principles of the Way. Recognize security and danger. Distinguish the wise from the foolish. Discriminate between right and wrong. This is the knowledge of the ruler. Militancy does not mean violent ferocity. Choose that which agrees with the Way and do not doubt. Slander cannot confuse him. Flattery cannot move him. This is the militancy of the ruler. To be benevolent yet not know is like having good fields without ploughing them. To know yet not be militant is like finding the weeds around the shoots but not pulling them. To be militant yet not benevolent is like knowing how to harvest but not to plant. If these are all complete together, the state will be ordered and strong. If one is lacking, it will decline. If two are lacking, it will be in danger. And if none of the three is present, it will be lost. Ever since there were people this has never changed.^[54]

Ssu-ma repeated this to every ruler he served. Within the context of these general values of rulership, Ssu-ma identifies three essential tasks that the ruler must carry out to ensure that the parts of the structure fulfill their assigned functions. This Ssu-ma repeated to all his emperors as well.^[55]

I have heard that the way of achieving order depends on but three things: first, the assignment of offices; second, reliable rewards; and third, necessary punishments.^[56]

[53] SMK 21:307.

[54] Ssu-ma kept repeating these prescriptions for rulers throughout his career, with later memorials often citing earlier memorials. The memorial cited here is SMK 20.296, "Ch'en san te shang-tien cha-tzu" (1061). Others are SMK 24.346, "shang-tien hsieh kuan cha-tsu" (1062); 27.275, "Shang huang t'ai-hou shu" (1063); 27.383, "Shang huang-ti shu" (1063); 28.381, "Ch'i chien-sheng hsi-wu" (1063); 31.417, "Yen wei chih so hsien shang-tien cha-tzu" (1064); 32.427, "Ch'en chih yao shang-tien cha-tzu" (1064); 38.493, "Ch'u ch'i chung-ch'eng shang-tien cha-tzu" (1067); 46.568, "Chin hsiu-hsin chih-kuo chih yao cha-tzu" (1085).

[55] See the memorials listed in note 54.

[56] SMK 20.297-98.

The ruler's function is to see to it that all those responsible for the working of government fulfill their functions. This does not happen, however. Assignments and promotions are determined by mere longevity, and officials are shifted between offices before they learn their jobs, or when the difficulties their initiatives cause are still being felt but the good results are not yet apparent. Those who try are punished, while those who avoid problems are rewarded. The state cannot count on all officials to motivate themselves to fulfill their functions; few are able to be "concerned with the public and forget the private" irrespective of the situation.^[57] To make government effective requires institutional improvements: longer tenures, assignments by ability, and promotions by real merit. The first concern of policy is the effective administration of government.

When Ssu-ma wrote this he was fully aware that Sung was facing serious military and fiscal problems.^[58] We might well expect him to give priority to these pressing tasks of government. Why does he insist instead upon the importance of correct administration? The answer, I think, is evident in his analogy between the state and a building. The state is a structure of groups with different kinds of power and responsibility. (Note that the "people" are the foundation upon which the rest of the building-state rises.) Maintaining these groups in proper relation is the same as maintaining the state. The parts of the whole most subject to collapse are those with the most power; they are also fewest. If one part gets too much power, or gets the wrong kind, or fails to meet its responsibility, it will affect the work of the other parts. Thus, Ssu-ma concludes, make sure that the men assigned to each role can do the job, and use rewards and punishments to see that they continue to do so. The process of government, rather than the actual work of government, holds Ssu-ma's interest. This explains, I suspect, why Ssu-ma has a limited interest in the lower end of officialdom, in local government, where he supposes that if men are given the chance to develop competence and are rewarded for it, the tasks of government will be carried out. He is primarily interested in leaders of state as men directly responsible for making the government function.

Behind all this lies an assumption that men are generally guided by partiality and self-interest. Ssu-ma appeals even to the ruler's own partial interest: if he desires to keep the state for his descendants, he must have a total commitment to the "public." Institutions can be reformed, but whether officials do their duty depends upon the rewards and punishments that remind them where their own interests lie. This is not ideal, but, Ssu-ma points out, political authority must take the realities of "custom" (i.e., social values) into account. He does argue, at considerable length, that the

[57] SMK 21.312.

[58] SMK 20.298.

government must try to influence social values in its own interest. That is, it should make men "accustomed" once again to "the roles of superior and inferior." "Be Careful about Habits" recognizes that a willingness to accept hierarchy is a state of mind and that men must believe in its importance or come to accept it as a given.^[59] Ssu-ma recognizes, of course, that it is not a given; it is necessary to help contain the problem of partial and selfish interests. However, once those with authority understand this, they can act so as to "make the people accustomed to the roles of superior and inferior."^[60]

The argument of this memorial, developed through a historical review of political change, is essentially this: the survival of a state is a function of its success in getting men accustomed to hierarchical relations of authority. This "custom" is the safety net of the polity, for the human resistance to change (which Ssu-ma takes as a given) entails a willingness to follow established authority as long as it does not itself involve costly change. In antiquity this habit was sustained by ritual, but with the decline of the Chou it gradually disappeared until only the powerless Chou king survived to testify to its value. During the Han and T'ang not only was ritual not restored, but men became accustomed to challenging superiors. By the end of T'ang men no longer spoke of "the ranking of honored and humble or the principles of right and wrong," and the ephemeral states of the Five Dynasties period were the result. The Sung founders "understood that all misfortune arises from an absence of ritual" and took measures to establish the authority of the ruler and reduce that of the provincial governors. They unified the hierarchy of authority so that it extended from the court to fiscal intendants (Ssu-ma's modern equivalent for the feudal lords of Chou) and thence to local officials and the people; "thus the ranking of superior and inferior was correct and rules and principles were established."^[61]

As Ssu-ma would later assert in the Comprehensive Mirror , a hierarchy of authority and clearly defined levels and areas of functional responsibility were basic to order and unity. These concerns dominate Ssu-ma's memorials as a policy critic. We find typical examples in his opposition to the regular granting of amnesties, on the grounds that they make rewards and punishments ineffective; his call for changes in the examination system, on the grounds that exams that "value literary writing highly" promote only one kind of talent; his proposal for a reform of the promotion system, to be undertaken to guarantee that only the competent rise; and his criticism of particular cases of rewards for men he finds lacking true merit.^[62]

[59] SMK 24.347-51.

[60] SMK 24.349.

[61] SMK 24:348-49.

[62] These examples are taken only from 1061 but are representative of the 1061-1065 memorials in SMK , ch. 20-36. On amnesties, see SMK 20:300; on exams, 20:302; on the promotion system, 21:314; for an example of a particular case, see his attack on Chang Fang-p'ing, 23:325.

When Ssu-ma does turn to national defense and state finances, he again argues for bureaucratic and administrative solutions. Writing during a period of increasing tensions with the Tangut's state of Hsia, he argues against all attempts to increase the size of the military, conscript northern farmers, and establish a more aggressive posture on the borders. His aim is to keep the two parties, Sung and Hsia, in a stable relationship, one in which neither will disturb the other. But he also notes that a large military depletes the treasury, thus making the military itself a threat to the state. His advice is relatively simple. Better training and better selection of the officer corps will make it possible to return the military to its proper size; and he cites the achievements of the Sung founder, who had far fewer troops.^[63]

"On Wealth and Profit," his most extensive memorial on fiscal affairs from this period, takes a similar view. He opposes "literati of literary talent" whom he sees dominating the Finance Commission and proposing reforms aimed at increasing the state's share of the national wealth. The restoration of fiscal stability is to be achieved by three kinds of measures. First, improve the financial administration: fill financial posts with specialists in financial affairs and create a special career path of financial offices, separate from the career path of "literary talent." Second, bring the common people back to farming, their proper role, and thereby restore production and revenue. He advocates reducing farmers' taxes by tapping urban wealth and hiring ya-ch'ien servicemen. But since he sees poor administration as the main problem, he pays most attention to bureaucratic reforms that will ensure competence in local government. Third, he calls for a reduction in expenses. The stipends of the imperial clan and gifts to officialdom can be reduced. But the real problem is that the numbers in civil and military service are growing while "the production of heaven-and-earth is constant." Thus instead of discussing ways of increasing production and revenues to cover costs, Ssu-ma concludes that the state must manage its revenue better to make ends meet while reducing expenses. He calls for the unification of financial controls in a single organ under the chief councillor, with the responsibility of matching expenses with revenue, finding the causes of deficits and possible savings, and evaluating the financial bureaucracy.^[64]

Ssu-ma was well aware that his policy advice was not being taken.

[63] Ssu-ma's discussions of military policy became more specific as tensions with the Tan-guts increased during the 1060s; see SMK 20:298; 34:449-59; 35:461; 35:464.

[64] SMK 25:356; 25:361; 35:353-62. Note that Ssu-ma would bring the Privy Purse into this unified fiscal system as well.

Moreover, he saw that rulers, in particular Ying-tsung, did not really believe that "to order the self nothing is prior to filial piety; to order the state nothing is prior to impartiality (kung )."^[65] To demonstrate the necessity of his views Ssu-ma turned to scholarship.

Learning from History . Long "fond of historical learning (shih hsueh )," Ssu-ma in the 1060s turned to historical writing in earnest. In 1064 he submitted to Ying-tsung the Chronological Charts (Li-nien t'u ), a chronology of events from 403 B.C. to A.D. 959, which was to be incorporated in a longer work, Record of Examinations into the Past (Chi ku lu ), in the 1080s.^[66] And in 1066 he submitted an eight-chapter work entitled Comprehensive Treatise , covering the years 403 to 207 B.C. The Charts appears to have become the outline for the Comprehensive Mirror ; the Treatise became its first eight chapters.

In the Mirror Ssu-ma would claim that learning from history was the "single starting point" (i tuan ) for knowing the Way to achieve order through government.

The Changes says, "The superior man acquaints himself with many sayings of antiquity and many deeds of the past in order to increase his virtue." Confucius said, "It is enough that the language one uses gets the point across." Thus history is the single starting point for Ju, wen is a superfluous affair for Ju. As for Lao and Chuang, their "void" and "nothingness" certainly are not that with which we give instruction. Learning is that by which we seek the Way. There are not two Ways in the world. How could there be four learnings!^[67]

Ssu-ma was sure that history revealed the principles or rules (Ssu-ma uses such terms as kang-yao and chi-kang ) necessary for establishing and maintaining the state. As he asserted in the Chronological Charts , "From when the people first came into being to the end of heaven and earth, those who have had states, in spite of myriad sorts of change and transformation, have not gone beyond [the rules set forth here]."^[68] Ssu-ma freely admitted that

[65] SMK 31:417. For an example of Ssu-ma's view of Ying-tsung, see 36:471-75.

[66] For an account of this work, which briefly covers the period before 403 and Sung through 1066, see Wang Ching-chih, "Chi ku lu chien t'an," Chung-hua wen shih lun-ts'ung , vol. 14 (Shanghai: Shanghai ku-chi, 1980), pp. 121-31. I have followed the text of the Charts Ssu-ma reproduced in his Chi ku lu of 1086 (Hsueh chin t'ai yuan ed.), 1 l:58b-16:79b.

[67] Quoted in Lin Jui-han, "Ssu-ma Kuang ti shih-hsueh yü cheng-shu," reprinted in Sung shih yen-chiu chi , vol. 8 (Taipei, 1974), pp. 59-60. The passage is found in Ssu-ma's Tzu-chih t'ung-chien (Peking: Ku-chi ch'u-pan-she, 1956) 123:3868. Ssu-ma is here commenting on a passage relating the simultaneous establishment of "schools" of Neo-Taoism (hsuan-hsueh ), Ju, History, and Wen under the Liu Sung dynasty in the fifth century. Ssu-ma cites the I Ching , Ta-ch'u hexagram, trans. Wilhelm and Baynes, 516, modified, and the Analects , 15:41, trans. Lau, p. 137.

[68] Chi ku lu 16:78a.

he had chosen those events that he saw as essential to tracing the rise and fall of states. But he held that the principles he derived from these were valid for the present because they could account for both the success and the failure of past political leaders to establish and maintain order. A single set of rules made history consistent and coherent: "The Way of order and chaos is on a single thread in past and present."^[69] Thus his history becomes a record of political conduct rather than of cumulative institutional or cultural change. He is after timeless principles, valid whether or not rulers followed them consciously. But they were principles that men should obviously try to follow.

What then is the content of these principles? The view that Ssu-ma was a monarchical absolutist and defender of autocracy, although still current, has been effectively challenged by Anthony Sariti, who argues that for Ssu-ma, "the emperor was confined in his own stratum, namely the bureaucracy," that Ssu-ma believed in the necessity of remonstrance, and that the ruler was to subordinate himself to "Confucian principles." While it is certainly true that Ssu-ma limits the ruler, demanding that he adhere to the lessons of history, Ssu-ma also makes the ruler bear full responsibility for the survival of the state. Precisely because government functions through a hierarchy of authority, the conduct of the ruler, more than that of any other position, affects the working of the entire government. Ssu-ma defines the functioning of the ruler according to principles that apply to the administration as a whole. No doubt these principles are more Confucian than Taoist, but we should not assume that there was a given set of clearly defined and well-established Confucian principles to which Ssu-ma and other officials could appeal. One of the purposes of Ssu-ma's historiography was to define what the principles of the Ju were.

The introduction to the Chronological Charts gives the essential rules that determine order and chaos, placing the ruler at the center:

The Way of order and chaos is on a single thread in past and present. The duration [of a state] is a matter of virtue alone. I am ignorant and of shallow learning, incapable of understanding the larger form of the state. But whenever I have divined it with the traces of previous ages as carried in the records I have boldly dared pronounce upon it. Now whether the state is in order or chaos depends entirely upon the ruler of men. The Way of the ruler of men is one. His virtues are three. His talents are five.^[70]

[69] Chi ku lu 16:75b, written in 1064. Compare, in 1066: "The sources of order and chaos have the same necessary form in past and present" (ku chin t'ung ti ): SMK 17:254-55.

[70] Chi ku lu 16:75a-75b. The Chronological Charts begins in 403 B.C. , as does the Comprehensive Mirror , because the division of Chin into three states shows that "the rules of ritual were exhausted." See Chi ku lu 11:58b. This judgment is explained and elaborated in the introductory comment in the Comprehensive Mirror ; see Tzu-chih t'ung-chien 1:2-6. Both works end in 959 A.D. with the demise of Later Chou, which Ssu-ma treats as having provided the foundation without which Sung could not have unified the world. The point here, I take it, is that Sung cannot truly distance itself from immediate pre-Sung history and simply return to antiquity. See Chi ku lu 15:74, Tzu-chih t'ung-chien 294:9599-9600. The twenty-five charts are now lost, but the year-by-year record of events has survived together with thirty-six comments interjected by Ssu-ma. In contrast to the comments found in the Mirror , Ssu-ma's comments here are largely devoted to evaluating rulers through history. For example, in the evaluations of Eastern and Western Han (12:87b-88b, 13B:103a-104a) and T'ang (15:61b-65b), every ruler is discussed, while ministers are introduced only in relation to the rulers' conduct of affairs. Although he wrote the work for Ying-tsung, it gained a broader readership with the publication of an unauthorized edition. This edition appeared under the title The Line of Emperors (Ti t'ung ), a title that Ssu-ma claimed "was not my intent." See SMK 71:877, "Chi Li-nien t'u hou," on the format of the original work and the problems with the unauthorized edition. The introduction presents the rules in general terms, while the various comments explain how particular rulers variously succeeded or failed to realize those principles.

The one Way is the correct way to employ men; the three virtues, as before, are benevolence, knowledge, and militance; and the five talents are styles of rulership. In elaborating on the one Way Ssu-ma repeats much of what he has said before but also makes the new claim that it is the shih that the ruler must employ. They exist in every state as the men of either great wisdom or ability. They must be employed because the people follow them, as the leaves and branches follow the "root." To gain the support of the people the ruler must gain the support of the shih . Once he has them he must choose among them, assign them functions according to their real · talent and achievements, and give them full authority to carry out their tasks. He gives them rank and salary, encourages them with rewards, and punishes those who stray. In this view the shih , the elite, exist already in society. They are not created by the state, but they are essential to the state, which, Ssu-ma appears to imply, must make those members of the elite in government do its bidding yet not go so far as to threaten the leading position of the elite among the people.

The three virtues concern how the ruler employs men. The first two need no further explanation; the third, militance, concerns the ruler's ability to stand behind those he has chosen. The five talents are indicative of the ruler's success. There are founders, able to unify the world, and four classes of successors: those who keep the system in good repair, reforming those received institutions that have developed faults; those who do not pay attention to the historical situation; those who do and so "restore" the state; and those who are so utterly unconcerned that the state is lost. The point here is constant attention to the direction of events and an ability to see that a deterioration in the structure of relations presages trouble. The Way, virtues, and talents hold through all historical change, Ssu-ma concludes, yet an ordered world has been a rare achievement. The first two Sung rulers were true founders. Implicitly, readers are invited to ask what kind of successor the current ruler will choose to be.

Thus by the end of Ying-tsung's reign Ssu-ma had articulated a body of principles and a general program. When Shen-tsung came to the throne in 1067, Ssu-ma was ready to provide advice. Instead, the emperor turned to Wang An-shih and proceeded to contravene the rules of government Ssu-ma had defined for him.

Wang An-shih's Program

Wang An-shih set forth his program for achieving political-social integration in his famous "Ten-thousand Word Memorial" of 1058, repeating its essential points in 1060 and 1061.^[71] The distance between Wang and Ssu-ma is clear. All that Wang requires of the ruler is that he support the program. Nor, for him, are the shih already present in society and waiting only to be correctly employed: Wang focuses on forming the kind of shih that will be suitable for carrying out his program.

Wang's memorial follows directly from his concerns in the 1040s. We can read it as an answer to one of his own examination questions:

There were root and branch in the sages' ordering of the age. There were what came first and what came last in their applying [of root and branch]. Today the problems of the world have been left uncorrected for a long time, and teaching and policy have not yet been put in accord with the intentions of the sages. We have lost sight of the root, seeking it in the branch; we have taken what should come last and put it first. Thus the world careens toward disorder. Now if it is so that the world will not be ordered except through the means the sages used to achieve order, then to be considered a true shih one must attend to how the sages achieved order. I want you gentlemen to relate in full the root and branch of how the sages achieved order and what they put first and last.^[72]

The memorial also responds to two objections, raised by the historically minded, to a policy of returning to antiquity. First, the policies of the sages addressed their times; times have changed and so must policy. Second, if there was an integrated order in antiquity and sages governed, then men seeking to cope with an age of decline clearly should not adopt policies suited only for an age of perfect order.

Wang begins the memorial by attributing contemporary problems (domestic and foreign, moral and financial) to a failure "to understand institutions" (fa tu ). By this he means understanding that to secure order institutions must "agree with the policies of the Former Kings." But more precisely: "imitating the policies of the Former Kings means we should only

[71] LCC 39:410-23. For Wang's restatements, see LCC 41:438 and 39.423. The memorial is fully translated in H. R. Williamson, Wang An-shih: A Chinese Statesman and Educationalist of the Sung Dynasty (London: Probsthain, 1935), vol. 1, pp. 48-84.

[72] LCC 70.747, third question.

imitate their intentions." Moreover, these intentions, which ground a systematic program of action, stood the test of historical change in antiquity.

The two emperors [Yao and Shun] and the three kings [Yü, T'ang, and Wen] were removed from each other by over a thousand years. Order and chaos followed one upon the other and periods of splendor and decline were fully present. The changes they encountered and the situations they faced differed, and the measures they adopted varied as well. But their intentions in making society and state (wei t'ien-hsia kuo-chia chih i ), the root and branch and the first and last, were always the same. I therefore say: we ought simply to imitate their intentions. If we imitate their intentions, then whatever changes and reforms we make will not shock people or cause complaint, yet will surely be in agreement with the policies of the Former Kings.^[73]

In short, policies that accord with the intentions of the sages, and are systematically organized to correspond to—not to replicate—the sages' program of action, can be put into effect without fear of the consequences.

But, Wang continues, "to fit the intentions of the Former Kings to the changed situation of the present" is not possible as long as "human talent in society (t'ien-hsia ) is inadequate." Talent is the "root" of an integrated order; the government must set about "completing it by molding and casting it." The government creates the talent it needs through a sequence of measures founded in the model of antiquity.

The first requirement is "Instruction" through government schools, established to train all those capable of being of use in realizing the program. Schooling should immerse men in the integrated systems of antiquity. They will learn "the affairs of rites and music, punishment and policy" by living in an environment where all that they see and practice are the "model sayings, moral conduct, and intentions for ordering society of the Former Kings." Second is "Nurturing." The government nurtures all the people by providing economic support, establishing rituals of passage and of daily life appropriate to their economic stations, and finally by controlling them with penal law, thus to "unify social customs and bring about order." Third, through "Selection," the schools recommend the most wise and able to the leaders who may, on examining their speech and conduct, assign them probationary employment and titles. Finally, in "Employment," those selected and proven are assigned ranks and responsibilities commensurate with their talents. Such officials are to be given long tenures and left unfettered by regulations so that they can develop and complete projects and do what ought to be done.

Wang goes on to compare the present with the model. "Instruction" is incomplete, for it deals only with the civil and literary and ignores the military. Salaries are too low to "nurture" honesty among officials, while rites

[73] LCC 39.410-11.

fail to restrain men, and the law does not punish basic faults. "Selection" is based on literary skill and memorization; it fails to garner men of real use. "Employment" is determined by a seniority system that assigns men to positions outside their competence. Wang concludes that the first step now is to devise policies to form talent under present conditions but in a manner in line with the intentions of the Former Kings. To accomplish this there is a necessary sequence: first think out broad strategies, then make precise calculations, then gradually put them into effect, and finally bring them to fruition. Offer rewards to those who further the cause and punish those who hinder it. Let the ruler take the intentions of the sage kings as his guide.

The "Ten-thousand Word Memorial" as I read it, is about the idea of a perfect, self-contained, and self-perpetuating system. The forming of talents is not separable from policies to provide for the livelihood, morality, and discipline of society. Producing men devoted to the system requires establishing the system at the same time. However, when the memorial turns to actual measures appropriate to the present, its aims are more limited. Rather than proposing a school system for all the people so that officials can be chosen from among them, Wang asks only that, as an initial step, the government restructure its relationship with the shih . In effect, Wang calls for integrating the shih with the government, rather than allowing the shih to exist independently of the government and its purpose. But we may also say, and perhaps just as accurately, that Wang is asking that the government be made one with the shih and with their aspirations as Wang presents them. In either case Wang's program demands the unification of politics and morality, of government and society. This is clear in his claim that education must "unify all who learn" to prevent disagreement among literati about what is to be done. It is clear too in his view of the economy: he holds it essential that the government "manage wealth" in order to keep private wealth from increasing to the point that people become beholden to private interests. That would irreversibly divide authority and so hinder the uniform implementation of policy.^[74]

For Wang political leadership belongs to those who have truly "learned," men committed to realizing the intentions of the Former Kings. "Even the son of heaven, facing north [thus acknowledging such men as teachers], will inquire of them and take turns acting as host and guest with them." In letters to followers in the 1060s, Wang insists that achieving an integrated system can arise only when men who have truly learned gain authority. Such men will be able to direct all the parts to fulfill their proper roles as pieces of a larger whole; "We cause others to take what is correct

[74] LCC 82.860-61; 82.862-63. Other accounts of the ideal school include 82.858 and 83.870.

from us so that they are able to be correct." Once they have the knowledge in them, all they need is the political authority to put it into effect.^[75]

Learning from the Sages . Wang An-shih's essays show that he sought to justify his claim that literati ought to take the sages as their guide to political action. The model for the present must be the "completed model" (ch'eng-fa ) developed by the sages when they took over what heaven-and-earth had begun, "completing" the things nature had brought into being. (Contrast this with Ssu-ma Kuang's view that the truly relevant models are those of the dynasty's founders.) "Thus in the past, when the sages were in power and took all things as their responsibility, they necessarily instituted the four methods. The four methods are ritual, music, punishment, and policy; these are the means for completing things."

The sages' completed model, which was free of any onesidedness and partiality, was established over time in a cumulative manner. Later sages did not imitate the "traces" of former sages, Wang explains, because sages responded "according to changes in the times" (ch'üan shih chih pien ). Nevertheless, there was a basic uniformity to their intentions. Confucius, the last sage and completer of the system, responded to changed times according to the "intent of ritual"; he did not stick to "the forms of ritual [already] instituted." This meant, in the case of ritual, anchoring it in the communal feeling of man as it was at that moment while incorporating the sages' own tradition. Wang extends this to all institutions. A system must be built from the common desires of those affected, and it must be instituted in a way that allows the "average man" (chung jen ) to meet its requirements. Imitating the forms and imitating the intentions are not mutually exclusive, however. Those who successfully thought out and established integrated systems inferred intentions from the forms of the past. Thus they were better able to accommodate change and maintain an integrated system than were mere imitators of past forms. The latter, trying to preserve forms, were increasingly unable to incorporate new developments. True sages were able to create new systems on the basis of past intentions. Their willingness to make changes in forms to realize the same idea was a sign of their sincere desire to benefit all. Imitators of forms lacked that self-confidence; they copied the past to impress others but did not know the intentions on which past models were based. This accounts for the decline of Chou: men started imitating forms and forgot how to act as sages.^[76] Having the intent as opposed to imitating the form was what made kings and hegemons fun-

[75] LCC 82.858; 72.766-768. See also LCC 75.790-91: "In regard to learning, the superior man's intention is surely directed to society, but first to ourselves and then to our fellows. If we ourselves are in order, then whether others are in order or not depends on whether we realize our intention or not."

[76] LCC 69.730; 67.713; 66.701; 67.717; 67.714.

damentally different. Ssu-ma Kuang did not share this Mencian view. Ssu-ma saw only a difference of degree, since political order under any circumstances depended upon the same principles: if order was achieved by hegemons, they must have been following the same principles as kings.^[77]

For Wang antiquity has authority in three ways. On the first level, he can justify the creation of institutions on the grounds that they formed part of the "completed model" provided by the ancients. Here it is a matter of "imitating the traces." On a second level, antiquity provides a set of ideas for organizing society and state into a single system. These ideas were constant, even though the institutional structure grew and changed. Third, antiquity provides, in the sages, a model for carrying out the reorganization of society and state. One who shares the intentions known from the forms of the past can do what is necessary to reestablish a coherent, unified system. Antiquity thus has authority as a source of models, purposes, and methods. These three aspects join in the term fa (model, rule, method, law, and to imitate). Knowing the fa of the sages is more important than historical knowledge, for their fa were effective in achieving true order. In what might well be seen as a response to Ssu-ma Kuang, Wang writes, "I hold that one who understands order and chaos ought to discuss the methods [the sages] used to transform men."^[78] Clearly Wang and Ssu-ma had different ideas about how the sages had civilized men.

Because Wang is sure that there are models, intentions, and methods, he can make claims about how sages in antiquity must have operated, even when evidence is lacking. This is illustrated in "The Duke of Chou," where he defends the idea of a state school system to form shih against those who think (like Hsun Tzu in this essay, or Ssu-ma Kuang) that the sages relied on a system of personal connections and recommendations to find shih . He appeals first to the "system of the Three Eras," a time when "fa was truly perfected," which is said to have included schools. But then he argues that this manner of forming shih is necessary for fa to be good. "If the Duke of Chou knew how to govern, then by rights he would have established the fa of schools under heaven."^[79]

This suggests that for Wang the real test of whether he is right is the ability of his own mind to envision integrated systems. To deal with the present he must, of course, fit in developments since antiquity.^[80] Essays

[77] SMK 73.896. For a discussion of Sung literati debates on this issue, see Hoyt Cleveland Tillman, Utilitarian Confucianism: Ch'en Liang's Challenge to Chu Hsi (Cambridge: Council on East Asian Studies, Harvard University, 1982), pp. 46-53.

[78] LCC 69.731.

[79] LCC 64.677-78.

[80] He admits this when he writes that in learning men must "comprehend past and present, acquaint [themselves] with rites and rules, the patterns of heaven and the affairs of man, and the changes in policy and instruction." See LCC 69.734.

such as "On Ritual and Music," where ritual and music are presented as models for true "learning," and "On Attaining and Using Unity," which defines the process of "learning," suggest that the process of mentally envisioning how various parts can be formed into a coherently integrated system is basic to Wang's view of learning. These are very difficult essays, in part because we find Wang trying to unite internal and external, human nature and sentiment, man and heaven-and-earth, sages and common men. If they are not persuasive as philosophical treatises, they still indicate what Wang was trying to accomplish. He begins the second essay with the statement "All things have their perfect patterns (li ). If one can grasp the essential of the patterns, one is a sage. The way to grasp the essential of the patterns lies simply in attaining [and using] their unity." This is applied to human affairs in the first essay when Wang concludes that all enduring creations "have been established by sages who attained [and used] the essential and loved learning." For Wang attaining and using the essential or unity enables one to hold all dualities together and create things that harmonize and synthesize diverse interests and traditions. By contrasting this with merely "maintaining completed models" he argues that this kind of learning can and must be a basis for reorganizing modern society.^[81]

There is a dialectic between Wang's attempt to achieve a coherent, unified understanding of the Classics and the sages and his efforts to envision a system that can incorporate all aspects of the world. We can see a similar dialectic, in Ssu-ma Kuang's thinking, between the effort to define the constant principles of political process and the attempt to make history coherent. To a greater degree than Ssu-ma, I think, Wang appeals to the coherence of his conclusions to justify their validity. He will later write his cousin Tseng Kung:

For long the world has not seen the complete Classics. If one were only to read the Classics, this would not be enough to [let one] know the Classics. I thus read everything, from the hundred schools and various masters to [such medical texts as] the Nan ching and Su wen , the pharmacopeia, and various minor theories, and I inquire of everyone, down to the farmer and the crafts-woman. Only then am I able to know the great system (ta t'i ) of the Classics and to be free of doubt. The later ages in which we learn are different from the time of the Former Kings. We must do this if we are fully to know the sages.^[82]

[81] LCC 66.707; 66.702-6; quoting from 66.706. Winston Lo, in "Wang An-shih and the Confucian Ideal of Inner Sageliness," and K'o Ch'ang-i, in his account of Wang's philosophical thought (pp. 194-96), both treat "On Attaining and Using Unity" as a statement of Wang's basic philosophy. I suspect that this essay and "On Ritual and Music" are late writings because they are concerned with the internal and the cosmological. Hsia Chang-pu has argued that after 1068 Wang changed his view of human nature to argue that good values were innate; see "Wang An-shih ssu-hsiang yü Meng-tzu," p. 315.

[82] LCC 73.779.

Knowing the sages and the Classics means envisioning a perfect system that comprehends and integrates everything. Wang's claim, I think, is that if he has envisioned that system he knows the sages and he can act.

Advancing and Withdrawing, 1067-1086

Wang's rise to power began with his appointment to the Council of State early in 1069; at the end of 1070 he became chief councillor. He left central government for good in 1076 and soon retired. Ssu-ma left central government in the fall of 1070. Until 1085, when he was appointed to the Council of State after the death of Emperor Shen-tsung, he lived mainly in Lo-yang, working on the Comprehensive Mirror . He died in 1086 in office as chief councillor, surviving Wang by half a year.^[83] Once Wang's "New Policies" (hsin fa ), as they were soon known, were put into effect, Ssu-ma set forth his objections. When Ssu-ma gained power he fought to rescind the New Policies in their entirety.

The likelihood of disagreement between Ssu-ma and Wang had already become apparent in 1061, when Ssu-ma proposed a reform of the promotion system to ensure the effective bureaucratic functioning of government. He envisioned a twelve-grade system of assignments (ch'ai-ch'ien ), from chief councillor to sheriff and registrar, and new rules for determining assignments, tenures, and promotions. This was a reform close to Ssu-ma's heart, an essential part of his larger scheme. Called upon to review the proposal, Wang responded by dismissing it as a "petty reform, in the end offering no solutions to real problems and not worth undertaking." Improving the functioning of government missed the point, Wang held, unless it was accompanied by a complete reexamination of the system. "If the court is committed to greatly perfecting institutions and molding and ranking human talent," it should call upon all "literati with proposals" to debate.^[84]

When the two men met again, in 1068, they again disagreed. But the new emperor was interested in "accomplishing something," in spite of the growing deficit, and he asked for advice. Ssu-ma called for reducing ex-

[83] In 1068/4 Wang was summoned to the capital as Han-lin academician, in 1069/2 he was made vice-councillor, in 1070/12 he became one of two chief councillors, between 1072/3 and 1074/4 he was sole chief councillor, and, after having been briefly forced out, served again as chief councillor from 1075/2 to 1076/10 (alone from 1075/8). He then went to a prefectural post in Chiang-ning. He retired fully in 1077/6, and remained in Chiang-ning until his death in 1086/4. In 1068 Ssu-ma was Han-lin academician and reader-in-waiting, in 1069/2 and 1070/8 he requested outside appointments, in 1070/9 he took a provincial post in Yung-hsing but left that for a censorial position in Loyang in 1071/4. From 1073 he held a sinecure in Loyang. In 1084 he began to become active again, and in 1085/4 he was appointed to a prefectship. But in 1085/5 he was appointed vice-councillor, and then, 1086/interc.2, he was made a chief councillor. He died in office in 1086/9.

[84] SMK 21.314-16; LCC 62.667.

penses (although he declined appointment to a committee to propose cuts) and suggested that the traditional gifts to high officials not be given at the sacrifices this year.^[85] This led to a debate between Wang and Ssu-ma before the emperor. Wang responded that Ssu-ma's proposed measures were a meaningless exercise that would do nothing to "enrich the state [kuo ]." His following comment was surprising: "Moreover, the inadequacy of revenue is not the priority of the present moment." Ssu-ma rejoined:

"Since the end of Chen-tsung's reign the state's revenues have been inadequate. How can you say this is not a priority?"

Wang replied, "Revenue is inadequate because we still lack men good at managing wealth (li ts'ai )."

Ssu-ma responded, "Men good at managing wealth merely tax, thus exhausting the wealth of the people. As a result the common folk are impoverished, they flee and become bandits; how does that benefit the state?"

Wang : "This is not what being good at managing wealth means. When one is good at managing wealth, revenues will be more than adequate without increasing taxes on the people."

Ssu-ma : "Those are the words with which Sang Hung-yang deceived Emperor Wu [of Han]; Ssu-ma Ch'ien recorded them to criticize Emperor Wu's obtuseness. The wealth and products heaven-and-earth produce are of a fixed amount. What is not with the people is with the government. If Sang Hung-yang was able to increase revenues, then, if he did not get it from the people, where did he get it?"^[86]

In the end, Wang An-shih's definition of the priorities of the day—man-aging wealth, changing customs, and establishing institutions—prevailed.^[87]

Debate over the New Policies

When Wang came to court in 1068, he immediately called for action on education, the civil service, agriculture, the military, and finances.^[88] Seen in isolation, the policies that he and his allies eventually proposed addressed specific problems, some internal to the working of government itself, others relating to the demands government made on society in order to maintain its tax base, its supply of manpower for local service and national defense, and its pool of potential officials. In many instances, the policies either had been proposed during earlier reigns or had been developed by creative local officials. And while Wang consistently cited ancient models to justify the institutions his policies created, he was not averse to citing Han and T'ang

[85] SMK 42.533-34 and 42.535-37.

[86] SMK 42.543-45. I am not sure this debate actually took place (the text is out of place in Ssu-ma's collection), but it does accurately reflect the positions of both men.

[87] HCP: shih-i 4.3b-4a (1070/2) and 5.19 (1070/9).

[88] LCC 41.444.

precedents. But the New Policies were not intended to be discrete remedial measures. They were to work together to change the way government operated, change its relation to society, and transform society itself. Both Wang and Ssu-ma understood that the program was a vehicle for larger ends. The following chronology traces the institution of many of the New Policies.^[89]

1069, second month: Finance Planning Commission (Chih-chih san-ssu t'iao-li ssu). Codirected by Wang; absorbed into the Secretariat in 1070, fifth month. This office drew up the plans for the key policies.

1069, fourth month: Investigating Commissioners . Commissioners appointed to investigate local conditions in agriculture, irrigation, and obligatory local service and recommend action.

1069, sixth month: Administrative Regulations Commission (Pien-hsiu chungshu t'iao ssu). Abolished in 1075, tenth month. This office planned the restructuring of the bureaucracy completed in 1082. It resulted in placing the Council of State in charge of organs for financial, administrative, and military planning and operation, where hitherto the leading organs in these three areas had reported directly to the emperor.

1069, seventh month: Equitable Transport and [Price] Equalization Policy (Chün-shu p'ing-chun fa). Fiscal intendants for six southern circuits were given the authority and capital to supply government requirements by buying and selling according to market conditions, rather than relying on fixed local quotas and transport obligations.

1069, ninth month: Green Shoots Policy (Ch'ing miao fa). Using the Ever-Normal Granary (Ch'ang-p'ing tsang ) reserves as capital; led by Ever-Normal Granary commissioners in each circuit. Also known as the Ever-Normal [Granary] Policy (Ch'ang-p'ing fa ). Provided loans to farmers, and eventually to urban dwellers, at 20 percent interest (plus surcharges). Loan amounts were determined by household grade.

1069, eleventh month: Regulations on Agriculture and Water Conservancy (Nungt'ien shui-li t'iao-yueh). Based on reports from the investigating commissioners. Set out rules for recovering fallow land, carrying out local irrigation projects, and undertaking river conservancy.

1070, twelfth month: Tithing Policy (Pao-chia fa). Begun in the capital region; organized households into units of ten, fifty and five hundred. Each household with two or more adult males was to supply one as a militiaman. Militiamen received training, patrolled, and caught bandits for rewards. The basic unit was responsible for mutual surveil-

[89] For an overview of most of the New Policies, see Liu, Reform in Sung China , pp. 4-7. The most extensive account in English is still Williamson, Wang An-shih , pp. 131-346. Liu, pp. 98-113, discusses the Hired Service Policy in detail. Paul Smith, in chapter 2 in this volume, discusses the Green Sprouts Policy. For a brief account of Higashi Ichio's findings on some of the most important policies and planning offices, see OAnseki jiten , pp. 43-90.

lance: it reported any crimes of members, new residents, and so forth. Eventually the tithing units were charged with tax collection.

1071, second month: New Examination System and Schools Policy . Followed discussions begun in 1069, fourth month. The various fields were dropped in favor of a single chin-shih examination. In place of the test in poetry, candidates wrote on the greater meaning of ten items from the Classic of their choice (Odes, Documents, Change, Rites of Chou, Book of Rites ) and from the Analects or the Mencius . There were also one essay and three policy proposals. Education officials were to be appointed to all prefectures, with special provisions for the northern circuits. Fields were ordered to be set aside to provision the schools in 1071, third month. The National University was restructured in 1071, tenth month, making it possible for those who passed through the three levels to be appointed directly to office.

1071, tenth month: Hired Service Policy (Mu i fa). Followed discussion and experiments in the K'ai-feng area. Obligatory local service was abolished in favor of hiring men to fill local subbureaucratic posts. Households were assessed a cash tax, graduated according to household wealth (thus requiring an assessment and grading of the wealth of rural and urban households).

1072, third month: [Government] Trade Policy (Shih i fa). Offices were created in major commercial centers to replace wealthy guild merchants as wholesalers, buying from and selling to smaller merchants and traders, as well as loaning money to smaller merchants at interest.

1072, eighth month: Land Survey and Equitable Tax Policies (Fang-t'ien chün-shui fa). Beginning with K'ai-feng and five northern border circuits. To survey land according to standard units, assess the quality of land, and determine ownership. The results, together with other investigations, to be used in correcting household grades so that they accurately reflected wealth (household grades were integral to determining loan eligibility under the Green Shoots and Market Trade Policies, Hired Service tax, and Tithing duties).

1073, third month: Bureau for Commentaries on the Classics (Ching i chü). Responsible for preparing official commentaries on the Odes, Documents , and Rites of Chou . Completed in 1075, sixth month.

According to Wang the most important policies were the policy of expansion in the northwest, the Green Shoots Policy, Hired Service, Tithing, and Government Trading. These were, he noted, also the most controversial, yet their beneficial effects would only gradually become apparent.^[90] These, together with Wang's approach to the bureaucratic system, were also the focus of Ssu-ma Kuang's critique. A comparison of what Wang saw

[90] LCC 41.440.

as the promise of the policies with what Ssu-ma saw as their predictable consequences suggests that the real issue between them was the role of government in society.

Wang An-shih's Defense . What did Wang An-shih think he was trying to accomplish? His immediate concerns largely corresponded to his announced priorities: establishing institutions, changing customs, and managing wealth. Policies in one area were related to those in other areas. To manage wealth required establishing institutions whose effectiveness depended upon changing the values of the officials and the customary practices of the people. I shall make a more general distinction, however, between Wang's efforts to control the activities of government, officialdom, and the pool of potential officials (the literati) and his efforts to establish control over society. This last item will be clarified shortly.

The Finance Planning Commission was Wang's first move toward controlling the operations of government. Wang intended it to be the single organ for formulating an integrated set of policies to "manage wealth." It took authority for fiscal policy away from its traditional home in the Finance Commission, yet it was separate from the Council of State, thus depriving its members of a say in the initial stages of planning. Wang defended this bureaucratic anomaly: it was not incompatible with the structure of ancient government; it was the ancient rule that managing wealth should come before correcting the bureaucratic structure, and effective planning required that the planners share a unity of purpose and not be deterred by "divergent opinion." Wang avoided normal channels again when he appointed investigating commissioners to conduct local inspections and provide policy recommendations to Finance Planning, and yet again when he bypassed the fiscal intendants, the ranking officials of the circuits, and worked through newly created Ever-Normal Granary intendants. In each of these cases he appointed men he believed shared his policy goals. Here too he broke with normal procedure, appointing relatively young, lower-ranking officials to offices that gave them practical authority over older and higher-ranking officials. He then used these offices to promote those who had proved their commitment and effectiveness.^[91] (On all of this see chapter 2 in this volume.)

This creation of a government within a government can be seen as pragmatic, given that Wang aimed to change well-established policies supported by well-established officials. But Wang was unusually concerned with the threat of "divergent opinion" (i lun ), even after he had fully gained the imperial ear as sole chief councillor. He suspected that even objections

[91] HCP: shih-i 4.5 (1069/2); HCP: shih-i 6.1a-2b (1069/11). For an account of Wang's appointments in these areas, see Higashi, OAnseki jiten , pp. 81-86.

to particulars of a policy were really aimed at the basic principles of his program. As Liu Chih put it, two "opinions" or views had emerged within the government:

[One] is secure in the constant and practices the old, happy when there are no problems. [The other] transforms the old and changes policy, pleased by the daring to act. . . . [The first] holds that in keeping to the successful policies of the [imperial] ancestors we need only follow those that are beneficial and, on the basis of traditional [policies], repair their deficiencies in order to achieve order. . . . [The second] holds that the [old] policies are rotten and the [old] way exhausted, and that without a great transformation we will be incapable of managing things and realizing our purposes.^[92]

Wang's policies amounted to a change in what was called the kuo-t'i , the normative form of the state, although he could see them as bringing the form of the state into line with antiquity. Effecting such a change required a unity of purpose, or at least general agreement, among the officials who made up the government. Those who subscribed to "divergent opinion" could be expected to try to thwart the change by stalling the implementation of the policies. Thus Wang did his best to see that leading dissenters were kept out of leading positions. Ssu-ma Kuang should not be appointed to the Council of State, he explained, because he "likes to offer divergent opinions," and his appointment, at a time when "customs are not yet fixed," would encourage lower officials to keep holding off. "If we give divergent opinion an authority to which to appeal nothing can be accomplished."^[93]

However, Wang's antipathy toward divergent opinion went well beyond the pragmatic. Unity was not simply necessary in order to carry out policy, it was itself a sign of the true achievement of the policy goals, and thus part of the goal of policy. For Wang, opposition was immoral because it was divisive; a moral world was one where all men shared the same fundamental values. The imperial mind seems not to have grasped this, for Wang found it necessary to lecture the emperor repeatedly. For example:

It is only that Your Majesty is not firm enough, so that you have not been able to unify morality and change customs. Therefore the clamor of divergent opinion does not cease. If you can act forcefully without tiring and decide every matter according to moral principles (i-li ), then men's sentiments (jench'ing ) eventually ought to change of their own accord. Your Majesty will

[92] HCP 224.5442-43 (1071/6). As Lü Hui-ch'ing explained in a debate with Ssu-ma Kuang, the regime was going beyond correcting faults in the established policies of the dynasty to change the policies themselves. HCP: shih-i 6.6a-10a (1069/11).

[93] HCP: shih-i 7.17b (1070/2) and HCP 213.5168 (1070/7). As James Liu notes, in his useful discussion of factionalism during the time, Wang has not entirely successful. He did not, for example, wrest control over the Bureau of Military Affairs from Wen Yen-po until 1073, See Reform in Sung China , pp. 59-79.

observe that sentiments this fall are already different from those of the spring, so you may be certain they will gradually change.^[94]

The emperor did agree that if policies accorded with moral principles they would be of benefit.^[95]

Wang assumed that since his policies were based on a vision of how each part should function in the larger whole, and since they accorded with the real interests of men, people would eventually become accustomed to them and receive their benefits. It was only necessary to force them to become accustomed.

"At ease with it because it is accustomed; uneasy because it is unaccustomed." That describes the common folk. . . . Your Majesty must act as Heaven acts; then you will be able to protect all under Heaven. . . . [To bring about the year, Heaven must create and kill, and it does so] because it relies on true patterns and is free of sentiment.^[96]

Policies were intended to set out the normative order according to which action should take place. Wang rejected the criticism that he should leave the details of policy to specialized offices closer to the immediate situation:

The Duke of Chou made policy like this; he was not ashamed if it went into complex detail. Taking charge of both the minute and the large is the proper form of government, but the high-ranking are responsible for the great, and the low-ranking for the minute. Such is the model of the Former Kings and the natural pattern of heaven-and-earth. Just as with a man's body: seeing, hearing, eating, and breathing all [take place] in the head, but if he wants to scratch he needs fingernails. The forms are small and big, and the responsibilities differ, but all are necessary. Heaven-and-earth create the myriad things, yet each thing, even something as small as a blade of grass, has its pattern. In governing today we should only ask whether the policies established harm man or not; we should not rescind them because they deal with the minute.^[97]

In dealing with the government Wang had other means of garnering unity besides relying on the emperor. He could, and did, arrange special promotions and rewards for active supporters of the policies. This was, he explained, a means of persuading the majority of officials, men of "medium talent and less," to "achieve things by following the court's laws and institutions" instead of holding back out of fear of his opponents' criticism.^[98] This was fairly effective (helped along by increasingly harsh treatment of opponents), and by 1085 even Ssu-ma Kuang admitted that few men in

[94] HCP 215.5232 (1070/9). Cf. 225.5474-75 (1071/7), 250.6089-90 (1074/2), and 251.6129 (1074.3).

[95] HCP 219.5321 (1071/1).

[96] HCP 236.5742 (1072/interc.7).

[97] HCP 240.5827 (1072/11).

[98] HCP 240.5827 (1072/11).

leading offices were dissenting. The new policies on the examinations, schools, and curriculum also promised to secure intellectually compatible literati for official service. But they went further: henceforth all those who wished to be considered shih and gain office through examinations would be taught the correctness of Wang's policy goals. Wang's memorial on the examinations stated:

The selection of shih in antiquity was rooted in schools in every instance. Therefore when morality was unified above, customs were perfected below and the talents of men such that all were capable of accomplishing something (yu wei ).^[99]

There were limits to this, of course. Men passed without subscribing to Wang's learning, provoking the comment that, while "eight or nine tenths of the literati have been transformed by the methods of the classics," just as many "have not sought mental comprehension."^[100] The restructuring of the National University perhaps offered more promise of ensuring the correct education of future officials.

Wang also sought unified control over government through his policies on clerks and local service personnel. Salaries were instituted for clerks to free them from dependence on bribes and make them more dependent on the institutions they served. The Hired Service policy in effect removed the leading local families from the local institutions that most affected their interests, replacing them with full-time personnel who lacked independent means. When this did not prevent corruption, Wang added a further enticement: he ordered that the best local clerks be recommended for ranked appointments. In doing so he had a further goal: that "henceforth good literati may be willing to serve as clerks; when good literati are willing to serve as clerks, then clerks and literati may once again be united as one, as in antiquity. . . . This was a priority of the Former Kings."^[101]

Once government began to function as a single whole, the policies would take full effect. But what did Wang think they would achieve? They would enable the government to organize the distribution of all wealth under heaven. As Wang explained: "To manage wealth, the ruler should see public and private [wealth] as a single whole."^[102] All wealth was to be subject to government control, just as all people were subject to social organization through the Tithing system (if they were not already organized as officials or government personnel). Wang maintained that managing wealth was far more important than increasing government revenues. When the wealth of society was correctly managed, the revenue problem would disappear of

[99] HCP 220.5334 (1071/2).

[100] HCP 248.6056 (1073/12).

[101] HCP 214.5223 (1070/8); 237.5764 (1072/8); cf. 215.5230-31 (1070/7).

[102] HCP 214.5223 (1070/8).

itself, but to make the increasing of revenues the end was to miss the point.^[103] Managing wealth was the basic means of establishing the control over society necessary for creating an integrated human order that could be commanded from above. Just as control over government required silencing those independent-minded officials on whom other officials depended, so control over society required suppressing those of independent means on whom the poor depended. Benefiting society (li min, li t'ien-hsia ) involved controlling those private interests that, by their simple existence, threatened the government's command of society. Wang's logic might be stated in this fashion: for the people to receive benefits, they must be organized; to be organized, they must be willing to take direction from the government; to be willing to take direction, they must find that their material interests depend upon their doing so. Those social elements that came between the people and the government, that is, those who used their wealth to make others dependent on them, were an obstacle to this end.

Wang called these elements "engrossers" (chien-ping ). By this he meant all those who used wealth to make others dependent upon them, in either agriculture or commerce. Just how central "suppressing engrossers" was to Wang is evident from his initial justification for the Finance Planning Commission. Such an organ, he claimed, was "how the Former Kings suppressed engrossers, aided the poor, circulated the world's wealth, and caused benefit [or material interests] to come from a single source."^[104]

Only the greatly activist rulers of antiquity were able to suppress the engrossers. Those called engrossers are all leading men of means; their opinions are capable of moving the shih ta-fu . If in instituting policy today we merely go along with what suits popular sentiment in every instance, we will not be able to control engrossers. . . . [If we do not], how will we be able to equalize the world's wealth and free the common folk from poverty?^[105]

Suppressing engrossers required that the government use its power to affect material interests. In the following passage Wang is ruminating on the advantages of the T'ang tax system, which realized somewhat "the intent of the Former Kings" by taxing adult males, not household wealth, and by distributing land equitably:

If the ruler truly knows the interests of society (t'ien-hsia li-hai ) and institutes policies that confer what society regards as "loss" upon the engrossers, then they will not dare keep fields beyond the allotment. If he institutes policies that confer what society regards as "profit" upon those who till, then they will be encouraged to till and will not take fields beyond the allotment. But this must be done gradually in order to become established policy. Now if the

[103] HCP 251.6129 (1074/3).

[104] HCP: shih-i 4.5 (1069/2).

[105] HCP 223.5433-34 (1071/5).

ruler can truly control profit and loss, and according to his values confer them [upon the people], he will not have to worry about getting people to do what he likes and avoid what he dislikes.^[106]

Suppressing engrossers was of such importance that it alone made Sung superior to past dynasties; even Ch'in had been unable to accomplish this. "In my opinion, from Ch'in on there have never been methods for suppressing engrossers, until today. Your minister thinks that if we are able to suppress engrossers, then managing wealth will work and we will not have to worry about a lack of wealth."^[107] Suppressing engrossers was a moral issue. In defending a proposal to have local state trading offices "investigate engrosser families that control profit and harm the New Policies" Wang asserted: "Heaven has given Your Majesty the nine regions within the four seas; certainly it intends Your Majesty to suppress the powerful and raise up the poor, so that poor and rich receive [heaven's] benefits equally." Such men were truly unworthy, yet some objected that the salaries Wang wished to give his officials were too munificent. Wang responded: "Today in every prefecture and subprefecture there are engrosser families who annually collect interest amounting to several myriad strings of cash without doing anything. . . . What contribution have they made to the state to [warrant] enjoying such a good salary?"^[108]

The New Policies created institutions that took over the functions of the engrossers. But this did not mean that the government was after material gain. "As to establishing the correct form of the state: suppressing engrossers and collecting their takings in order to accomplish things of benefit and relieve distress were the affairs of government of the Former Kings. These are not to be termed 'valuing profit.'"^[109]

For Wang a moral world was a unified, integrated world, free of engrossing. The Green Shoots and Government Trading policies were intended to supplant the engrossers as the source of rural and commercial credit. Suppressing engrossers, that is, meant depriving them of investment income, making the majority of small farmers and traders independent of them. It also meant making the people dependent on government institutions, thus allowing the government to reduce interest rates somewhat while using the profits to invest in economic development.^[110] As Wang explained with regard to Government Trading: "When the small traders have to rely on the great houses, the great houses take the larger share of the profit and the

[106] HCP 223.5419 (1071/5).

[107] HCP 262.6407 (1075/4).

[108] HCP 232.5640-41 (1072/4); 240.5829 (1072/11).

[109] HCP 240.5828 (107/11).

[110] For the goals of the Green Shoots Policy, see, for example, HCP: shih-i 5.19 (1069/5) and 7.7 (1070/2). For Government Trading, see HCP 236:5736-39 (1072/interc.7) and 240.5826-28 (1072/11).

small traders' share is slim. Now officials take the slim profit, and the small traders gain the larger profit for themselves; how does this harm the small traders?"^[111]

Although the Hired Service policy collected the service-exemption tax from poorer households as well, the tax was graduated. Wang argued that the burden on poorer households was slight, that their real burden was exploitation by engrossers. When the service-exemption tax produced far more than was needed to hire local service personnel, Wang defended the policy: because the engrossers paid more, the poor could pay less; and because there was a surplus, there was now money available for famine relief and agricultural projects. No longer would the government need to demand relief loans from the rich. As long as the money was not wasted on imperial extravagances, Wang concluded, it could not be called exploitation. We might note that government agricultural projects and famine relief funds also challenged the wealthiest families' role as the leading investors in local projects.^[112]

The Tithing policy (pao-chia fa ) exemplified Wang's ideal union between government and society. While the initial rules made local security the point of the policy, Wang repeatedly said that his real goal was to "make farmers and soldiers one."^[113] With training, he claimed, the militia units would eventually become better than regular soldiers, and they would be cheaper.^[114] At the same time, it promised to establish a new set of local leaders, men who would command the respect of the militiamen and the community and lead the people to take direction from the government.^[115] The Tithing policy offered an alternative basis for community, one that disregarded family ties and private status and kept the people oriented toward the state. Wang soon concluded that the units could take on other official responsibilities. By 1075 they were collecting the land tax, service tax, and Green Shoots loans. Charging the people with many offices was, Wang asserted, the way the Former Kings employed the people.^[116] "Once the policies are in effect everyone will be of use. We will be employing all the

[111] HCP 251.6124 (1073/3).

[112] HCP 223.5433 (1071/5); 237.5776-77 (1072/8). For the promulgation of the policy, see 227.5521-24 (1071/10).

[113] HCP 218.5299-300 (1070/12) and 237.5764 (1072/7); for the initial regulations, see 218.5297. For a full account of Wang's plans and the emperor's doubts, see Teng Kuang-ming, "Wang An-shih tui Pei Sung ping-chih kai-ko ts'o-shih chi she-hsing," in Sung shih yen-chiu lun-wen chi (Shanghai: Shang-hai ku-chi, 1982), pp. 311-20.

[114] HCP 235.5697 (1072/7). Regular military training took place mainly in the northern border circuits.

[115] HCP 218.5297 (1070/10). The rules required that pao-chia leaders be men of ability and intelligence from landowning households of any grade.

[116] HCP 263.6436-37, 6451 (1975/interc.4).

people under heaven to do all the affairs under heaven and we will be free of the problem of useless people."^[117]

Ssu-ma Kuang's Critique . Ssu-ma began to memorialize on Wang's program in 1069/8. He repeated the same objections in 1085-1086, when he was in power. Knowing what Wang An-shih thought he was doing makes it easy to get the point of Ssu-ma's "divergent opinions" on the normative form of the state, the structure of government, the relation between government and society, and the necessary order in society.

Ssu-ma continued to insist that "order" depended on maintaining a clearly defined hierarchy of delegated authority, with a corresponding division of functional responsibilities. Thus while Wang was willing to break down established divisions within government in order to establish unified control and carry out his policies, Ssu-ma wanted to keep levels of government discrete. Wang wanted to integrate government and society; Ssu-ma argued for a necessary boundary between the institutional activities of government and the traditional procedures through which the people pursued their material interests. Whereas Wang aimed to break the power of private wealth, Ssu-ma defended the particular and necessary social function of the rich. And even in foreign policy, Wang could imagine integrating barbarians into China, while Ssu-ma favored a balance of power among states.

One of Ssu-ma's first targets was the way Wang was changing the administrative process, using the Finance Planning Commission to subvert the discrete levels that properly constituted the pyramid of government. This placed inexperienced, low-ranking literati in charge of formulating policy in areas where they lacked competence without advice from the responsible offices, and it produced policy so detailed that local variation was impossible. The use of commissioners to investigate local conditions, and later to oversee policy implementation, was wrong for the same reasons. The fiscal intendants were ignored, local government was investigated by men without adequate knowledge and experience, and proposals were made and policy effected without input from local officials. If established policies were not working well—and Ssu-ma agreed there were real problems—the solution lay in better personnel administration to insure competence in planning and practice. Later Ssu-ma publicly attacked Wang's conduct as chief councillor. Wang, Ssu-ma said, saw himself as another Duke of Chou aiding King Ch'eng, but was nothing more than a man who insisted that everyone share his opinions.^[118]

[117] HCP 215.2312 (1070/9).

[118] SMK 43.547-50; cf. 60.719 (1070/2); 45.572 (1074/8) and 17.258 (1082). He had already taken Wang to task in 1070/2 in a series of three letters: see 60.719-27.

Ssu-ma had not lost his belief in the primary importance of administrative process. But he quickly realized that the real issue lay in what Wang hoped to use government to do. This was an issue of principle. Should the government be concerned with "profit" (li )? That is, should the government try to manage the wealth of society as a whole and so interfere in the way the people realized their material interests, instead of letting them pursue those interests for themselves? Ssu-ma did not think "managing wealth" was an integral part of ancient government. At most, government should insure that its policies did not prevent the people from enriching themselves. Ssu-ma held that the government should tax only to satisfy the minimal requirements for maintaining order. There was a world of difference in his eyes between that goal and an attempt to control how society produced wealth. .^[119]

To prove that Wang's policy goals were wrong, Ssu-ma defined what he saw as their long-term consequences and explained how these would eventually deprive the government of the resources necessary to maintain the state. The problem was that the policies changed existing relationships in society that were necessary for social stability and prosperity. The rich, for example, were both socially useful and politically necessary. The difference between rich and poor arose from natural differences in intelligence and ability among the people. The rich were those who planned ahead and worked hard, while the poor were those who did not. There was a mutually beneficial relationship between the two: the rich lent to the poor to enrich themselves, and the poor borrowed from the rich in order to survive. Now, through the Green Sprouts loans, the government was establishing itself as the sole source of rural credit, which appeared to serve the interests of the poor who needed to borrow. But in fact the institutional demands placed on officials kept them from fulfilling the traditional role of the rich, who after all needed to keep the poor in place to profit from them. To meet their loan quotas officials had to force all households to accept loans. To cover defaults, rich and poor were forced to join in loan guarantee units; thus in a bad harvest the poor (who could not manage their finances) would default, leaving the rich responsible for their debts. The cumulative increase in indebtedness would soon carry over into good years as well. As the poor fled, the rich would gradually be impoverished (they would no longer hold mortgages on the land of the poor borrower). Eventually the government would have to excuse the debt. When debts were excused the Ever-Normal Reserves, originally intended to be the major source of emergency funds for famine relief and military exigencies, would be depleted and their benefits lost to the state. The impoverishment of the rich, a traditional source of local relief funds, further harmed famine relief policy. In effect the govern-

[119] SMK 60.719-24, 726-27; 41.523-24.

ment was creating the source of rebellion—landless refugees—while giving up the fiscal means to cope with it.^[120]

The creation of a service-exemption tax to be collected in cash led Ssu-ma to his second major argument against "managing wealth." Collecting taxes (and loans) in cash, he held, fundamentally changed the nature of the rural economy, which had previously been based on the products of the people's own labor with taxes levied only in kind. The new system made producers dependent on the market, since they had to sell their goods to get the cash to pay their taxes. Cash taxes made the poorest households suffer the most, for they were most easily harmed by price fluctuations. To pay their taxes the people would have to accept the lower prices available at harvesttime; this and the increased demand for cash would cause the value of goods relative to money to decline. Ssu-ma described the problem of what others called "cash famine": goods remained cheap, even in bad years, because money was expensive; thus marginal farmers were forced to sell real property. At the same time, the creation of official trade offices reduced the profit margins in commerce, decreased the number of traveling merchants, and harmed the flow of goods.^[121]

The monetarization brought about by the new tax system, according to Ssu-ma, led to commercialization, destroying rural self-sufficiency and so rural stability. The pao-chia policy further aggravated the situation. It forced the people to take on duties that interfered with agriculture, their proper function in the larger structure of the state. Yet they did not receive the kind of training in police methods and military arts they needed to serve competently as soldiers and police. In sum, the policies to manage wealth forced the people into flight and banditry, while the pao-chia system taught the people enough martial technique to become bandits yet deprived local government of effective security forces.^[122]

Ssu-ma's analogy between the state and a building can be applied to his critique of the New Policies. The people are the foundation on which the government is constructed, while high and low officials, generals and soldiers serve as different parts. Wang was taking the building apart in order to put it back together in a new way. Ssu-ma Kuang objected that the different pieces could only serve certain purposes effectively. Low officials could not play the role of high officials, farmers could not function as soldiers. In Ssu-ma's view society (i.e., the people) was the foundation on which government existed. Wang was trying to change the internal structure of that foundation by depriving the rich of their function. In the end this destroyed the stability and prosperity of society, and thus, however intricately the

[120] MK 44.559-63.

[121] .SMK 45.575-77, 17.258-59.

[122] .SMK 17.259.

government was structured, it destroyed the foundation of government and led inevitably to the collapse of the state. As long as political unity did not depend upon the government's ability to command society, whether the rich kept the poor in a state of dependency or not was of no concern to the government. On the contrary, it was in the interest of the state to leave wealth in private hands in order to maintain social stability, as long as the government could fulfill its fiscal needs. In Ssu-ma's scheme of things the government had no legitimate reason to threaten the interests of the rich, but it did have legitimate cause for taking the interests of the rich into account. They, after all, had direct responsibility for the poor.

Ssu-ma's conviction that the state should come to terms with society as it was, rather than try to control it, was also evident in his attitude toward the relation between the government and the literati. In contrast to Wang's desire to bring the literati under government tutelage and integrate them into the government and its policies, Ssu-ma preferred to ignore state schools and prefectural qualifying examinations in favor of a recommendation system. The best way of staffing government, he argued, was to admit to the examinations only those men recommended by high officials (who were free to recommend kin), giving preference to those with the greatest number of sponsors.^[123] Government was to be an enterprise of an existing national elite. It should not try to create a new elite.

Ssu-ma Kuang's and Wang An-shih's foreign policies reflected the same division. Wang supported Wang Shao's efforts to recover territory and absorb new population in the northwest, and, although he warned that his domestic program should precede further foreign entanglements, he could envision integrating the barbarian peoples into China. "If we can transform the barbarians into Chinese [Han], that is good." In contrast to Wang's desire for integration and unity, Ssu-ma consistently argued in favor of a balance of powers between Sung and foreign states. These were different peoples with different interests. He even called for returning territory taken from Hsia.^[124]

Ssu-ma was never in any doubt about what should be done to resolve the crisis of state the New Policies had brought about. When he came into power he moved immediately to rescind all of the policies.^[125] At the same time he restated his general principles on the roles of emperor, court, and bureaucracy,^[126] and he proposed various measures to improve administra-

[123] SMK 40.517-22.

[124] HCP 221.5370, 5384-85 (1071/3); 245.5963 (1073/5); SMK 42.537-38; 45.569-71; 45.571-78; 17.257-60; 50.631-35.

[125] For a general statement, see SMK 46.588-91. This is detailed in memorials on the pao-chia , military, and hired service; 46.591-94; 48.611-13; 47.608-9. Not all were persuaded that the hired-service system should be abolished; see 49.626-28 and 55.669-71.

[126] SMK 46.586-88.

tive process.^[127] For the first time Ssu-ma Kuang argued that the only justification for the New Policies had been Shen-tsung's desire to extend the borders of Sung to match those of Han and T'ang. Placing the blame on the emperor was consistent with Ssu-ma's understanding of his role in the polity. He concluded that if there was no longer a desire to "employ the troops," there was no longer a rationale for any of the New Policies.^[128]

Scholarship

The scholarly activities Wang and Ssu-ma pursued during these years gave further support to their political visions. Here I shall only note the larger outline. Wang oversaw the preparation of commentaries on the three Classics, writing that on the Rites of Chou himself. More than any other work from antiquity, he claimed, this text revealed what it was like when "The Way was present in the affairs of government" and true order prevailed; it was "the best source for the policies that can be applied to later ages."^[129] Wang's commentary does make the Classic support the New Policies, but it does more. It provides students with a model for finding meaning through envisioning the coherence of the text as a whole. Wang's Explanations for Characters was a unique work. Surviving explanations, and examples from his Rites of Chou commentary, illustrate his method of integrating the parts into a coherent whole as a means of defining values. Here too he adduced later knowledge to achieve that coherence (in a manner that seems quite ad hoc), just as he brought later knowledge to bear on the Classics to illuminate their great system. His confidence that his book had set out normative values for all affairs is evident: "Is it not that Heaven intends to revive this culture of ours (ssu wen ) and has used me to aid in its beginning? Therefore instruction and learning must begin with this [book]. Those able to know this [book] will have nine-tenths of the meaning of morality."^[130] In retirement, as is well known, Wang turned to spiritual interests. Although their exact nature remains vague, Wang's interest in Buddhism suggests that he was still intent on seeing the unity of Way with his own mind.

Ssu-ma Kuang finished the Comprehensive Mirror for Aid of Government in 1084. It was massive evidence for his views on the timeless principles of government. Those views were succinctly stated in the first of his interjected comments, which serves as introduction to the entire work. It tells

[127] SMK 54.661-67; 55.671-75; 57.685-90, 51.641-42; 50.638-40. These all date from 1086.

[128] SMK 49,624-26.

[129] LCC 84.878. On the connection between the Rites of Chou commentary and the New Policies, see Liu K'un-t'ai, "Wang An-shih 'Chou kuan hsin i' ch'ien-chih," Ho-nan ta-hsueh hsueh-pao , 1985, no. 4:87-92. The best edition of this commentary is Ch'ien I-chi's edition in the Ching yuan .

[130] LCC 84.879; cf. 56.608.

how the Chou order finally fell apart when, in 403 B.C. , the Chou king Weilieh allowed the three ministers (ta-fu ) of Chin to divide their state into three. Chou's failure to maintain the correct forms of political authority set the stage for its own destruction; the fault lay with the king himself. The first part of his comment will suffice to show that Ssu-ma's understanding of the correct form of the state was not the same as Wang An-shih's:

Your Minister Kuang says, "Your Minister has heard that of the Son of Heaven's responsibilities (chih ) none is greater than ritual (li ), that in ritual nothing is greater than roles (fen ), and that for roles nothing is greater than names (ming ). What is ritual? It is rules (chi-kang ). What are roles? They are ruler and minister. What are names? They are duke, lord [of a subordinate state], grand minister, and minister.

Now, that the broad land within the four seas and the multitude of people took direction from a single man, so that even those of exceptional strength and extraordinary talent dared not but rush to serve, was because that man used ritual to make rules for them. For this reason the Son of Heaven brought together under his control the Three Dukes, the Three Dukes led the Feudal Lords, the Feudal Lords directed the Grand Ministers and Ministers, and the Grand Ministers and Ministers ordered the shih and common people. The noble ruled the humble and the humble served the noble. Superiors' directing inferiors was like the heart's employing the limbs and the root and trunk's ordering the branches and leaves. Inferiors' serving superiors was like the limbs' guarding the heart and the branches and leaves' screening the root and trunk. Only then was it possible for superiors and inferiors to guard each other and for the state to be ordered and secure. Therefore I say that among the Son of Heaven's responsibilities nothing is greater than ritual.^[131]

Ssu-ma's commitment to a hierarchy of delegated authority extending from the ruler down to the lower officials is also evident in what he says about the Way in his commentaries on the Book of Change and Yang Hsiung's T'ai hsuan .^[132] Although he treated heaven and man, number and principle as parallels, his conclusion had more to do with "man" than with "heaven" and self-consciously avoided mysticism.^[133]

Ssu-ma did have spiritual interests as well, or at least an interest in "mind." He concluded that when men devoted to learning and acquiring knowledge succeeded in keeping their thinking mind undisturbed by external things, the mind would see all sides of a matter and find the mean. Achieving this state of mind he called arriving at chung or chung-ho (glossed as neither exceeding nor falling short). Men who attained it could keep the body healthy without medicine (and thus control their "destiny") and know

[131] Tzu-chih t'ung-chien 1.2-3.

[132] I shuo (SKCS ), "Tsung lun." SMK 67.834-35. His interest in numerology dated back to the 1040s; see T'ai hsuan (SPPY ), preface.

[133] SMK 74.913-16.

which choices were necessary to bring order to all under heaven.^[134] Perhaps we can see in Ssu-ma's ideas about chung his conviction that harmony arose from balancing opposing parts and interests.

Conclusions

What were the essential differences between the political visions of Wang An-shih and Ssu-ma Kuang? Why were two such different visions articulated during the eleventh century? How was the historical choice they represented related to the course of Chinese history?

One obvious difference became apparent in the 1070s: Wang expanded the scope of government activity, while Ssu-ma wished to limit it. Yet a relative difference over the appropriate scope of policy, in my view, cannot by itself account for half a century of acrimony and partisanship. Rather, we must look to the underlying political visions. Not all would accept this. Some might interpret Wang's policies simply as a response to the fiscal and defense problems of the Sung government and suppose that the problems generated the solution. Wang, then, was a reformer who saw the problems and knew how they could be solved, though he was perhaps too idealistic.^[135] But Ssu-ma saw the same problems and proposed solutions no less plausible. Wang's program was not the only available choice. We can still ask why these two different solutions gained such followings. But the best reason for taking their underlying visions seriously is that Wang and Ssu-ma each saw his own normative form of the state as the necessary basis for

[134] Ssu-ma developed his ideas about chung in 1084-1085 in correspondence with Fan Chen (T. Ching-jen), who argued, to the contrary, that bringing order depended on supreme political authority, and Han Wei (T. Ping-kuo), who tried to persuade him that if the mind was made "empty" it would gain access to the "great basis," which was the one source of the "multitudinous bases" of the many things. The correspondence appears to have taken place in the sequence that follows: (A) 62.752, SMK to Fan with whom he had been discussing music; (B) 62.758-59, Fan's reply to (A); (C) 62.759-60, SMK responds to Fan (B); (D) 62.670-761, Fan replies to SMK (C); (E) 62.761, SMK replies generally and sends an essay, "On chung ho "; (F) 64.793-96. At this point Han Wei joins in the debate; see (I) and on; (G) 62.761-62, SMK notes Fan's disagreement and sends another essay, "Attaining Knowledge Depends on Guarding against Things," and a piece written in 1083 (included in the "Impractical Writings" entitled "Forsaking the Four"; (H) 65.800 and 74.911-12; (I) 62.765, Han informs SMK of his intention to participate; (J) 62.765, Han gives SMK his understanding of chung ; (K) 62.766-68, SMK replies to Han; (L) 62.769-70, Han replies to SMK (K) (note that this is incorrectly given as a reply from Fan Ching-jen in the collection); (M) 62.768-69, SMK replies to Han (L). The immediate origin for Ssu-ma's theory of chung appears to have been his interest in cultivating physical health (yang sheng ); see 61.734-35, "Ta Li Ta-ch'ing shu" from 1072, and 62.751, "Yü Wang Lo-tao shu" from 1080. Tanaka Kenji, in Shisei tsuken (Tokyo: Asahi, 1974), pp. 8-10, has noted the value of this debate.

[135] This seems to be the general thrust of the views collected in John Meskill, Wang An-shih: Practical Reformer? (D.C. Heath, 1963).

resolving all the problems the government faced. Both were willing to sacrifice expedient, short-term solutions in the interest of their larger principles. As Ssu-ma pointed out, Wang had no interest in trying remedial measures; he countenanced continued deficits throughout his reign. Ssu-ma, for his part, was willing to restore the obligatory-service system, the burdens of which he had pointed out, in order to avoid the further monetarization of the rural economy he attributed to the hired-service system—this though many of his allies believed that hired service was preferred by the people.

What were the visions? Wang imagined a state without a distinction between government and society, between the political and the moral, and whose institutions fulfilled the common desires and needs of all men. Ssu-ma, however, aimed at securing the survival of the state as a political entity created on the basis of the existing society. To accomplish this, he tried to perfect the structure and operation of government, drawing a sharp line between public, institutional responsibilities and the private interests of those who bore them, and seeking to block the encroachment of one on the other.

Wang An-shih believed that public power should command private interests. Ssu-ma believed that this would destroy wealth, turn the people against the government, and make it impossible for the government to maintain the state. Wang, however, thought that Ssu-ma's efforts merely to defend the public realm from private interests and to perfect the administration of government were incapable of accomplishing anything of lasting value. In practice both saw the state as containing both government (court, civil officials, the military, laws and institutions, etc.) and society (literati, farmers, merchants, etc.). They both assumed, moreover, that there was a necessary relationship between government and society, that government policy directly affected (for better and worse) the values and material welfare of society, and that it was the responsibility of literati to use their learning to determine how government acted. They understood the correct form of the state (kuo t'i ) differently, and they understood differently the relationship between government and society necessary to establish and maintain that form.

Wang did not disagree with Ssu-ma's assessment that government existed within the context of private interests, which could threaten the government's power to maintain the state, but he believed that government had the power to keep those interests in check by making it advantageous for them to accept its own powers and functions. Ssu-ma's response was to ask how government could keep the support of private interests without harming its own functioning. His answer comprised two main elements: first, the ruler should ensure that the functions of government necessary to maintain the state (defense, judicature, revenue supply) were carried out fairly and impartially (i.e., that they had no other purpose than to preserve the state); and second, the men to whom the ruler delegated authority to

administer government should be selected from among the shih as the most powerful, wealthy, and talented families in society. Thus, while government worked to defend political unity and stability, so that men would be free to pursue their interests and "enrich themselves," all at minimum expense to society, it co-opted into its service those who had the most to defend. Wang An-shih, however, saw no need to accommodate the interests of wealthy, independent families, for he believed that if government took direct institutional responsibility for social values and welfare, it could keep the support of society without their mediation and, at the same time, ensure an equitable distribution of wealth. But this also required actively suppressing those who used their wealth and position to make others dependent on them and so prevented government from gaining command over all under heaven.

It is certainly fair to call Ssu-ma a "conservative." He generally accepted the existing structure of society, with its unequal distribution of wealth; he held that the existing political order could be made to work effectively; and he saw no need to create new institutions or for the government to claim new social responsibilities. He accepted the existing social order because he believed there was an enduring balance between rich and poor in a self-sufficient rural economy. He saw commercialization as destroying that balance, but he blamed commercialization on government policy and the self-interested behavior of the shih , both of which he thought could be controlled. He assumed that the government could block political and social change. I am much less sure that we can call Wang a reformer. He was demanding radical social change. He was trying to bring about a "revolution" in the true sense of the word, a return to the beginning of civilization as the sage kings of antiquity had created it.

Wang and Ssu-ma claimed to have gained their knowledge of the principles they espoused from studying the human past, Wang claiming the Classics and Ssu-ma the historical record. But since not all classicists reached Wang's conclusions, nor all historians Ssu-ma's, it behooves us to look at their methods. Here I would emphasize the very different ways the two understood "unity" or "oneness." The issue may appear overly abstract, but both men wrote on it explicitly as a philosophical question; and their notions of unity were reflected in their scholarship and in their ideas on the form of the state. The difference can be seen clearly in their view of how parts must go together to form a coherent whole. Wang assumes an integrated, organic whole, and then figures out what the function of each part is relative to all the other parts given. In his true union each part takes direction from the guiding intent of the whole and functions in the interest of the whole. Wang's task, in scholarship and in government, was to see what the "intent" of the whole was and then to determine the "intents" the respective parts fulfilled in working together to fulfill that unifying intent. I am quite persuaded by Ssu-ma's charge that Wang was "literary." For

Wang the Classics are a single, coherent literary work, in full accord with an inclusive, integrated natural order.

Ssu-ma's view of unity is not organic. It is a construct of divergent tendencies or forces, each playing a role, according to its ability, in a balanced structure. When the pieces are in good repair and kept in place, the structure will not collapse. The union survives, not because when joined correctly the parts are transformed into a single body taking directions from a single mind, but rather through diligent and deliberate maintenance. The difference is illustrated in the two men's greatest scholarly works. Wang's commentary on the Rites of Chou and his Explanations for Characters begin from the assumption that all the pieces are organically related. Ssu-ma, in his Comprehensive Mirror , consciously constructs a historical narrative from a collection of acknowledgedly independent texts and contradictory accounts.

This divergence bears on how they thought about the state. For Wang the issue was the "original intent" of the sages in creating the state (to benefit all under heaven), and he assumes that if he can see the unifying "pattern" (li ) of the original creation, it can guide the present. Moreover, since all things are in principle part of a greater whole, he assumes that they can be reintegrated according to how their individual patterns fit into the larger one. Wang can accommodate historical change and can change the way things are without fear that the state will collapse, as long as he is able to control the relations between all things. Ssu-ma, however, sees the state as the creation of a particular historical moment by a particular person. It exists because rivals were vanquished and the existing groups (armies, officials, shih , the people) were brought together in a mutually supportive structure that satisfied the particular interests of each. Looking back on a succession of historical states, he sees that there are necessary rules (chi kang ) for maintaining any such structure. But once the pieces are put together at a particular moment, the point of following the rules is to prevent any structural changes from taking place. An illustration of this difference is the contrast between Ssu-ma's profound fear that commercialization would distort the balance between rich and poor (by giving greater advantages to the wealthy and driving the poor off the land), and Wang's ability to incorporate the ongoing commercialization of Sung society into his program.

The models of the unified state Wang and Ssu-ma articulated were not unique to the eleventh century, although they were formulated in eleventh-century terms. What Wang had in mind should remind us of the Ch'in dynasty and of Wang Mang's Hsin ("New") dynasty, just as Ssu-ma's ideas are more consonant with Han and T'ang. Similarly, one might associate the Kuomintang's willingness to have government accommodate private interests with Ssu-ma Kuang, and the Communist Party's commitment to make private interests dependent upon public institutions with Wang An-

shih. The current trend in the People's Republic toward elevating Ssu-ma Kuang and rejecting Wang is quite consciously part of a general rethinking of the proper relationship between government and society.

It is striking that although New Policies advocates dominated the government of Sung for almost half a century, from Southern Sung on the relationship between government and society in China generally accorded more nearly with what Ssu-ma had envisioned. Some would see exceptions in the Ming dynasty. But Chu Yuan-chang, the founder, seems to have adopted a similar vision of the state. He saw a need for clearly dividing public and private realms, preventing the shih from using government to further their private interests, and securing a noncommercial, self-sufficient agricultural economy. In his relations with the bureaucracy he did, of course, go well beyond the limits of his function as Ssu-ma defined it. Chang Chü-cheng, the great Ming statesman, insisted, like Ssu-ma, on the value of the dynasty's institutions. He too saw the task of government as ensuring the survival of the state and sought to defend the public nature of government against private interests. Unlike Ssu-ma, he did not develop a principled political vision, and he accepted the reality of commercialization, but he did not try to make government directly responsible for benefiting society.^[136] The fiscal reforms of the Yung-cheng period in Ch'ing, too, were remedial measures; they did not change the form of the state.

We can partially account for the turn away from the New Policies approach to government by referring to the realities of the 1120s and 1130s. The restoration of Sung under Emperor Kao-tsung, an admirer of Ssu-ma's Comprehensive Mirror , required the accommodation of existing interests—the issue was simply whether they would support the restoration of Sung political authority—and efforts to institutionalize government command over society were detrimental to that end. But in a larger sense, the abandonment of attempts to transform society through institutions has to be attributed to the unwillingness of the shih as a group to support such changes. The opposition of many bureaucrats to the New Policies was not, after all, an insurmountable problem; bureaucrats could be replaced. It was more difficult to replace the shih . How should we account for their opposition?

Here I propose a way of understanding intellectual and social change in the eleventh and twelfth centuries that will allow us to answer three questions at once: Why were larger political visions being articulated in the eleventh century? Why did Ssu-ma Kuang's view on the relation between government and society prove more acceptable to the shih ? And why did literati learning shift in the direction of Tao-hsueh ?

[136] My reading of Chang Chü-cheng is based on Robert Crawford, "Chang Chü-cheng's Confucian Legalism," in Self and Society in Ming Thought , ed. Wm. Theodore de Bary and the Conference on Ming Thought (New York: Columbia University Press, 1970), pp. 367-414.

Much of what I shall say revolves around three issues: first, the role of learning in establishing men as shih ; second, the tension between the political meaning of shih (those who "serve" the state) and the social meaning of shih (those who are distinct from the "people" [min ] and are elite rather than "common" [shu ]); and third, the problem created for the shih when it became impossible for the vast majority to serve in government.

Through most of the eleventh century the shih saw themselves as men who "served" (shih ), and they assumed that the purpose of "learning" was to prepare for sharing political responsibility. Until the 1070s most scholarly or polemical writing assumed an audience primarily concerned with government. In the early eleventh century the shih were still assumed to be an existing group of families with traditions of official service. When the examination system was first expanded in the late tenth century, as north and south were unified, the aim was to provide a mechanism that would bring into government men from shih families lacking personal connections to the founding elite. But the examinations complicated things, for in effect they made "learning" the primary criterion for official service. As was noted at the time, men were now claiming to be shih because they had mastered the learning necessary for examinations, even though they did not come from families with traditions of service. Once they became officials it was too late. They were now unquestionably shih , and they had established the beginning of what their descendants could see as a family tradition of service. The practice of posthumously granting official rank to the ancestors of high officials was an institutional recognition of this desire.

Those who began writing on the ideal goals that ought to inform politics, men like Ou-yang Hsiu, were speaking to and for the ambitious men who were establishing themselves as shih through learning. Such writers accomplished two things. First, by presenting themselves as men truly committed to the public good, they justified their own desire to participate in policy-making at a time when the upper levels of government were dominated by shih from already established families. Ou-yang's attack on the Five Dynasties statesman Feng Tao, for example, was an attack on all shih who put family above state.^[137] Second, they explained to those who relied on learning rather than family why their learning could justify their claim to political responsibility. Indeed, they were asserting that the only true shih were those who rose through learning and who could determine political values through their learning. For these new shih , who lacked family traditions of service (or whose traditions were short and of recent beginnings), elite social status—equality with men of longer pedigree—was a direct conse-

[137] Ou-yang wrote the New History of the Five Dynasties in the 1020s. For his critique of Feng Tao, see Wang Gung-wu, "Feng Tao: An Essay on Confucian Loyalty," in Confucian Personalities , ed. Arthur F. Wright and Denis Twitchett (Stanford: Stanford University Press, 1962), pp. 123-45.

quence of their official career. Anxious to gain office and rise in the bureaucracy, they tended to see the ideals they found through learning as goals that could only be realized through government action, and they assumed that policy should be a function of learning. Men from long-established shih families, who had access to yin privilege, facilitated examinations, and family connections to the highest officials, were born shih . Their family traditions suggested they had the "right" to an official career and gave them a certain interest in maintaining the status quo. Their elite status was mediated by the fact of their birth.

The "ancient style" writers' idealization of antiquity as an integrated order, in which government and society were indistinguishable and the government's institutions were also the institutions of social morality, provided a "teaching" that could guide "policy." But the idea that government should take institutional responsibility for the welfare of society also served the new shih . Such a program gave government a reason to promote shih with plans for new government action (which would also bring a need for more officials), and it envisioned the elite of society as the officials of government (who in turn were men working to change things). Wang An-shih made his reputation in pursuing this vision, and his rise to power was a sign that it had come to be shared. Wang, from a family of as yet only relatively low-ranking officials, envisioned the shih as men produced through government schools whose social role would be their political role. His "Ten-thousand Word Memorial" of 1058 presented the transformation of the shih as the basis for realizing the antique ideal. Ssu-ma Kuang, alarmed by the idea that an ideal model should determine policy choices without respect to consequences, attacked all those who relied on talent and ideal plans to justify a claim to political power. The son of a high-ranking official and holding official rank before he took the examinations, Ssu-ma put his formidable scholarship to the task of establishing the correct form of the state. He also viewed the shih as a historical elite, existing independently of the state yet necessary to every state because of their social leadership, scholarly and administrative traditions, and moral superiority. The claim to moral superiority was, I think, a way of justifying the dominant role in government of older shih families under pressure from newer men. Recall also that Ssu-ma wished to leave education in private hands and to admit men to the examinations through recommendation by high officials, and that he stressed the importance of "ethical conduct" in learning. His promotion of the virtue of filial piety illustrates his view that learning should begin in the family, in contrast to Wang's willingness to rely on state schools and his emphasis on the literary side of learning.

This leads me to suggest that grand political visions were being formulated in the eleventh century in response to a struggle between new and old shih , a struggle that arose because the examination system was creating new

shih —many of whom used the idea that learning should guide politics to push for a greater role in government—at the expense of established shih families. The greatest of these visions were Wang An-shih's and Ssu-ma Kuang's. They defined very differently what shih could try to accomplish through government, and they justified their claims with different scholarly methods. Yet both continued to see the shih primarily as those who served, a political elite whose social privileges were justified by their potential service to the state through the institutions of government. And both continued to treat learning for the shih in the context of the shih's political responsibilities.

But by the end of the twelfth century leading scholars no longer shared these assumptions. Indeed, Tao-hsueh , as represented by Chu Hsi, insisted that the only true goal of shih learning was the moral transformation of the individual and his advance toward sagehood. Chu objected to examination learning because it failed to turn students toward the cultivation of a moral self, and because it made the selfish desire to succeed the goal of learning. He worked to establish alternative schools, he envisioned schemes that would allow local shih to take responsibility for local mores and welfare without official supervision, and he gave intellectual interests priority over official duties in his own career. Tao-hsueh represents the acceptance of Ssu-ma Kuang's view of the proper relationship between government and society, yet it no longer defined the shih as those who served in government. It may seem paradoxical that Tao-hsueh advocates also subscribed to the idealization of antiquity promoted by Wang An-shih. But since their vision of an integrated order was not based on the satisfaction of material interests, they could see government from Ssu-ma Kuang's perspective. Instead, they defined the real common interests of men as moral interests. An integrated order would emerge only when men realized in practice the moral principles possessed by all men alike. Integration of government and society did not require government command over society or social transformation, because it would exist when all men behaved morally. The primary responsibility of the shih and the purpose of learning had shifted from politics to morality.

To see why this happened, and why the shih chose Ssu-ma's view of government rather than Wang's, we must turn back to the eleventh century. The examination system attracted ever-larger numbers of candidates, although the absolute number of degrees awarded remained constant. It thus created an ever-larger group of men who by virtue of their participation in the examinations alone—whether they attained a degree or not—thought of themselves as shih . The New Policies exacerbated this by establishing a national school system that regarded all students as shih . By the end of the eleventh century there were about 200,000 registered students, half of whom were competing for about 500 examination degrees in hope of

joining a civil service of possibly 20,000 men (but with probably only half that number of actual posts). The number of candidates continued to increase in Southern Sung. In short, the examinations made it possible for far more men to claim that they were shih , because they had cultivated the learning that made men shih , while denying them the opportunity to serve in government. Under these-circumstances the notion that shih were those, and just those, who served became implausible, and a real gap began to emerge between the great pool of shih and the relatively small group of shih who held rank as officials.

Wang's vision failed to keep the support of the shih because it could not (despite its ambition to do just this) provide them with official careers. Left without the possibility of government service, yet seeing themselves as the elite by virtue of their education, shih had to defend their social status as the elite by making connections outside of government. As shih did this, by establishing their families as the dominant families in local society, by taking up careers as teachers, and by joining intellectual movements such as Tao-hsueh that gave precedence to moral cultivation, their interests as shih tended to lie with an approach to government that did not interfere with the structure of society, shih learning, and private wealth.

Wang An-shih had seen this problem and had worked to keep this larger pool of shih dependent upon government. The idea that policy and teaching, and local government in general, should come from the state's local schools was a solution. Moreover, the idea of using local shih to staff the unranked positions in local government that had practical responsibility for administration (the "clerks") promised in theory to absorb the shih the schools were creating. In the long run, Wang also understood, this required providing full support to students, eliminating the need for private education and private funding, and either abolishing the examinations entirely or limiting access to the examinations to graduates of the schools; for the prefectural qualifying examinations and the custom (new in Sung) of blind examinations continued to allow men to enter the bureaucracy independently of state institutions. But the social division between shih and clerks had apparently become too great for the shih to accept this solution, and the government lacked the funds necessary to pay those student-officials who remained clerks all their lives the kind of salaries shih would find becoming. Thus the New Policies regimes were left with more shih than ever before, whose chances of a career in government were smaller than ever before. Just as the purges of opponents among the old guard kept them from placing their descendants in official positions and pushed them toward consolidating their status in local society, so did the lack of prospects make the new shih increasingly dependent upon their relation to local society to establish or maintain themselves as the best families. Wang's attack on private wealth must also have harmed the families of men who were trying to

establish themselves as shih during what was supposed to be a transitional period, although this would have mattered less if the government had been able to offer all shih an official career.

My conclusion is simple. When the shih found themselves increasingly dependent on local society to maintain the material advantages they thought becoming to men of their status, they concluded that their interests lay with the kind of government that did not try to manage the wealth of society. They had an interest, to be sure, in the government's survival, for its institutions were necessary to their definition of themselves as shih and to the privileges that flowed from being recognized as such. The results would not necessarily have been to Ssu-ma Kuang's liking. One of the conditions for his balance of interests was, after all, a noncommercialized rural economy that tended to limit the social distance between rich and poor. But one may speculate that the ongoing commercialization of the economy, not balanced with checks on the wealthy, was vital to the transformation of the shih from a political to a cultural and social elite. And as the rich became richer, private interests became an even greater threat to the impartial functioning of public institutions, and local shih families made it increasingly difficult for the government to realize the minimal administrative duties Ssu-ma Kuang envisioned.

Without the prospect of official careers, learning became essential for distinguishing the shih from the common people, because it allowed shih to claim moral and cultural superiority over those who were merely wealthy or powerful. Being a shih still required claiming responsibility for the public good. For most shih this was accomplished merely by participating in the examination system, thus proving one's cultural training. For others it could be accomplished independently of government through private learning, such as Tao-hsueh , proving one's idealistic and moral commitment. The examination system thus provided the government with institutional means to ensure that the social elite had a vested interest in the survival of the state that recognized them as shih . The restoration of a Sung-style examination system in Chin, the system of state schools and official teachers in Yuan, and the school and examination system of Ming, with its formal acceptance of Tao-hsueh teachings, all served to establish ties between local elites and the government. The survival of the state had come to depend, just as Ssu-ma thought, upon the government's willingness to accommodate the interests of the existing social elite. But the society on which government was constructed was not as stable as Ssu-ma would have wished.