Premodern Chinese Political Thought

Helen Dunstan. Handbook of Political Theory. Editor: Gerald F Gaus & Chandran Kukathas. 2004. Sage Publication.

Political thought in ancient and imperial China could be studied from two opposite perspectives. An ‘externalist’ agenda would reflect the preoccupations of political theorists outside the Chinese tradition. Given the persistence of Western hegemony in scholarship, the preoccupations of the European heritage would doubtless be privileged, the questions those that would occur to Western intellectual historians. The aim of an ‘internalist’ approach, by contrast, is to see the development of Chinese political thought from the inside. This may mean operating in a conceptual world that is opaque and not necessarily interesting to historians of Western thought. The codes are different, the allusions obscure; one seems to encounter a perversely enduring fixation with the hermeneutics of scraps of ancient text. Yet much premodern Chinese political thought indeed took place within such scriptural and backward-looking frameworks. To understand it, one must attempt to sojourn in Chinese cognitive structures. ‘Internalist’ perspectives are essential.

The dichotomy between the two approaches need not be absolute. While Eurocentrism is perhaps betrayed by utterances that begin ‘What was the nearest Chinese equivalent of…’ (cf L. Liu, 1995: 7), ‘externalist’ questions are more likely to spring from curiosity than arrogance. As stimulants to enquiry, they usefully complement sinologists's research agendas. Indeed, sinologists may ask them. The following discussion, intended for both non-sinologists with basic knowledge of the main schools of Chinese philosophy, and sinologists interested in English-language scholarly developments outside their own fields, draws on both approaches. Where the ‘externalist’ approach predominates—as in the discussion of Chinese ideas on (1) the origin of the state and civil society, and (2) provision for popular participation in government the findings illustrate both its heuristic value and its limitations.

The New Textual Scholarship on Ancient Chinese Political Theory (Late Zhou to Early Han, c. 500 to c. 180 bc)


Scholarship on ancient Chinese political theory has been transformed since 1980 through investigation of recently discovered texts and radical rethinking of the concepts of author, book and text as applied to pre-Qin (antiquity to 221 bc) writing. The silk manuscript versions of the Dao de jing (Tao te ching, hereafter ‘the Laoz’) and the previously unknown ‘Yellow Emperor’ texts found with them in 1973 are only the most famous of the archaeological discoveries that repeatedly jolt our understanding of ancient China's intellectual vitality.1 Recent inventories of known texts reveal a diverse corpus, some of it reflecting the cultural background to the emergence of philosophy, some of it explicitly political or governmental (e.g. Loewe, 1993; Giele, 1998-9: 306-37). Meanwhile, attention to the physical characteristics of ancient Chinese books has stimulated both experimentation with the received arrangement of surviving text and a displacement of authors. The surveys of ancient Chinese philosophy thought standard in the early 1980s (Fung, 1952; Hsiao, 1979) now serve as rich statements of the conventional understanding that was the point of departure for more recent scholarship.

The New Approach to Ancient Text


In the English-speaking world, the harbinger of the new scholarship was Angus Graham, whose supremely perceptive Disputers of the Tao (1989) is priority reading. Perhaps more influential among specialists was his retranslation of the Zhuangzi (Chuangtzu, fourth to second centuries bc), one of the two great classics of what became known as the Daoist (Taoist) school (Graham, 1981). Graham's translation was distinguished from earlier attempts by his greater attention both to the work's generically variegated content and to the probability that parts of the received text are out of sequence. Ancient Chinese philosophical texts were usually written on bamboo strips that were linked with thongs and bundled. Strips usually outlasted thongs. Later editors inherited disintegrating bundles and, where necessary, rearranged the contents into chapters. This might involve juxtaposing generically disparate material, both verse and prose, perhaps including unmarked quotations. Recognizing the range of material in the Zhuangzi's core chapters, Graham made a point of adopting appropriate diction and layout for the different types of text. He transposed passages that seemed out of place and rearranged much of the material in the later chapters by theme or philosophical tendency (1981: 31-2). His reconstructed Zhuangzi is a miscellany of work by multiple authors and of different dates. It represents at least three other tendencies besides the ‘school’ of Zhuangzi and his followers.

Parts of the Zhuangzi are important for the study of ancient Chinese political thought. Graham identified two groups of utopian writings, including a chapter that represents ‘the first documented instance of a true anarchist in China’ (1981: 170) and a set of chapters by a radical anti-moralist whom he called ‘the Primitivist’ (1981: 170-5, 195-217). He clarified the political significance of the ‘egoist’ doctrines of the school of Yang Zhu (Chu), writings from which constitute another section of the Zhuangzi (1981: 223, 219-53). However, Graham's Zhuangzi is relevant here mainly because of its methodological influence on work on other texts. Consensus seems to have emerged that to treat thematic collocations of material like unitary essays is to misrepresent the thought of the original. While it has long been recognized that many ancient works include writings by several hands, there is a new interest in breaking received textual units into their component parts, the dating and authorship of which are the next challenge.

A deeper rationale for this new ‘deconstructionism’ is supplied by Mark Lewis (1999), who draws radical implications from the ‘fluidity’ of texts written on bamboo strips. For at least two centuries from the age of Confucius (fl. c. 500 bc), he suggests, text strips were media through which philosophical schools elaborated and transmitted their doctrines. Possession, custodianship and ongoing creation of the texts helped to define the schools; the doctrines were attributed to a founding ‘master,’ but the texts grew by accretion as successive generations added new material and rearranged the old. To give their pronouncements authority, the later-generation disciples represented them as sayings of the master, a quasi-fiction whose own authority came from the fact that he addressed disciples (or rulers in disciple role). Historical masters, such as Confucius, presumably existed, but the masters whom we perceive through texts were constructs of disciples with their own agendas (1999: 54-8).

The views of Lewis and those on whose work he draws profoundly challenge previous understandings, including some of Graham's. The clearest illustration of this is the fragmentation of Confucius by E. Bruce Brooks and Taeko Brooks (1998).

The Confucian ‘Analects’ of Brooks and Brooks


In his attempt to reconstruct Confucius's intellectual personality, Graham (1989) was unperturbed by the existing consensus that not every dictum preserved in the Analects transcribes the master's words. It was enough that the Analects is ‘a book homogeneous in thought, marked by a strong and individual mind’; one could assume that it represented ‘the earliest stage of Confucianism’ without worrying about its literal authenticity (1989: 10). Graham convincingly portrayed a unitary Confucius, believer in the full efficacy for government of an ethicized tradition of aristocratic ‘ceremony,’ and advocate of rulership through de (te), conventionally translated ‘virtue.’ De had previously meant ‘the power… to move others without exerting physical force,’ but for Confucius it became ‘the capacity [or Potency,’ Graham's preferred rendering] to act according to and bring others to the [moral] Way's (1989: 13-15). If the Brookses are right, this reconstruction is untenable.

In their recent translation, Brooks and Brooks (1998) rearrange the Analects dicta by suggested order of accretion. For them, out of the entire Analects, only 16 of the 26 short utterances in Chapter 4 possibly represent the authentic voice of ‘the historical Confucius’: a ‘mentor’ advising ‘protégés’ on the importance, for would-be courtiers, of maintaining a morality befitting the hereditary nobility (1998: 1, 11, 13-16, 203-4, 208-9). The famous sayings taken to epitomize Confucius's approach to government are typically late additions to a work that took over two centuries to reach its final form. Some of these dicta are from chapters added in the late fourth century bc, in the days of Mencius (Confucian doctrine's most influential ancient expositor). From Chapter 12, which Brooks and Brooks date to c. 326, comes the proposition that rulers, vassals, fathers and sons should all be what their names imply, and the analogy between the influence of the ‘virtue’ (Potency) of the gentleman in government and wind blowing over grass. The even later Chapter 2 includes the declaration that rule by ‘virtue’ and ‘ritual’ has superior effects to that by ‘government’ and punishments, and the analogy between the efficacy of governing ‘by virtue’ and the still force of the ‘North [polar] Star.’ The paradoxical claim that the prehistoric sage emperor Shun did nothing but respectfully face south is from Chapter 15, whose core Brooks and Brooks date to the very end of the fourth century (1998: 89, 92-4, 109-10, 131, 226-30, 234). The Brookses suggest that much in these later chapters reflects, or reacts to, the ideas or concerns of rival schools—the Legalist preoccupation with order, the Daoist promotion of non-action, the Mohist belief in undifferentiating love for others, and a new interest in theorizing the cosmos (1998: 95, 97, 109-10, 137, 226-31).

Many assumptions will have to be rethought if Brooks and Brooks are right. They and others in the Warring States Working Group (a network centred on the University of Massachusetts, Amherst) have proposed yet further conjectures. Prima facie, their reading makes exciting sense, but their methodology is debatable. The scholarly community needs time to reach mature consensus as to the value of their work. The overall approach, already enshrined in The Cambridge History of Ancient China, may win more widespread approval than the details of the Brooks interpretation (compare Nivison, 1999: 745-6, 755-9, with Makeham, 1999: 1-15).

The Xunzi of John Knoblock


Especially important for studies of ancient Chinese political thought is the late John Knoblock's three-volume study-cum-translation of the works of Xunzi (Hsün-tzu: Knoblock, 1988; 1990; 1994). The Confucian Xunzi (c. 310 to c. 215 bc) arouses Western curiosity because some of his ideas seem reminiscent of Hobbes and Locke. Knoblock's integral translation shows how limited the resemblance is. Perusal of the most obviously ‘political’ chapters (9-16) reveals Xunzi as, above all, a relentless moralist. His concern was with urging better ways upon contemporary rulers and refuting the errors of rival schools and some fellow-Confucians. Western readers will escape disappointment with Xunzi only if they accept him on these terms.

Knoblock's study features close textual analysis and reconstruction, explications of content, context and allusions, and a biography of the presumed primary author. The translation's layout reflects the discontinuous structure of a work that may, in part, have been reconstituted from Xunzi's jottings or disciples’ notes. Sometimes ‘Xunzi’ himself is editor, commentator or transmitter of earlier writing (1988: 6, 127-8). Knoblock nonetheless took the historical Xun Kuang as the main intelligence behind the book. His study clarifies how one of China's major political thinkers used pre-existing historical and rhetorical traditions to discuss what it meant to posit morality and ceremony as the only sound bases for government. It also shows Xunzi responding (not always negatively) to the realpolitik concerns and doctrines of the turbulent third century.

Concerns about the translation's accuracy have been raised (Harbsmeier, 1997: 183-95). Indeed, the rendering of key terms, such as fen (basically ‘to divide’), is sometimes overdetermined. For Graham, fencould represent the concept of ‘allotting] portions,’ while in another school's usage, fen (‘portion’) as a noun could signify the duties incurred through one's position relative to others. It is troubling that where Graham understands fen as ‘apportion,’ Knoblock has ‘divide society into classes,’ while Knoblock renders fen zhi (‘divide them’) as ‘create proper social class divisions’ (1990: 96, 104; cf Graham, 1978: 46, 255-6; 1989: 255ff). Without dogmatically endorsing Graham's reading, one should warn against uncritical reliance upon Knoblock.

Legalism, Syncretism, and the Political Wing of Daoism


The retranslation of the writings of the ancient Legalists (political realists) is still in progress. J. J. L. Duyvendak's (1928) pre-war translation of the Shang jun (Shang-chün) shu (Book of Lord Shang) has not been superseded, but a new, annotated translation of the Han Feizi (Han Fei Tzu) by Christoph Harbsmeier is forthcoming with Yale University Press, and Allyn Rickett (1985; 1998) has completed his translation of that partially Legalist miscellany, the Guanzi (Kuan-tzu). Meanwhile, archaeological discoveries have stimulated new work on the syncretism of the late Warring States and early Han, especially the rapprochement between the Laozi (Lao-tzu) wing of Daoism and the Legalists.

The Laozi has itself come under scrutiny. Until recently there was no proof of its existence before the mid-third century bc, but in 1993 bamboo strips with an earlier version of parts of the received text were found in a late-fourth-century (bc) tomb at Guodian in central China. Experts differ as to both the likely ‘date of composition’ of the Guodian text, and the transmission processes through which the familiar version emerged not later than the early second century bc (Allan and Williams, 2000: 118-20, 142-6). The new material may help decide where the truth lies between Graham's intuition of a single authorial personality behind this ‘long philosophical poem or poem cycle,’ and the subsequent hypothesis of a multi-authored work that took ‘almost a century’ from the mid fourth to gain its present length (Graham, 1989: 216-19; Brooks, 1994: 64-6). Such issues of textual chronology are crucial for speculation about inter-school influences. The discovery that ‘about a third’ of the material in the receivedLaozi existed, in a different sequence, in the late fourth century (Allan and Williams, 2000: 128) makes it completely possible that the most sophisticated Legalist, Han Fei (died 233 bc), was influenced substantially by this text.

Graham (1989) found it ‘debatable’ whether Han Fei wrote those Han Feizi chapters that discuss the Laozi or blend its ideas with Legalism. However, if indeed the Laozi ‘presents itself as another guide to the art of rulership,’ there is nothing incongruous about its ‘mystical statecraft’ appealing to a pragmatist with no time for Confucian moral gestures (1989: 170, 285). Both Graham (1989: 286-9) and Wang Hsiao-po and Leo Chang (1986: passim) convincingly portrayed a synthesis in which metaphysical and meditational ideas from the Laozi underpin the ideal of the impartial, non-assertive monarch. Neither imposing his will on reality nor taking personal initiatives, this monarch conducts government as personnel management, administering rewards and punishments according to objective facts and standards. The poetic expositions of this view in the Han Feizi are expressions, whoever wrote them, of a synthetic philosophy that was fashionable by the late third century BC.

Somewhat similar texts, newly accessible through Rickett's translation, appear in the Guanzi in chapters, one of which Rickett dates to the early third century. Despite its title, ‘Clearing the mind’ has more political than meditational content. It expounds the approach of a sage ruler, who is receptive to forces outside himself, such as a mysterious ‘Great Brilliance,’ but otherwise does little but maintain a constant set of laws, name things correctly (to create order), and verify subordinates's accomplishments. Importantly, however, he does engage in warfare (Rickett, 1998: 85-97). Rickett associates this text with a strain of thought that is loosely called ‘Huang-Lao,’ a Han-dynasty term projected back into the Warring States because Sima Qian (Ssu-ma Ch'sien, 145 to c. 86 bc) identified several Warring States thinkers, including Han Fei, as students of Huang-Lao doctrines. ‘Huang’ refers to the Yellow Emperor, mythical ‘inventor of the state and of war,’ and through him to ‘the Legalist strand’ within the synthesis (Graham, 1989: 170-1). This culture hero fittingly represents an intellectual tradition centred on advising rulers on the sociopolitical preconditions of military conquest.

Four texts, discovered in 1973 and identified by some as the lost ‘The Four Scriptures of the Yellow Emperor,’ lend plausibility to the notion of a pre-Qin Huang-Lao movement. Only the second text features the Yellow Emperor, shown consulting his advisers, whose purported replies are reproduced at length. The texts, disparate in content, format and, probably, authorship, are addressed to rulers and reflect a fusion of ‘Daoist’ and ‘Legalist’ ideas, plus elements from other traditions. Particularly intriguing is the advice to adopt ‘feminine [or ‘soft’ or ‘weak’] conduct’ (humility and yielding, coupled with a benevolent disposition) rather than the assertive ‘masculine’ counterpart. However, the ideal is not pacifism but military success (L. Chang and Feng, 1998: 67-70, 163-5, 177-8). Another theme is that the ruler should model his government on the operations of Heaven and Earth, complementing civil governance with resort to force, just as Heaven has seasons for life-giving and life-taking. As the cosmological references suggest imperial pretensions, it is no surprise to find passages envisaging a universal sovereign—a uniquely informed autocrat who values educated men who ‘understand [the] Dao [Way]’ (1998: 33-7, 42, 46-7, 104, 111, 116-20).

Two translators of the four texts opine that they were probably written ‘around 290’ bc (1998: 214). Other scholars argue for a date between the mid-third century and early Han, pointing out, for example, that the breadth of the syncretism resembles that of the ‘philosophical encyclopedia’ Lüshi chunqiu (The Spring and Autumn Annals of Lü Buwei) (Peerenboom, 1993: 18-19; Puett, 2001: 239-40, n. 111). The eclectic Lüshi chunqiu was compiled about two decades before the Qin unification (221 bc). A plea for morality in government that draws little from the Laozi (Graham, 1989: 373-4), it is newly accessible through a bilingual translation volume (Knoblock and Riegel, 2000). Understanding of the four syncretist texts is less definitive. Unreadable in places, the manuscripts are riddled with ‘loan’ characters. Alternative translations remain desirable.

The Origin of Civil Society and the State


Xunzi: Civilization as the Sages' ‘Artifice’


The best-known premodern Chinese theorist of how political organization arose was Xunzi. Xunzi's best-known proposition (that human nature is bad) seems too inconsistent with much else that he wrote to be considered the foundation of his political philosophy, although it is commonly taken as such. For Xunzi, it was the introduction of ‘ceremony’ (Graham's translation) and morality that marked the emergence of civil society. The ‘Former Kings’ (sage monarchs of antiquity) had established ceremony and morality to institute principles of allocation, thus ending the chaos that had prevailed when the people were left to compete for means of satisfying their innate desires without attention to ‘measures’ and ‘boundaries’ (Graham, 1989: 257; Knoblock, 1994: 55). Crediting sages with the invention of specific aspects of culture was a convention in the Warring States, but Xunzi almost dispensed with sages. Elsewhere he located the origin of civil society in that ability, indeed lifelong compulsion, to ‘associate],’ that is the secret of man's dominion over physically superior animals. Sage founders appear here only in the definition of a ‘lord’ as ‘one who is accomplished at causing men to form societies.’ Doing without the myth of sagely ‘artifice,’ Xunzi ascribed moral sense to humankind's distinctively ‘exalted’ nature (translations variously from Knoblock, 1990: 103-5; Graham, 1989: 244, 255; Lewis, 1990: 171-2). He further identified the human emotions, senses and intellect as ‘Heavenly,’ the sage being one who trains to perfection that which Heaven has implanted in him (Puett, 2001: 67-9).

The claim that human nature is bad is consistent with, but not necessarily entailed by, Xunzi's assumption of primeval chaos. The pertinent fact about human nature is that, ‘born with desires,’ man is incapable of not pursuing them. However, the desires need not lead to chaos if controlled through civilization's artifices. Graham wisely suggested that Xunzi's oversimplifying ‘slogan’ about human nature was adopted for debating purposes and should not be mistaken for a fundamental tenet (1989: 250-1). Knoblock's efforts to date the separate chapters of the Xunzi might have clarified the relationship between the ‘slogan’ and the account of man's need for culture. Unfortunately, he changed his views on a key question while producing his three-volume translation (compare Knoblock, 1988: 9-11 with 1994: vii). At present, one can only note that his attempt to date the various materials in the Xunzi raised for the first time the possibility of tracing the chronological development of an ancient Chinese intellectual's political philosophy.

Whether the historical Xunzi really thought human nature bad is less important than how he elaborated the notion that principled apportionment is fundamental to civil society. Knoblock's retranslation of Xunzi's chapter on ceremony shows how inappropriate it would be to see him as a contactarian manqué. His central interest was in the rationale for ritual and ceremony, subjects he discussed in celebratory detail. This was reasonable, as he considered ceremony ultimately more important than military might for enabling rulers to extend their sway while remaining secure at home. Ceremony, embodying differential entitlements and graduated expressions of respect and love, safeguarded order from the lurking threat of ‘anarchy’ (1994: 57-61, 70-1).

The Yellow Emperor and the Origin of the Martial State


Mark Lewis (1990), by contrast, has drawn attention to the martial characteristics of the pre-unification Chinese states as a problem in ancient Chinese political thought. Using disparate texts, he reconstructed the Yellow Emperor myth as an ancient rationalization of the emergence of ‘sanctioned [governmental] violence.’ The Yellow Emperor brought order to a chaotic age by subduing warring nobles, an oppressive Fiery Emperor, and that savage, bestial-looking rebel, Chi You. Thereafter, he ruled as a travelling order-keeper, instituting other prerequisites of civilization such as the calendar. Sima Qian represented him as civilization's original founder (Lewis, 1990: 174-6).

To Lewis, the precedence that Sima gave the Yellow Emperor over more pacific culture heroes reflects elite espousal of a doctrine of legitimate and necessary force. Probing of other layers of the myth reveals that the Yellow Emperor was also lord of storms. Chi You was a rival lord of storms, weapons and warfare. Uncontrolled, he represented brute, anarchic violence. The Yellow Emperor was the originator of organized, cosmically sanctioned, strategically guided warfare, as well as due judicial process. His story, emerging (through elite reinter pretation) in the Warring States, was a mythological representation of those social, political and military transformations whose theorists were the Legalists and Strategists. It eventually ‘became a charter myth for the absolutist state’ (1990: 176-84, 195-212).

Michael Puett (2001) criticizes Lewis's presumption of a single myth complex. For Puett, the rewritings of the Yellow Emperor story by exponents of rival Warring States philosophies should be analysed as conflicting accounts of the origin of punishment, warfare and the force-using state. Puett posits the establishment of three narrative patterns and associated messages before the Yellow Emperor entered the debate in the third century bc. Pattern One, first found in a Confucian text, displaces the responsibility for creating punishments (or weapons) onto barbarians or reprobates, and has the civilizing powers to ‘appropriate’ these tools. In Pattern Two, there is nothing ethically problematic about the invention of weapons and punishments; the issue is how they are used. Pattern Three evades the problem of creation by envisaging the sage as ‘organizer.’ Rather than imposing novelty on nature, the organizing sage brings out the order inherent therein (2001: 101-11).

In Warring States stories that include the Yellow Emperor and/or Chi You, the assignment of characters to roles reflects the author's favoured pattern. The rebel figure is not always Chi You, who may be minister, not rebel. Chi You as rebel originator of violence is paired with an ‘appropriating’ and/or ‘organizing’ Yellow Emperor, Chi You as minister with a ‘creating’ Yellow Emperor (2001: 120-7, 131-3). Chi You's multivalency illustrates the richness not of ancient Chinese myth but of the rhetorical strategies used to debate the morality of government's coercive aspects. The issue of origins is, after all, not central: origins feature as symbols of moral status.

Other Ideas about the Origin of Culture and the State


Puett (2001) relocates Xunzi's speculations in the context of the Warring States assumption that civilization's material, organizational and ceremonial constituents began as the creations of specific individuals (although the establishment of rulers might be attributed to Heaven). Opining that whether sages should ‘create’ was a contentious issue in pre-Qin times; he discusses references to acts of cultural creation as reiterations, variants or hybrids of three basic positions. The Confucian view, adumbrated in the Analects and developed by Xunzi, was that the sages's innovations had been ‘patterned’ upon Heaven; the sages had only ‘brought forth’ the constituents of culture, thereby completing the generative processes of the natural world. Mohists validated cultural creation—for them, the invention of useful techniques and artefacts as having imitated Heaven's creative acts; Daoistic texts reject it as ‘transgress[ive]’ against nature (2001: 44-55, 62-3, 68-73). Thus the (Confucian) ‘Xici’ (Hsi-tz'su) appendix to the Book of Changes represents institutions, techniques and implements as having been derived, at one remove, from nature. These innovations of the sages were inspired by the trigrams and hexagrams, which were themselves sage-made abstractions from nature's patterns (2001: 86-90). In partial contrast, three Lüshi chunqiu chapters combine the Mohist appreciation of useful inventions with the Daoist viewpoint that the sage does not impose himself on nature. Past sages had succeeded by leaving creation to their able ministers, thus freeing themselves to cultivate the stillness through which they identified with Heaven (2001: 81-6).

The Legalist Shang jun shu represents the state's emergence in secular, developmental terms, positing a sequence of approaches to the problem of disorder. While accepting the sagely origin of mankind's means of quelling chaos, it grounds the need for sagely intervention in population growth. Ancient people started quarrelling because population was outstripping resources. Impartiality and disinterestedness began as society's response to its first age of disorder, when the old kin-based groups learned the disadvantages of pursuing self-interest by force. Unfortunately, instituting the norms of justice meant ‘elevating’ men of superior ability, and this, combined with further population growth, engendered a new phase of disorder, marked by competition among the able. Sages then moved to establish the state, instituting demarcations of landed property and between the sexes, ‘prohibitions,’ ‘officials,’ and, finally, ‘a ruler.’ Only now was hierarchy established as the fundamental principle of order. While this account was offered to justify the principle of institutional and legislative innovation, it created a prototype for historicizing explanations of contemporary political arrangements (Fung, 1952: 315; Graham, 1989: 271-2; Y. Liu, 1998: 177-80).

How was the legacy of pre-Qin thought about the early history of human institutions developed in the imperial period (221 bc to 1911)? Hoyt Tillman's (1994) work on the mid-imperial ‘utilitarian’ Confucian Chen Liang (1143-94) illustrates the interest of this under-researched question. Influenced by Xunzi, Chen diverged from him on the origin of rites and ceremony, tracing the latter to norms implanted in human hearts by Heaven (1994: 32). Chen had his own theory both of the origin of civil society, and of the emergence and morality of hereditary monarchy, which he discussed in terms of the polarity between public spirit, gong (kung), and self-interest, si (ssu). In the earliest times of greatest public spirit, the non-hereditary rulers were chosen by the communities they ruled. Moral deterioration led to formalization in the governmental structure; thus the Confucian culture heroes Yao and Shun selected their successors (while refraining from appointing their own sons). By the time of the fully historical dynasties, the hereditary principle had long been established, but public spirit remained necessary for dynastic success. It was just that the founders of successful dynasties could not match the superb public spirit of the prehistoric rulers (1994: 34-7).

Popular Participation in Government, Elite Criticism, and the Notion of a Chinese ‘Liberalism’


That Chen could imagine a utopian antiquity in which communities chose their rulers suggests that it need not be Eurocentric to adopt ‘Chinese ideas on popular political participation’ as a research topic. Admittedly, there remains the danger of reading Western concepts into Chinese writings—as illustrated by the treatment of Mencius in a long-influential textbook. Here, the notion of a popular ‘right of revolution’ is discerned in Mencius's claims that a righteous conqueror will be welcomed as liberator by entire peoples, that a vassal who kills a reprobate ruler need not be said to have murdered his lord, and that the legitimacy of a new overlord is manifested through popular acceptance (de Bary, Chan and Watson, 1960: 87, 95-7). As Graham pointed out, however, in Mencius it is Heaven and distinguished nobles who appoint and depose rulers, the people being little more than Heaven's mouthpieces (1989: 115-17). The Chinese rebel commoner could appeal for justification only to the righteousness of Heaven, not the rights of man; he could legitimately expect government for the people, but sovereignty was a matter between Heaven and the ruler. The new edition of this textbook sets the record straight (de Bary and Bloom, 1999: 124).

Some ancient Chinese authors did nonetheless rhetorically envisage an autonomous political role for commoners. This was as junior members of remonstrance hierarchies: schematically defined sets of people from whom wise rulers accepted feedback on their governance. These hierarchies-expounded in speeches recorded in the pre-Qin works Zuo zhuan (Tso chuan) and Guo yu (Kuo yü) probably reflect fourth-century (bc) political ideas. They are conveniently assembled in David Schaberg's study of the rhetorical structure of remonstrance speeches, texts that took the past as principal source of authority and used ‘inherited’ material to make their point effectively (1997: 135-7, 140-2; cf Schaberg, 2001).

One speech claims that Heaven, having created the people and established rulers for them, gave the rulers helpers ‘to serve as [their] teachers and protectors and to keep [them] from exceeding proper measures’ (1997: 144). The rulers' helpers (loyal critics) comprise the entire social order. Scribes and blind musicians give criticisms literary clothing, performers ‘recite their remonstrances,’ senior administrators ‘correct and instruct,’ knights ‘pass on words,’ and ordinary people murmur in the marketplace or work criticisms into their manufactures. Another speech warns that, after perfecting their moral potency, the ancient kings ‘listened to the people,’ soliciting remonstrances and poems from their officers. They heeded popular ditties, the gossip of the marketplace, and evaluations of their governance ‘along the roads.’ A third text likens the people's words to flowing water: blocked, either may burst out disastrously, but prudently drawn forth, either can enrich the kingdom (1997: 143-8).

The ideal of remonstrance, supposedly endorsed by Confucius, found institutional embodiment in later Chinese history in the form of the censorate, a branch of government whose functions included loyal criticism of the ruler. However, institutional provision for the upward flow of popular opinion was usually confined to token gestures. Similarly, later discussion of public opinion's importance generally focused on the contribution that members of the Confucian-educated elite could make from outside the bureaucracy, if granted a respectful hearing. Thus, when the private intellectual Fang Dongshu (1772-1851) alluded to the admonition about the people's words resembling water, the context was his vindication of political debate by educated men of principle who did not currently hold power. There was a continuum between jiangxue (seminar-style) exploration of issues in moral philosophy and morally inspired discussion of ‘the evils of the day.’ Both were necessary, both could enlighten the ruler, and the natural home of both was the academy (de Bary, 1991: 80-5).

Huang Zongxi and Confucian ‘Liberalism’


The intellectual best known for his supposed advocacy of jiangxue-style political debate is Huang Zongxi (Huang Tsung-hsi, 1610-95), whose loyalty to the Ming dynasty (1368-1644) precluded his taking office under the succeeding Qing (Ch'sing, 1644-1911). Huang's status in modern Chinese nationalist historiography, as China's greatest proto-democrat, rests on passages in his rhetorical critique of despotism, the Mingyi daifang lu (1663). Lynn Struve has warned against overestimating this work's significance, for many of Huang's ideas had been anticipated in the critical writings, reform proposals and scholarly practice of the previous generation of Confucian dissidents (1988: 475-9). However, the recent publication of Wm. Theodore de Bary's (1993a) book-length study and translation of Huang's tract presumably represents a claim for its importance. Its centrality in de Bary's earlier (1983) attempt to delineate a ‘Confucian liberalism’ raises interesting issues.

Two passages in particular encourage the view of Huang as proto-democrat. In one, Huang deplored the role reversal that he claimed had taken place since the ancient era when ‘the [people of the] empire’ had been recognized as the realm's proprietors, and their rulers merely as retainers. Elsewhere, Huang suggested that in antiquity, schools had provided moral guidance to the ruler, for in those days ‘even the Son of Heaven [emperor] did not dare to decide right and wrong for himself, but shared with the schools the determination of right and wrong’ (1993a: 104). Huang proposed the restoration of this ancient ideal through adaptation of more recent institutions. The emperor and key high officials should attend monthly seminars conducted by the Chancellor of the Imperial Academy, who would be empowered to address the emperor frankly on flaws in his governance. In the provinces, the renowned scholars appointed, after public discussion, as directors of education would hold twice-monthly seminars, with the local officials in studential role. Thus could minor flaws in the officials's governance be reproved, and major flaws denounced to the beating of drums. The local degree candidates would be empowered to repudiate collectively a director whose morality had provoked adverse discussion, and the assembled elders to give the local officials feedback and advice at periodic drinking ceremonies (Z. Huang, 1663: 2, 11-13; de Bary, 1993a: 92, 106-8).

This is less than a blueprint for democracy. While the metaphor of society as proprietor (literally, ‘host’) is powerful, the context is mere polemic against the historical emperors's usurpation of the people's role. Huang does not say explicitly that current governmental policies would be subject to open seminar-style discussion, or that local literati would have powers beyond participation in denunciations. If, then, the notion of Huang as a proto-democrat is problematic, what of de Bary's ‘Confucian liberalism’ thesis as applied to Huang? De Bary adopted the word ‘liberal’ to posit continuity between the great Neo-Confucians of the Song (Sung) dynasty (960-1279) and Huang. This ‘liberalism’ featured both rejection of conservative rigidity in favour of humanist political reformism, and an ‘individualism’ that stressed personal responsibility for internalizing moral doctrine (1983: 5-9). The intellectual independence achieved through broad learning would give one a basis for reformist activism (or criticism) in the political sphere.

In suggesting that, in the Mingyi daifang lu, Huang ‘advanced Neo-Confucian liberal thought,’ de Bary (1983: 85) probably referred both to the potential of Huang's proposed educational reforms to nurture larger numbers of independent-minded moral persons, and to his ideas for change in governmental institutions. These themes were linked, for institutional change would be required to liberate the full political potential of the moral individual. The ‘unlawful laws’ of despotism must yield to a more fundamental kind of law intended ‘to protect and promote impartially the general interests of mankind’; the reformed schools would be ‘institutions through which a broader, more informed public could participate in the political process’; the seminars would ‘provide a firm institutional basis’ for ‘open discussion of public and intellectual issues,’ thereby ‘[bringing] to a climax’ the Neo-Confucian tradition of discussion as a pedagogical technique (1983: 81-8). Huang was, however, distinctly illiberal on such issues as whether others might dress as they liked, enjoy fiction and drama, buy ‘useless’ objects, or embrace un-Confucian doctrines (de Bary, 1993a: 106-7, 109-10, 159-60). Could ‘something like’ Huang's positive concept of law ‘have provided a framework for what we call today “human rights”’ (1983: 85, 89)? Huang's discussion of law (1993a: 97-9) is largely polemical and shows no interest in procedure. Where were the rights to ‘due process’ of educational directors whose private lives upset some local people, or county officials whose controversial actions could be dubbed major flaws, fit for denunciation to the sound of drums?

The claims advanced in de Bary's (1993a) study of the Mingyi daifang lu are bolder yet. Huang ‘intends that [schools] should perform much the same purpose as political parties or parliaments’; he ‘can reasonably qualify as a constitutionalist, albeit a Confucian one’; the constructive proposals in the Mingyi daifang lu arguably comprise ‘a kind of Confucian constitution’ (1993a: 56, 63, 68). De Bary both expounds the differences between Huang's ‘constitutionalism’ and the Western liberal-democratic type, and draws inspiring lessons from the Mingyi daifang lu for a contemporary China that has left Maoism behind but not embraced democracy (1993a: 69-71). It is nonetheless hard to discern the careful prescriptiveness of modern constitutions in the rhetoric of the Mingyi daifang lu. The work is better viewed as a provocative attempt to rethink the design of China's polity, including its military and economic systems. Huang's institutional proposals are better compared with those in other seventeenth-century works, whether ‘utopias’ such as Wang Yuan's Pingshu, or essay collections such as Tang Zhen's and Gu Yanwu's.

A Chinese Constitutionalism at the Dawn of China's Modern Age?


More convincing, yet claiming less, is Philip Kuhn's (2002) discussion of the thought of Wei Yuan (1794-1857) in terms of constitutionalism. Wei's collected essays date from the transitional period before the collapse of the imperial system, but after the First Opium War. His political thought (as represented by Kuhn) borrows an ancient format comment on the pre-Confucian Book of Odes to advocate remedies for the ineffective autocracy established by late eighteenth-century misrule. In advocating broader political participation to strengthen the state, he inadvertently foreshadowed a key difference between much modern Chinese democratic thought and the Western preoccupation with limiting the power of the executive (2002: 32, 47-53; cf. Nathan, 1986: 45-66).

Politically conscious literati of Wei's day were frustrated by the undersupply of bureaucratic posts relative to the numbers qualified to fill them, the inherited tendency to view private political associations as unprincipled factions, and the lack of openness to outsiders' suggestions for improving government. Wei's ‘constitutional’ opinions reflected his position as a graduate inhabiting the margins of the political establishment, although exceptionally knowledgeable on public affairs. Concerned to redraw the ‘boundary’ delineating ‘that part of the community that properly participates in national politics,’ Wei resembled Huang in advocating some role for the Confucian-educated elite and envisaging urban academic institutions as fit venues for ‘discussions of ideas’ (2002: 27, 42). Wei, however, would have had consultation confined to those non-office-holding literati with solid academic qualifications. He interpreted an image of deer calling to each other as referring to the need for elite political discussion, and read the line ‘I shall seek everywhere for information and advice’ as encouraging ‘a broad search for policy opinion.’ He recognized the importance of rulers having divergent views from which to choose (2002: 39-44). Kuhn's study indicates the need for an annotated translation of Wei's political essays, permitting his ‘constitutionalist’ remarks to be assessed in context.

Scholarship as Ersatz Political Participation


Even granted that the Odes had canonical status for Confucians, how could Wei read ‘constitutional’ lessons into these diverse poems? Wei was affiliated with the New Text school, a revived Han-dynasty hermeneutic tradition that sought hidden messages in texts whose surface meaning suggested no particular moral or political intent. For Wei, interpreting the Odes should transcend mere scholasticism to inspire reformist moral action (Kuhn, 2002: 34-9). Better known for their reinter pretations by New Text adherents are the ultraconcise Chunqiu (Ch'sun ch'siu, Spring and Autumn) annals covering the years 722-481 bc from the perspective of the ducal court of Lu (Confucius's native state). Confucius, the supposed compiler, was thought to have conveyed judgements on the events recorded through subtle vocabulary choices. One more radical interpretation, exemplified in the Yuan dynasty (1272-1368), was to read the Chunqiu as a manual of statecraft, or even penal law (Langlois, 1982). Another—the New Text school's approached the Chunqiu through its most visionary early commentary, the Gongyang (Kung-yang) Tradition.

The Gongyang understanding of Confucius is best known through its reshaping by Kang Youwei (K'sang Yu-wei, 1858-1927). Kang represented Confucius as an ‘institutional reformer’ who would have recognized the need for constitutional monarchy had he lived in Kang's day. Kang took a Han-dynasty Gongyangite three-stage theory of history, originally intended to apply to the era covered by the Chunqiu, and elevated it into a doctrine of global political development. The world was passing from an ‘age of disorder,’ heyday of absolute monarchy, to one of ‘approaching peace’ and constitutional monarchy, whence it would eventually enter an age of ‘universal peace’ and republicanism (H. Chang, 1980: 287-8; Fung, 1953: 81-5).

Benjamin Elman (1990) has investigated Kang's intellectual precursors in the late-eighteenth-century New Text school of Changzhou, Jiangsu province. The first was Zhuang Cunyu (1719-88), an educational official and Grand Secretariat academician who briefly held vice-ministerial office at the time when the emperor's favourite Heshen (Ho-shen) was consolidating his infamous ascendancy. Elman views Zhuang's Gongyang style interpretation of the Chunqiu as an oblique lamentation of Heshen's rise. In advocating the revival of a moralistic reading of the Chunqiu, Zhuang expressed his opposition to changes that he could not fight directly (1990: 108-16, 171-85).

It was Zhuang's grandson Liu Fenglu (1776-1829) who, with his followers, supplies the missing link between Zhuang's turn to the Gongyang commentary and Kang's invention of Confucius as utopian reformer. Several years before attaining governmental office, Liu was working to restore the interpretations of He Xiu (Ho Hsiu, ad 129-82), exponent of the three-stage periodization of the Chunqiu era and other Gongyanginspired notions. For He, as explicated by Liu, the ‘uncrowned king’ Confucius had a ‘mandate to establish institutions’ in Lu, the putative future springboard for a new dynastic order. To avoid presumptuousness, Confucius worked his governmental ‘models’ into the Lu court annals, thereby ‘providing] lessons for ten thousand generations.’ His historiography offered inspiration for ‘epochal change’—ultimately, ‘great unification,’ a concept echoed in the name of Kang Youwei's utopia (1990: 233-4, 240, 255; H. Chang, 1980: 288-9).

Elman rightly stresses the Confucian scriptures' importance as the ideological mainstay of the imperial government establishment (1990: 74-5). However, it was the tragedy of both Kang and the so-called historical Confucius to be political outsiders. Liu's notion of Confucius as a Heaven-appointed prophet (Elman, 1990: 231) seems fundamentally an outsider's fantasy. Would systematic study of the sociology of Qing New Text Confucianism confirm that the image of Confucius as prophet and reformer appealed chiefly to politically powerless and disempowered literati? New Text Confucianism may have been most significant as an ideology of scholars who did not hold office or had little power within it—although adherents who joined the bureaucracy might apply New Text perspectives to official business (1990: 215-18).

Mainstream Confucianism and the Imperial State


What of Confucianism in power: the main overt ideology of government for most of the imperial age? What, in particular, of the Song reinterpretations and elaborations that are conventionally regarded, under the name ‘Neo-Confucianism,’ as the official orthodoxy of the remaining dynasties? For James Liu, the introspective self-cultivation urged by Neo-Confucian moralism implied a retreat from engagement with external reality that was partially responsible for the lack of dynamism and creative change in China's subsequent development (1988: 9-11, 149-53). In Ray Huang's brilliant critique of the Confucian political order in Ming China, the thought of the great synthesizer Zhu Xi (Chu Hsi, 1130-1200) is lampooned for ‘committing] every literate person within the empire to a lifetime of study whose only purpose was to affirm that the world is organic and that he was bound by law of nature to perform his assigned duties in society’ (1981: 204). Other Chinese scholars, however, affirm Neo-Confucianism's continuing validity and positive role in modernized East Asia. Tu Weiming has even tentatively anticipated a ‘third epoch of the Confucian Way’ (1993: 214-22).

Acceptance that a long-established ideology of moral governance and social (or sociocosmic) harmony has lessons for the modern age inspires volumes with such titles as Confucianism and Human Rights orConfucianism and Ecology (de Bary and Tu, 1998; Tucker and Berthrong, 1998). Thoughtful, rewarding essays have been written on these themes (e.g. W. Chang, 1998; Twiss, 1998). Yet the scholarly priority remains accurate understanding of the content, scope and functions of Confucian ideology while it still underpinned the Chinese polity (until about 1905). It would be an outrageous exaggeration to call the Confucian tradition itself stagnant over two millennia. The major changes and subtle refinements are explored in numerous publications, of which Peter Bol's (1992) study of the early evolution of Neo-Confucian moral doctrine is a magisterial example.

Neo-Confucianism as Call for the Rule of the Moral Mind


Neo-Confucianism is indeed best known for its Buddhist-influenced emphasis on self-cultivation and the systematic metaphysical and cosmological speculation of some of its founders (Fung, 1953: chs 10-14). However, self-cultivation within the governing elite was to serve a larger purpose. The concept of moral self-development was central to a reasserted political idealism reflected in the chapters on government in the Neo-Confucian anthology of Zhu Xi and Lü Zuqian (1137-81) (Chan, 1967: 202-59). Here we find Cheng Yi (Ch'seng I, 1033-1107) asserting that, fundamentally, the ‘way of government’ is ‘nothing but “rectifying what is wrong in the ruler's mind” and “rectifying one's mind in order to rectify the minds of”’ other officials, starting with those at court (1967: 213). Still more revealing is Cheng's ideal of the sage's mirror-like heart-mind that objectively identifies the good or evil confronting it, responds with the appropriate emotion and action, and yet remains detached (Fung, 1953: 525). A Neo-Confucian paragon would conduct administration in precisely such a spirit.

For a sympathetic exploration of Neo-Confucian political moralism, one may consult de Bary's studies of the thought of Zhen Dexiu (Chen Te-hsiu, 1178-1235) (de Bary, 1981: 67-126; de Bary, 1993b). De Bary analysed Zhen's Canonical Writings on the Heart-Mind and Extended Meaning of the Great Learning as culminations of the Neo-Confucian insistence, articulated in an earlier Song didactic tradition called the ‘Learning of the Emperors and Kings,’ on the ruler's self-cultivation as key to sound government. Zhen's Extended Meaning, an elaboration of his lectures as court scriptural expositor, was reportedly accepted as ‘a guide and model for the ruler’ (1981: 87). Its subject, the ‘Great Learning’ (a chapter of the ancient Book of Rites), contains the locus classicus for the doctrine that attaining governmental order begins with the ruler's moral and intellectual self-discipline (Graham, 1989: 132-4). Together with the Canonical Writings, which took an extreme position against human desires, theExtended Meaning offered the emperor learned advice on rectifying his thoughts and conducting his personal life. Supported by quotations from other authoritative texts and reinforced by historical examples, the central message was: ‘Though the four seas are vast, if the ruler's mind-and-heart are rectified, there is order; if not, there is disorder’ (de Bary, 1981: 115-16).

James Liu and Peter Bol have proposed different explanations for Neo-Confucian moralism's success in implanting itself in China's intellectual culture. Liu (1988) related the process by which Cheng Yi's and Zhu Xi's thought became state orthodoxy to Southern Song (1127-1279) dynastic politics, arguing that the espousal of Neo-Confucianism was a matter initially of political convenience, later of national defence. He represented the focus on the ruler's mind, from Zhu Xi on, as a rational response to the strengthening of imperial autocracy that he saw as a key trend in the early Southern Song (1988: 104, 146-8). This reverses the conventional assumption that Neo-Confucianism fostered autocracy, a view further challenged by Alan Wood in his (1995) study of Northern Song (960-1126) commentaries on the Chunqiu. But the Neo-Confucian call to perfect the self in order to transform society was not intended for rulers alone. Bol (1992) has considered Neo-Confucian self-cultivation's appeal to a large literati class, many of whom might never reach bureaucratic office, still less political power. As the growth of the civil service examination system progressively weakened the links between birth and government position, the educated elite needed a new source of esteem with which to validate their status and identity as shi (shih, scholars) (1992: 330-42). Neo-Confucian study and practice thus achieved a broadly based tenacity.

Neo-Confucianism Displaced: Indigenous and Manchu Challenges


Pamela Crossley and Benjamin Elman have highlighted the inadequacy of the assumption that ‘Cheng-Zhu orthodoxy’ provided the ideological framework of all post-Song imperial government. Crossley (1999) has emphasized constructed notions of ethnic identity in the changing political ideology of the Qing-dynasty Manchu rulers, who governed a growing multiethnic empire of which the Chinese world was only part. She discerns a transition from the ‘transformationalist’ ethos of the early Qing reigns, whose emperors increasingly represented themselves as aliens qualified to rule by their conversion to Confucian values and techniques, to a ‘universalist’ phase during the Qianlong reign (1735-96). Universalist ideology elevated the emperor as transcendent source of wisdom and authority; complementing his ‘culturally null’ ‘capacity to contain worlds’ were the ‘essentialist identities’ assigned to the peoples of his realm, who were to be controlled through definition (1999: 28, 38, 221 and Part 3).

This universalism, which borrowed the Buddhist symbol of the ‘wheel-turning king,’ deployed Confucianism without being Confucian. While the supplanted ‘transformationalism’ was not specifically Neo-Confucian, a central prop of the newly elevated emperorship affirmed a Han-dynasty contention that Zhu Xi had opposed. For Zheng Xuan (Cheng Hsüan, 127-200), only the ruler could attain the heights of moral efficacy, thus becoming Heaven and Earth's associate in exercising ‘transforming and nurturing powers.’ To Zhu, the scriptural passage that Zheng had so interpreted referred to sages, a category that Zhu considered open to all who cultivated moral prowess. Under Qianlong, Zhu and his chosen scriptures remained enshrined in the examination curriculum and much imperial rhetoric, but the court preferred Zheng's position as to who could fully embody ‘moral mind’ (1999: 225, 229-32).

Elman (1994) has argued that the Cheng-Zhu hold on the Qing examination system was compromised by the philological research on the Confucian scriptures known as ‘Han learning’ or kaozheng (k'sao-cheng, ‘evidential’) scholarship. If texts on which Song Neo-Confucian moralism had relied could be exposed as forgeries, what justified the requirement that candidates reproduce the orthodox interpretations? The system adjusted slowly to such challenges, but from c. 1770 on, examiners began to include questions on the problems that the new research had raised. Questions reflecting Cheng-Zhu orthodoxy still predominated, and the incorporation of Han learning into the examinations perhaps blunted its challenge. Nonetheless, ‘the straitjacket of Neo-Confucian orthodoxy’ is now a contested concept (1994: 133-43).

Confucian Statecraft and Political Economy


The notion of Confucianism as straitjacket has been further undermined with respect to Chinese statecraft and political economy. ‘Statecraft’ is the standard translation of jingshi (ching-shih), a term that can embrace what Chang Hao has termed ‘moral statesmanship’—Neo-Confucian self-perfection as the means to ordering society (1974: 38-46). In English-language scholarship, however, ‘statecraft’ commonly refers to writing about practical approaches. Such writing characteristically propounds schemes for institutional, administrative, or fiscal improvement, theoretical rationales being relegated to the preambles. One can use such documents as source materials for Chinese political thought by analysing the authors's assumptions and their allusions to classic texts. However, seekers of explicit political theory may find the conscientious thinking-through of practicalities that typifies these documents an unfruitful distraction.

Statecraft writings are nonetheless often valuable sources for opinions on ultimately political issues, such as social inequality and the stance of government towards it. The researcher must know some codes. For example, discussions about restoration of the ancient ‘well-field’ system are probably about agrarian inequality and reflect views on the state-society relationship. Below, I introduce recent scholarship on selected topics in which the state's role is an important theme. However, a prerequisite for discussion of Confucian statecraft is an introduction to the ancient Guanzi book on which it sometimes drew.

The Guanzi as a Manual of Statecraft


The Guanzi reflects the sophistication of ancient Chinese ideas of political economy. Its governmental sections had the Legalist virtue of being about method but lacked the taint of close association with the state of Qin. The work purported to be by Guan Zhong (Kuan Chung), a seventh-century chief minister of Qi supposedly praised by Confucius. While its advice to rulers seems largely pragmatic, it is not devoid of notions that Confucians could have welcomed. In Graham's view, the Guanzi ‘gives both morality and law places in the organization of the state, in proportions not very different from the Confucian [Xunzi]'s’ (1989: 268).

A key set of statecraft chapters in the Guanzi is that on ‘[The Art of] Light and Heavy’ (qing zhong), or economic management. Political morality, pace Graham, seems conspicuous by its absence from these chapters. They provide the theoretical rationale for that classic form of state economic intervention in premodern China, the maintenance of ‘ever-normal’ granaries for stabilizing grain prices in times of glut or shortage. This practice was usually represented as an expression of Confucian paternalism towards both growers (after harvest) and consumers (in the lean pre-harvest season). However, so redolent was theGuanzi account of cynical manipulation that it was the sanitized version in the Former Han dynastic history that became the locus classicus.

The tone of the ‘Qing zhong’ section probably reflects the context of its composition. Guanzi chapters may date from any century from the fourth to the first bc. Among various suggestions regarding the ‘Qing zhong’ section's date of composition, Rickett (1998) favours the view that its schemes for state enrichment are linked with the economic policies of the expansionist Han emperor, Wudi (Wu-ti, reigned 141-87 bc). Fiscal problems drove Wudi's court into a controversial series of government monopolies and other state trading operations. Rickett endorses the hypothesis that the ‘Qing zhong’ texts provided blueprints for these experiments. However, he qualifies previous findings through careful attention to the dating of separate chapters as well as differences of style, format and authorship throughout the section (1998: 345-57). He does not raise the possibility that those dialogues that propose gross or far-fetched schemes for exploiting society are not ‘promonopolistic propaganda’ (1998: 360), but caricatures intended to discredit unprincipled fiscality.

The ‘light’ and ‘heavy’ of ‘Qing zhong’ refer to value. The ruler should be able to change the value of commodities or means of exchange by manipulating supply and demand. The classic qing zhongoperation involved playing the speculator's game in order to quell private profiteering. State agents bought grain while it was ‘light’ (cheap) and sold it when it became ‘heavy’ (dear). Although this technique later became a tool of public welfare policy, the major Guanzi exposition mentions three goals: price stabilization, ‘tenfold’ profit for the ruler, and the curbing of private accumulations of commercial wealth, which sap the ruler's influence. The ‘Qing zhong’ section advocates rural credit schemes that exploit seasonal price fluctuations to the treasury's advantage, suggests government monopolies and other trading operations as ways of introducing indirect taxation, and explains how to manipulate the relative value of money and commodities so that all types of transaction between state and subjects benefit the former. Set in the multi-state world of Guan Zhong's day, it warns against market forces emanating from rival states. It even advocates a kind of mercantilism, with grain, rather than bullion, as the object of interstate competition (1998: 338-44, 378-84 and e.g. 362-75, 390-8, 411).

One ‘Qing zhong’ passage observes that ‘When the prince's demands are pressing, the value of gold increases. When they are relaxed, it decreases’ (1998: 425). Another mentions the legend that the founders of the ancient Xia and Shang dynasties cast coin from mountain ores to relieve famine (1998: 397). Such notions are frequently invoked in later statecraft writings. What issues do they raise about Chinese monetary theory and ideas of sovereignty?

Chinese Monetary Theory: The Political Implications of Cartalism


Once a state provides means of exchange, discussions of what money is reflect particular conceptions of state power. Richard von Glahn (1996) has argued that the story of money's creation as a famine relief tool contributed to the establishment of a form of cartalism as the dominant strand in Chinese monetary theory. Cartalism is the doctrine that money is deliberately created by a ‘monetary authority’ (here, the sovereign or state), it being this authority that determines money's nominal value, which in principle is arbitrary. The notion of money as a consciously created thing fitted well with the ancient Chinese tendency to conceive of the constituents of civilization as the inventions of sages (1996: 23, 25-8). Although the Guanzi ‘Qing zhong’ section introduced a quantity theory of money by which the (real) value of money was determined by its supply relative to that of commodities, the section's emphasis was on techniques for manipulating exchange values. The ruler could enhance the value of money by restricting its supply, or affect the value of specific currencies by adjusting his demand for them. He should be aware of the impact of his fiscal practice on the value of money, ‘arrogate to himself sole authority over the ratios of exchange,’ and deploy consummate skill in managing those ratios (1996: 29-33).

In imperial times, monetary thought and practice perforce recognized the limitations on the sovereign's power to determine the value of money by fiat. Premodern discussions reflect a ‘compromise between theoretical cartalism and practical metalism,’ the latter meaning a pragmatic understanding that the market value of monetary metals could not be ignored (1996: 34). Von Glahn traces the interplay of cartalism, practical metalism and occasional ‘catallacticism’ (a monetary theory stressing exchange) in an important survey of premodern Chinese ‘monetary analysis’ (1996: ch. 1). However, fundamentalist cartalist rhetoric survived into the seventeenth century, when an advocate of token coins declaimed: ‘The power of the ruler over men is such that he transforms the myriad things. If in an instant he can change the value of a man, can he not also in an instant change the value of a thing?’ (Gao Heng in He and Wei, 1827: ch. 53, pp. 21a-b). The increasing monetary use, beginning in the fifteenth century, of unminted, generally imported silver, and the resulting development of an asymmetrical bimetallism meanwhile created a new context for the Guanzi's advice that rulers manipulate exchange values. Guanzi-inspired arguments that the state should increase popular esteem for coin by demanding it in tax payments gained a new, anti-silver, meaning. Silver was an affront to cartalism: not instituted, as currency should be, by the sovereign, it weakened his grasp on the controls of the economy (e.g. Ren Yuanxiang in He and Wei, 1827: ch. 29, p. 12a; ch. 53, p. 12a).

The Tang poet Bo Juyi (Po Chü-i, 772-846) wrote in an examination essay that ‘One who reigns as king will level dear and cheap and adjust light and heavy, causing the hundred commodities to flow in every part, the people of the four directions to know mutuality of interest’ (quoted in He and Wei, 1827: ch. 53, p. 13b). The slogans of Chinese cartalism did not belong to an isolated realm of discourse, but should be addressed in the historiography of premodern Chinese conceptions of sovereignty.

The Debate over the Statist Policies of Wang Anshi


Premodern China's bitterest debate about how large and activist the state should be was that provoked by the ‘New Policies’ (1069-73) of the chief minister Wang Anshi (An-shih, 1021-86). Confronting high defence expenditures, Wang proceeded on the unconventional assumption that one could ‘create wealth.’ The prerequisite to enriching the state was to help society enrich itself, using a necessarily expanded civil service. State and society should indeed form a single body. Key parts of the economic programme would involve displacing private interests. Schemes to combat ‘engrossers,’ conceived in Guanzi-derived terms, were justified by invoking a dubious Confucian scripture called The Rites of Zhou (or The Officers of Zhou). This text contains detailed prescriptions for a highly complex, interventionist bureaucracy (Smith, 1993: 82-8; Bol, 1993: 144-5).

Robert Hartwell and Peter Bol have linked the New Policies with Wang's distinctive approach to the Confucian canon. Hartwell (1971) dubbed him a ‘classicist,’ referring to his belief in certain scriptures, including The Rites of Zhou, as depicting the hallowed governance of antiquity. Although Wang advocated the revival only of the intentions behind the ancient system, he singled these texts out for emphasis in the examinations, commissioning official commentaries to the three he valued most. What he discouraged, and his followers rejected, was the study of history as moral guide or source of statecraft lessons (1971: 690-4, 712-17). Hartwell focused on the historicist alternative to classicism, especially the tradition of ‘historical analogism’ developed by Wang's foremost adversary, Sima Guang (Ssu-ma Kuang, 1019-86). This was a sophisticated discipline, supported (since the eighth century) by classified encyclopaedias of historical policy precedents that, Hartwell suggested, could have contributed to the emergence of a science of political economy (1971: 701, 708-12, 717-27).

Hartwell's study delineates an opposition between fundamentalism (classicism) and realism (historical analogism). This opposition re-emerges, differently clad, in Bol's comparisons of Wang and Sima (1992: ch. 7; 1993). In Bol's view, Wang thought there was a unitary moral-political system underlying the diverse Confucian scriptures. Close study would reveal the coherence of the system and thus provide optimal training for prospective civil servants. Perhaps surprisingly, given Paul Smith's (1993) insistence that Wang wanted entrepreneurially minded bureaucrats, the goal of education was to produce not intellectual independence but uniformity of viewpoint. Only thus could society's leaders revert to antiquity's unitary standards. The shi should become identified with the government, and ‘divergent opinion’ should be recognized as undesirable in principle. The disaffected could be expected to mellow, if only the court held fast to ‘moral principles’ (Bol, 1993: 142-6, 160-3, 170-3; Smith, 1993: 87-8). To Sima, Wang's proposed dictatorship of dirigisme and virtue betrayed ignorance of, and jeopardized, the normal workings of society. It was preferable to have intermediaries between state and common people: the rich, on whom the poor depended, and the moral-intellectual elite—the shiwho emerged from society and were its natural leaders. Rather than presuming to mould new shi, the state must earn the support, and attract the service, of those who already existed (Bol, 1993: 159, 178-80).

Song Discussions of the State and Private Wealth


Zhihong Liang Oberst (1996) puts the Wang-Sima controversy in broader perspective by surveying Song writings on the well-field system—the ancient tenurial regime whereby equal-sized squares of land were supposedly arranged in groups of nine, loosely imitating the character for ‘well.’ Eight squares were individual household allotments; the ninth, central square was tilled by all eight households for the local lord. It was a cliché in imperial China that the well-field system could not be restored, although constraints on latifundia might be feasible. As Oberst notes, however, a system of state-allocated landholdings (‘equal fields’) had been implemented in north China between the fifth and eighth centuries ad (1996: 34). For eleventh-century writers, a theoretically egalitarian land tenure system was not necessarily a utopian ideal from remote antiquity, but a historical reality.

Oberst's account of Northern Song discussions of the well-field system is based largely on examination essays. Her material suggests that for junior intellectuals of this period, the well-field system could represent a number of ideals: socio-economic equality, diligence and thrift, a rational balance between production and consumption, a rationally planned society, public security, local harmony, famine preparedness, fair taxation, universal education, and a sustainable military system. Shang Yang's putative abolition of well-fields in the state of Qin was blamed for evils such as agrarian inequality and exploitation, vagrancy, and the abandonment of land, many of which caused concern during the Northern Song. While Oberst's essayists might lament the usurpation of the sovereign's role as patron of the poor, they generally stopped short of advocating restoration of the well-fields (1996: 42-9). She groups their proposals under two headings: the ‘land ownership’ and the ‘market’ approaches.

Representatives of the ‘land ownership approach’—measures to restrict landholdings without major social upheaval—included Li Gou (Kou, 1009-59) and Zhang Zai (Chang Tsai, 1020-77). The former advocated limiting landholdings as a means to fuller exploitation of the empire's agricultural potential; the latter suggested use of non-hereditary tax-collection rights to compensate rich landowners whose land would be absorbed in well-field units (1996: 52-7). ‘Market approach’ refers to the Guanzi-inspired strategy of helping poor farmers by substituting state grain trading and agricultural credit for the depredations of ‘engrossers.’ Oberst shows that Wang Anshi was not eccentric in his animus against ‘engrossers.’ Neither Zhang Fangping (Chang Fang-p'sing, 1007-91) nor Su Che (Ch'se, 1039-1112) was aligned with Wang's reform party, but the former blamed ‘engrossers’ for much of the agrarian misery that made peasants desert the land. He advocated that the state squeeze ‘engrossers’ out of the grain business, using sumptuary laws to remove the incentive for unconscionable profiteering. Su Che proposed state loans to the needy to undermine rich usurers in 1061, eight years before Wang launched a controversial rural credit scheme. Su represented usurers as oppressors of the poor and a potential challenge to the state (1996: 58-60). Wang acted on preoccupations that others left in the realm of academic exercises, while his schemes betrayed stronger fiscal concerns than are discernible in some of his contemporaries.

Wang's ‘market approach’ literalism provoked a backlash. Oberst documents the emergence of a counter-discourse stressing the social functions of the rich. Several writers responded to Wang's bureaucratic credit scheme with vindications of the private lender. Sima Guang's role was to broaden the defence of private credit into a doctrine of mutual reliance between rich and poor, and between the state and the rich (1996: 122-8, 134-5). His lead was followed by several Southern Song scholar-officials, such as Chen Liang and Ye Shi (Yeh Shih, 1150-1223). The pro-wealth discourse included arguments that the rich contributed to the polity's stability and local government effectiveness; that social inequality was natural; that resourceless people depended on the rich for land, employment, patronage and charity; that as the state could not supply the folk with livelihoods, it should not ruin those who did; and that the rich made manifold payments to the government and generally deserved their wealth. Government's proper role was to ensure that rich and poor remained harmonious and contented with their stations. This might require action to protect the poor, but not coercively except in cases of extreme recalcitrance (1996: 135-40; Lo, 1974: 117-20; Tillman, 1994: 53-4, 56).

A ‘Liberal’ Tendency in Chinese Political Economy?


The notion that society's own arrangements can be trusted to provide a certain level of security was reflected, under Song and Qing, in an economic discourse in many ways opposed to the Guanzi tradition. This discourse combined expressions of awareness of the functioning of market forces with suggestions that these forces were best left unimpeded. Such expressions—found particularly in discussions of state intervention in the grain trade have been uncovered by historians combing administrative documents to determine the concepts of political economy reflected therein. One result has been a cautious use of the term ‘economic liberalism’ to refer to certain tendencies in public policy discussion in specific periods, especially the mid-eighteenth century.

Exemplifying the methodology of such research is the close reading of a 1763 memorial on grain brokers by Pierre-étienne Will, who pioneered the application of the terms ‘liberalism’ and laissez-faire to mid-Qing imperial grain trade policy (1980: 186; cf Will, 1990: 213). Will shows that the 1763 document expresses understanding both of the discipline spontaneously imposed by competition, and of the need for a continuous flow of trade, undisrupted by blind action against ‘hoarders’ (1999: 335-49). This documentary approach is used extensively by Helen Dunstan, who has translated mid-Qing texts reflecting belief in self-correcting mechanisms, the social utility of grain speculation, and the functioning of price incentives and the profit motive (1996: esp. 97-9, 276-8, 324-6). She argues that the term ‘economic liberalism’ has a place in the analysis of indigenous Chinese political economy if used restrictively, and if the persistence of competing interventionist tendencies is recognized and no unwarranted assumptions are imported from European history. The emergence of a rudimentary economic liberalism did not depend upon the ideology of the Enlightenment (1996: 7-8, 327-30).

This research could be developed, first, by extending its temporal scope. Robert Hymes has shown that the argument that high grain prices trigger a self-correcting mechanism was made already in a famine relief manual dated c. 1200 (1993: 295-6). We know little about how such ideas fared between then and the eighteenth century, or later in the eighteenth century. Second, attempts should be made to relate the economic thought of individual civil servants to their philosophical and scholastic allegiances within the broader Confucian tradition. William Rowe's (2001) portrait of the eighteenth-century statesman Chen Hongmou constitutes a promising beginning. Finally, systematic comparison between Chinese economic liberalism and that of the French physiocrats and Adam Smith remains to be undertaken.

Conclusion


What can scholars of premodern Chinese political thought offer to those who see it as a ‘legacy,’ important for its presumed influences on modern Chinese political development and/or presumed potential contribution to the future evolution of East Asian political forms? Perhaps the most responsible answer is to warn against superficial judgements and teleological assumptions. To be sure, the tradition offers precedents, some less convincing than others, for various political styles and forms that different groups may advocate. Justification by indigenous precedent is a time-honoured Chinese technique for rendering the alien acceptable. Such justification may come at a price, if it tends to perpetuate essentialist ethnocultural criteria for evaluating institutions. It is one thing explicitly to argue ethnocultural particularism as a political principle, another to admit it unexamined.

As to influences, indigenous political culture was but one of many factors shaping modern Chinese political development; the premodern intellectual tradition (itself not monolithic) was but one factor shaping indigenous political culture. Perhaps the most responsible statements to be made about the Chinese intellectual tradition's influence are negative. ‘The Chinese’ are not doomed by their heritage to totalitarian forms of government requiring mindless intellectual conformity. The tradition offered precedents both for super-elevation of the ruler and for the promotion of rigid orthodoxy, but these precedents were neither dominant nor necessarily esteemed. However, it is not surprising in view of the tradition that, since the late nineteenth century, authoritarianism has generally prevailed over moves towards multi-party democracy in China, and that the principles and values of Western political and social liberalism have been thought problematic. Nor is it surprising that hierarchical principles have displaced radical egalitarianism, that moralism and paternalism have featured in Chinese communist political culture, or that some advocacy of popular political participation has had an elitist tone. What may be found intriguing is that the post-Mao combination of continuing political authoritarianism with trust in market forces is not the radical departure from premodern tradition that might have been assumed.

And yet the premodern intellectual tradition need not even be invoked in a convincing explanation of contemporary developments in Chinese political economy. It probably ‘caused’ few, if any, of the phenomena listed above. The value of studying it lies elsewhere: in its intrinsic interest and in the potential for accurate knowledge and responsible interpretation to combat ill-informed assumptions.