In the Western imagination, China's history has been inextricably linked to the notion of "empire." But in fact, more than a millennium of Chinese history passed before anything resembling an empire ever existed. For centuries, six separate states battled for military supremacy, until in 221 B.C. the Qin dynasty defeated the last of its rivals and unified the country. Military conquest is only part of the imperial story, however.

The defining characteristics of the Chinese empire-and, indeed, of all, empires-were its large scale and the diversity of its peoples. While all of China's inhabitants have retroactively become "Chinese" today, this term is anachronistic for the pre-imperial period. The peoples of that time would have been known as the Qin, Qi, Chu, or by the name of one of the other Warring States, or as the inhabitants of a particular region (for example, the people "within the mountain passes"). The Qin's conquests united these groups politically in the third century B.C., but distinct regional cultures and "temperaments" survived. Such regional variations were not an inconvenient fact of life but, rather, became essential to an empire that justified itself by making just this kind of hierarchical distinction-between the universal, superior culture of the imperial center and the limited, particular cultures of regions and localities. This fundamental distinction manifested itself in political service, religion, literature, and many other aspects of Chinese life. And following the Qin, the Han empire would come as we will see below.The most important change brought about by the Qin conquest however, was the universal use of a single non-alphabetic script. By standardizing written communication among groups that did not speak mutually intelligible tongues, this innovation bound together all the regions of the empire and allowed the establishment of a state-sanctioned literary canon. Thus Keith Buchanan explained that "The real history of China is not so much the history of the rise and fall of great dynasties as the history of the gradual occupation of the Chinese earth by untold generations of farming folk." Buchanan, Transformation of the Chinese Earth, pp. 5-6. Worster, Rivers of Empire, 1970, ch. 2, "The Flow of Power in History." See also complete bibliography used for this case study, underneath.

In later periods even areas that did not become part of modern China-Korea, Japan, and Vietnam-shared significant elements of culture through their use of a common written script. Eventually, a common literary culture linked all those engaged in, or aspiring to, state service. In later centuries literacy would permeate lower levels of society, through Chinese theater, popular fiction, and simplified manuals of instruction.

In the centuries following the Qin conquest, the gradual demilitarization of both peasant and urban populations and the delegation of military service to marginal elements of society reversed an earlier trend among the competing states which had extended military service throughout the peasantry. In 31. A.D. universal military service was formally abolished, not to reappear until after the end of the last empire in 1911. In place of a mobilized peasantry, military service was provided by non-Chinese tribesmen, who were particularly skilled in the forms of warfare used at the frontier, and by convicts or other violent elements of the population, who were transported from the interior to the major zones of military action at the outskirts of the empire. This demilitarization of the interior blocked the establishment of local powers that could challenge the empire, but also led to a recurrent pattern in which alien peoples conquered and ruled China.

Finally, "empire" as it developed in early China depended on the emergence of a new social elite-great families throughout the realm who combined landlordism and trade with political office-holding. Those families dominated local society through their wealth, which they invested primarily in land, and their ability to mobilize large numbers of kin and dependents. In the classical period, law and custom divided inherited property among sons, and therefore landed wealth was subject to constant dispersal. Even large estates (although no estates in this period were large by Western standards) devolved into a multitude of small plots within a few generations. In order to reproduce their wealth over time, families were obliged to find sources of income outside agriculture. Trade and money lending were vital occupations among the gentry, but the greatest source of wealth was imperial office-holding.

Like all of Chinese history, also the geography of the early empires is a tale of the country's many distinct regions. The state created by the Qin dynasty was not the modern China familiar from our maps. The western third of contemporary China (modern Xinjiang and Tibet) was an alien world unknown to the Qin and the early Han. Modern Inner Mongolia and Manchuria also lay outside their frontiers, as did the southwestern regions of modern Yunnan and Guizhou. While the modern southeast quadrant (Fujian, Guangdong, and Guangxi) was militarily occupied, it also remained outside the Chinese cultural sphere. The China of the early imperial period, and of much of its later history, consisted of the drainage basins of the Yellow River and the Yangzi. This area comprised all of the land that was flat enough and wet enough to be suitable for agriculture, and thus defined the historical limits of the Chinese heartland.

In the Roman Empire, it was cheaper to ship grain or wine all the way from one end of the Mediterranean Sea to the other than to transport it just a hundred miles overland by wagon. Regions without water links were not integrated in the Mediterranean economy. The same was true of China. Prior to the construction of railroads in the nineteenth century, carrying grain more than a hundred miles by pack animal cost more than producing the grain itself. Except for luxury goods such as spices, silks, or gems, where small amounts produced large profits, hauling goods overland was prohibitively expensive. And a lack of good natural harbors in north China made trade up and down the coast uneconomical. Consequently, almost all bulk trade relied on inland waterways. But even this mode of transportation had its limitations. Both of the major rivers-the Yellow River and the Yangzi-flowed from west to east, with no navigable water links between them. No natural intersecting lines of transport moved north and south. Over time, as the bottom of the channel gradually rose, the river overflowed its banks. Dikes were built ever higher to prevent flooding, and in some places the river started to flow above the surrounding countryside. Today, in a stretch of about 1,100 miles, the Yellow River moles along  yards above the plain. But dikes do not control silting, and floods continued to occur on an ever larger scale. On more than 1,500 occasions during the history of imperial China the Yellow River burst its dikes, destroying farmland, killing villagers, and earning its description as "China's sorrow." But under the Qin and Han empires, the Yellow River was the core of Chinese civilization, home to around 90 percent of the population. It was separated by mountains and hills into a northwestern region (modern Gansu and northern Shaanxi), the central loess highlands (modern Shaanxi, Shanxi, and Western Henan), and the alluvial floodplain (modern Henan, southern Hebei, Shandong, northern Anhui, and northern ]iangsu). The Yangzi drainage basin, still a frontier region in this period, was also naturally divided into three regions: the mountain-ringed Min River basin (modern Sichuan), the middle Yangzi (Hubei, Hunan, and ]iangxi), and the lower Yangzi (Zhejiang, southern Anhui, and ]iangsu).

The Qin state's conquests of its neighbors and the unified empire that emerged were built on a foundation of reforms that Shang Yang, a minister from the state of Wey, carried out in the years following 359 B.C. His radical, thoroughgoing transformations of Qin military and civil life grew out of practices that were first pioneered in Qi and in Jin and its successors. Internecine wars among the Zhou nobility following the monarchy's loss of power and the eastward shift of the capital in 770 B.C. had put pressure on Qi and Jin to increase the size of their armies. Gradually these states extended military service from the nobility and its followers to the entire population of the capital, and then on to certain segments of the rural population. Under Shang Yang's adaptation of these practices, Qin peasants who served in the army were rewarded with land that their individual households could hold and work and on which they paid taxes. But there were severe punishments as well as rewards.

When the  fall of his last rival left the king of Qin master of the civilized world, he and his court were fully aware of the unprecedented nature of their achievement. As one courtier remarked, they had surpassed the greatest feats of the legendary sages of antiquity. And now they would set about enacting visionary programs designed to institutionalize a new era in human history, the era of total unity.

Yet, the Qin dynasty collapsed within two decades because it did not change enough. Despite its proclamations of making a new start in a world utterly transformed, the Qin carried forward the fundamental institutions of the Warring State era, seeking to rule a unified realm with the techniques they had used to conquer it. The Qin's grandiose visions of transformation failed to confront the extensive changes that the end of permanent warfare had brought about. It fell to the Han, who took over the realm after the Qin dynasty's defeat, to carry out the major institutional programs and cultural innovations that gave form to the vision of world empire.

Although more than nine-tenths of the population worked on the land during the Qin and Han empires, little was written about peasants. Elites preferred the color and excitement of cities and the allure of power at court. Bound to the soil, rural life smacked of the brutish and vulgar. However, Shen Nong, the so-called Divine Farmer, figured in the Han pantheon. Credited with the invention of agriculture, he was the patron sage of a Warring States tradition that insisted all men should grow their own food. An early Han philosophical compendium Master of Huainan (Huainanzi) quotes him as a law-giver: "Therefore the law of Shen Nong says, 'If in the prime of life a man does not plow, someone in the world will go hungry. If in the prime of life a woman does not weave, someone in the world will be cold.' Therefore he himself plowed with his own hands, and his wife herself wove, to set an example to the world." (Huainanzi (The Master of Huainan). In Xin bian zhu zi ji cheng (New Compilation of the Comprehensive Collection of the Various Masters), Vol. 7. Taipei: Shijie, 1974.)

Some writers adapted this doctrine to support the Qin regime, which was dependent on rural households' productivity and suspicious of merchants' wealth, and it was carried forward into the Han. Farming was even incorporated into a rarely performed ritual in which Han emperors initiated the agricultural season with three pushes of a plow in a special field. Major officials then took a turn, in order to show through simulated labor the court's interest in agriculture. The empress did her part by engaging in ceremonial weaving for the feast of the First Sericulturist.

The limits of the Qin empire roughly defined the enduring borders of the Chinese people and their culture. Although the empire was sometimes extended into the northern steppes, Central Asia, southern Manchuria, Korea, and continental southeast Asia, these expansions were generally brief. The peoples of these regions remained beyond Chinese control until the final, non-Chinese Qing dynasty. The people surrounding China can be divided into two groups. To the north and west lay nomadic societies that lived on grasslands and formed states radically different from the Chinese model. Except for the oasis city-states of Central Asia, these regions would remain outside the Chinese cultural sphere. By contrast, the watery regions of the south and southeast, as well as the highland plateaus of the southwest, were progressively settled by Chinese emigrants. There, and in the northeast, sedentary agrarian states would gradually adopt Chinese forms of writing and state organization, but these developments had scarcely begun during the early imperial dynasties.

It was through the Han expansion that China made its first contacts with peoples outside of the traditional Chinese sphere, as its emissaries reached as far as Parthia (in modem Iran), China developed its earliest firsthand knowledge and understanding of other-particularly Western-cultural worlds. Second, the triumphant military expeditions implanted the Middle Kingdom idea firmly and visibly in the Chinese worldview of international relations, in which China was the center and superpower of the world and other peoples and countries were referred to only in tributary and subordinate terms. This replaced the conception of multination equality that had gradually formed through the preexpansion Han-Hsiung-nu relations from 200 to 133 B.C. Third, through the martial merits of Emperor Wu, empire building and its accompanying military expansion became in Imperial China a permanent part of the dual criteria of historical judgment of an emperor: wen-chih (civil and cultural merit) and wu-kung (martial achievements). Without wu-kung, wen-chih was not enough to make an emperor stand out in history. Many emperors actually became prisoners of such a concept and unwisely tried too hard to fit the pattern, only to ruin themselves and their empires. Furthermore, in institutional realms I1an Wu-ti's many new political organizations, intellectual and economic innovations, and legal measures, which were all instituted to meet needs created by military expansion, remained as permanent features of traditional China, and some even survive today.

In dynastic terms, Emperor Wu's reign reached a peak in the Western Han. His long military expeditions and colonial efforts eventually exhausted the nation's economic resources and manpower, and hence affected, and even broke in some cases, the established political, economic, and intellectual balance and stability of the Western Han empire. In long-range terms, it was Wuti's great expansion that eventually precipitated the decline of the Western Han dynasty. Han Wu-ti and his new empire have been a highly controversial topic in Chinese history. On the one hand, many regarded the emperor as a model ruler and empire builder. But others considered him the personification of pretentiousness, ruthlessness, and selfishness and his great empire a project of selfdestruction and a symbol of the misery of the common people. This controversy dates even to Wu-ti's own time. In 89 B.C., he issued an edict deploring his expansionist adventures and expressing his regret about the sufferings they had inflicted on the people.All of this has increased the complexity and variations in historical discourses on the origins, development, and consequences of Emperor Wu's new empire.

The high degree of centralization in the Han government contributed to the internal stability necessary for the Han Court to mobilize large military campaigns. From the time Kao-ti ascended the imperial throne in 202 to the end of his reign in 195 B.C., practically every year the Han Court was threatened by the rebellions of feudal states. These states controlled almost two-thirds of the Han territory, and some of them were extremely large and powerfuL The Ch'i kingdom controlled six provinces (chun) with seventy-three districts (hsien), the Tai kingdom three provinces with fifty-three districts, the Ch'u kingdom three provinces with thirty-six districts, and the Wu kingdom three provinces with fifty-three districts. These kingdoms were virtually independent in every aspect. Furthermore, they had their own royal courts and governing systems, independent economic resources and financial institutions, and, most important, independent armies. The Wu, for example, had an army of over five hundred thousand men, with an additional three hundred thousand from its ally Nan-yueh in the south. Moreover, the powerful generals who helped Uu Pang (the later Kao-ti) create the empire also constituted a threat to the court. The reign of Emperor Hui (Uu Ying) lasted only seven years (195-188 B.C.). The emperor was young, and Empress Lii held the real power. Even though this was a period of consolidation of Han rule, the struggle for power between the Lii clan and the imperial family (Uu clan) was already under way. In the next period, under the reign of Empress Lii (187-180 B.C.), this struggle reached its zenith. The empress ruled through members of her own family; the important members of the imperial clan were in their distant kingdoms and marquisates. At the same time, the threat of the powerful generals continued. The court was in a state of great tension. After Empress Lii's death in 180 B.C., the whole Lii clan was massacred by a joint force of members of the imperial family and Kaotsu's old loyal henchmen. Emperor Wen (Wen-ti, Liu Heng, 180-157 B.C.) was enthroned, although he had not been the heir apparent. He was the oldest living son of Kao-tsu and had been the king of Tai (mainly Shansi) before being chosen emperor. Moreover, as he stated in an edict in the first year of his reign, at this time the king of Ch'u (Uu Chiao, Kiangsu) was his youngest uncle, the king ofWu (Uu P'i, Kiangsu and Chekiang) was his brother (actually a cousin), and the king of Huai-nan (Liu Chang, Anhui) was his younger brother. There were other strong kingdoms of the Uu clan in outlying regions of the empire. At one time, he was even reluctant to accept the throne under these circumstances. The emperor was not in a position to deal effectively with these feudal states, and tensions certainly existed between the states and the Imperial Court. Some of these kings disregarded the orders of the court and plotted rebellion against the emperor. At least two feudal kings openly rebelled against him: the king of Chi-pei in 177 B.C. and the king of Huai-nan in 175 B.C. By the beginning of the reign of Emperor Ching, the conflict between the feudal kings and the Han Court had reached a climax. While the imperial government was preparing to curtail the power of the various feudal kings, a rebellion of seven of the strongest kingdoms-Wu, Ch'u, Chiao-hsi, Chiao-tung, Tzu-ch'uan, Chi-nan, and Chao-broke out in 154 B.C. It was the most serious revolt during the former Han period. It lasted several months and was finally suppressed by the imperial forces under the generals Chou Ya-fu and Tou Ying, which killed more than 130,000 rebel troops. The kings of the seven rebel states all were forced to commit suicide or killed by the imperial forces.

It is clear that during the period from 202 to 154 B.C. the Han empire was not politically stable. The court was frequently threatened by various unruly groups and rebellions. It was impossible for the court to concentrate on external problems or launch all-out military campaigns against the Hsiung-nu and others while the constant threat of internal rebellions continued. For instance, during the reign of Emperor Wen the Hsiung-nu menace became more serious, and so did the threat of the feudal kings. In 177 B.C., the Hsiung-nu's Worthy (Wise) King of the Right invaded the regions south of the Yellow River and northern Pei-ti (in modem Ninghsia). But when Emperor Wen went to Tai (in northern Hopei) and prepared to lead a campaign against the Hsiung-nu the king of Chi-pei (in southern Shantung) immediately took the opportunity afforded by the emperor's absence from the capital to start a rebellion. The emperor was forced to call off the expedition and order his forces to attack the rebellious king.

After 154 B.C., the feudal kingdoms were never again a major threat to the Han Court. From this year to the last quarter of the second century B.C., the Han Court used various means to render the existence of feudal kingdoms merely nominal, and its effort to eliminate more feudal kingdoms continued: three more were destroyed early in Wu-ti's reign. After the Rebellion of the Seven Kingdoms, Emperor Ching also undertook special measures to change the structure of the feudal kingdoms. First, he took away the independence of their political system. He eliminated the position of yu-shih tai-fu (the imperial secretary or deputy chancellor in the royal court) in 147 B.C. and degraded the status of the chancellor in the royal court by changing its title from ch' eng-hsiang to hsiang (chief adviser) in 145 B.C. The next year he changed the governing system in the kingdoms by drastically reducing the number of officials and assigning new titles to them, showing their inferior status compared to officials in the central government. To further eliminate any possible regional division, the emperor even ordered in 142 B.C. all marquises (ch'e-hou) not to assume their posts in their respective marquisates. Second, starting in 147 B.C. the emperor gradually eliminated the feudal kings' economic independence by nationalizing mintage and the currency system and imposing a monopoly on various essential material goods. Third, Emperor Ching broadened the base of entrance into officialdom to include the common people in order to reduce the monopoly of official positions by the hereditary aristocrats and their wealthy followers. In 142 B.C., he reduced the long-established financial requirement for official appointments by 60 percent, from one hundred thousand to forty thousand in cash (copper coins). In the same year, he decreed that merchants, who were usually required to register with the government and were the allies and supporters of the ambitious feudal kings, be prohibited to serve as officials either by merit or by open purchase. With all these aggressive measures, the possibility of any successful challenge to the Imperial Court by the feudal kingdoms and their local supporters was almost completely eradicated under Emperor Ching.

At the same time, there were conscious efforts to transfer the administrative power from regular cabinet members to officials close to the emperor, as symbolized by the rise of the Inner Court (Nei-t'ing or Nei-ch'ao) to usurp the power of the Chancellery and the Imperial Secretariat. All these measures plus other formal and informal means of control and inspection brought fundamental changes in the power structure and distribution of the Han government. The power of the emperor and the centralization of the government reached their highest degree during the reign of Emperor Wu. The long struggle for power between such powerful pressure groups as the imperial in-laws, the imperial family, and those who were instrumental in the founding of the empire was finally ended during Wu-ti's reign.

Basic changes in relationship between the central political power and local society also took place. Before Emperor Wu, the monarchy was without real and close ties to local society. Kao-ti, founder of the dynasty, followed a Ch'in practice of moving the rich local elites and the powerful aristocratic families, which numbered more than one hundred thousand, from the eastern regions to the Kuan-chung area (Shensi) under the direct supervision of the central government so that these people could not induce tension and disturbances with their wealth and influence. At the same time, families of his meritorious assistants, who were given high-ranking positions in the government, were moved to the district where his tomb was constructed, which was hence named Ch'angling. Following this latter practice, succeeding rulers moved the families of officials with an annual salary of 2,000 bushels (shih) of grain and local rich elites, merchants and stalwarts to districts and towns of their tombs (ling-i), located in the capital area and constructed in their own times. This measure evidently combined the control of certain potentially dangerous segments of population-in the case of local elites and stalwarts-and the traditional system of hostage taking-in the case of high-ranking officials.  All of these practices, however, were not strictly enforced on a large scale until Wu-ti's time. 's Moreover, even if they were, they would have achieved only one goal-social stability through population control-and that alone would not be effective enough to enable the central government to fully mobilize the massive manpower and economic resources needed for long and large-scale military expeditions against the Hsiung-nu.

Two basic measures were undertaken in the early period of Wu-ti's long reign. The first was the reinforcement of the practice of population migration. In 139 B.C., the second year of his reign, the emperor first established his tomb in Mou County, which was later called Mou-ling. Next year, for the sake of positive encouragement, he granted to those who moved to Mou-ling two hundred thousand ch'ien in cash to each household and 200 mu (or mou, Han acres) of land. In 127 B.C., two years after his new military offensives had begun, Emperor Wu ordered that stalwarts from provinces and kingdoms and those whose property was worth 3 million ch'ien in cash or more be moved to Mouling. The purpose was to increase the population of the capital area and at the same time prevent the spread of evil and vicious elements in the provinces and kingdoms. The second measure that Emperor Wu undertook to exert thorough control over local conditions was to gradually incorporate leaders of local pressure groups not moved to the capital area into governmental institutions as bureaucrats. This policy was usually carried out by the provincial governors. But their efforts were directed and controlled by the central government.

Another measure in the imperial government's quest for internal political and social control and stability was the use of "harsh officials" (k'u-li). These officials believed in strict legal order as the basis of good government, in the use of cruel measures against unlawful conduct, and in the equality of all peoplecommoners, noblemen and officials-before the law. They held positions of various types and ranks, such as governor (t'ai-shou or chun-shou), regional military commandant (tu-wei or chun-wei, chief commandant), capital commandant (chung-wei), prefect or governor of the capital (nei-shih), palace counselor (chung ta-fu), commandant of justice (t'ing-wei), general, clerks in offices of different levels, and others. They often employed tricky and vicious investigators to look for unlawful activities. Their main targets of investigation and persecution were members of the rich elite, the nobility, corrupt officials, stalwarts and men of evil influence, and racketeers and vicious merchants. With only one exception in Kao-ti's time, officials with such political and legal philosophies did not gain influence and prominence until late in Ching-ti's reign-after the Rebellion of the Seven Kingdoms was crushed in 154 B.C. Emperor Ching was the first Han ruler to send an official with a harsh reputation to a specific region. In about 151-150 B.C., the famous Chih Tu was appointed governor of Chi-nan (in central Shantung) to restrain the Hsien clan. This clan consisted of over three hundred households and was so notorious for its power and lawlessness that none of the two thousand officials in the area could do anything to control it. Chi executed the worst offenders of the Hsien clan, along w!th their families, and the rest were overwhelmed with fear. After a year or so under Chih T u's rule, no one in the province dared even to pick up belongings that had been dropped on the streets and roads. Other harsh officials operated in much the same way and at times employed even harsher measures. Among the people they arrested, prosecuted, and executed in the provinces and kingdoms, as well as in the capital area, were court ladies, feudal kings, high-ranking officials, local elites, rich merchants, and commoners. Such officials became more dominant early in Wu-ti's reign. They reached the highest echelon of the bureaucracy, and their measures became even more cruel. The governor of Tinghsiang (in southern Suiyuan) executed four hundred people in one day. The governor of Ho-nei (eastern Honan) executed over one thousand families. Their blood is said to have flowed over ten Ii (Han miles). In the inquest of conspiracy for rebellion conducted by the feudal kings of Huai-nan (mainly Anhui), Heng-shan (in Anhui), and Chiang-tu (central Kiangsu), Chang T'ang-the most influential harsh official of early Han times-was in charge of investigations and judgments. He put to death tens of thousands, at times merely on circumstantial evidence. These officials became so notorious that they were given such nicknames as Vicious Hawk, Vicious Killer, and the like.

Furthermore, in 130 B.G Chao Yii and Chang Tang, two of the most notorious and influential harsh officials at the time, were empowered by Emperor Wu to draw up various new statutes and ordinances. Among these were the laws that anyone who knowingly allowed a criminal act to go unreported was as guilty as the criminal and that officials could be prosecuted for the offenses of their inferiors or superiors in the same bureau. From this time on, the laws became more complicated, and they were applied with increasing strictness. This was a pronounced departure from earlier Han practice, which in general stressed simpler laws and lenient applications. In the earlier reigns of Emperors Kao and Hui and Empress Lii, as recently discovered Han era documents show, there were only twenty-eight sets of statutes and ordinances and a number of cases and precedents for reference and comparison. The principal legal philosophy was "following the established tradition."20 The clear intent of this change was to impose strict political and social order by means of harsh legal institutions and enforcement. Small wonder, then, that of the fifteen notorious harsh officials of Former Han times whose biographies appear in the Shih-chi (Historical Records) and Han-shu (History of the Former Han Dynasty), ten flourished in Wu-ti's time. All were extremely influential in decision making at the highest level, and a majority of them were instrumental in the formulation of the most important new political and economic measures undertaken during Wu-ti's reign.

During the seven years of the Ch'in-Han transition from 208 to 202 B.C., the main force of production in the economy, the population of able bodies, was drastically reduced because of the continuing war destruction and carnage. There were at least eighty large-scale battles and over one hundred fifty of lesser scale. However, the casualties in each of these bloody conflicts ranged from several thousands to tens of thousands and even hundreds of thousands. In 207 B.C., for example, over two hundred thousand Ch'in soldiers were buried alive by the rebel leader Hsiang Yii (232-202 B.C.) in Hsin-an (east of modern Mien-ch'ih in western Honan). Overall, the Chinese population at the beginning of the Han dynasty was reduced to less than one-half of the former Ch'in figure of 28 million by the end of the dynasty. In many regions, the loss was even greater. Chii-ni District (in Hopei), for example, had only one-sixth of the Ch'in era population left, down from thirty thousand households to five thousand. The large cities generally, retained only 20 to 30 percent of their former populations due to war deaths and flight. The great T'ang historian Tu Yu (A.D. 735-812) estimated that the population of the Han dynasty at its founding was less than one-third of the population of China in the Warring States period (404-222 B.c.). The modem scholar Liang Ch'i-ch'ao (1873-1929) estimated the Han population in Emperor Kao-tsu's time at only about 5 or 6 million, but more recently others have estimated it to have been in the range of 8.8 to 18 million; my own estimate is 12 to 16 million.

The founders of the Han took special measures to revitalize the bleak economic and social conditions. The government first instituted a general policy of economic relaxation and reduction of governmental spending. The theory was that government should interfere in the people's lives as little as possible. This was intended to correct the Ch'in policy of working the people so hard in public works and military expeditions that they eventually rebelled. At the same time, the Han Court also initiated various measures for economic recovery in different realms of concern. During the early Former Han, industrial and commercial growth was noticeable. But Han industry and commerce did not begin their full development in the first three reigns, 202-180 B.C., of the new dynasty, since the primary concern of the Han leaders then was full recovery of the agricultural sector. As it was in the Ch'in dynasty, the main economic concern of the Han founders was agriculturalism (nung-pen), with commerce and industry being regarded as "nonessential" economic pursuits (mo-yeh). During the reigns of Emperors Wen and Ching, 180-141 B.C., commerce and industry gradually achieved significant growth, as the empire's agricultural production had reached a very high level, population had increased by four times, and living standards (consumption of goods) had risen significantly. On the average, the minimum annual rate of business profits was 20 percent and higher. In fact, the profits of certain industries and businesses were so huge that later in the next reign, Han Wu-ti's time, they were channeled into a well-organized new system of governmeQt monopolies in order to finance the large-scale military campaigns and eliminate the economic threat to the imperial government posed by the tremendous wealth of the industrialists and merchants.

There were, according to Su-ma Ch'ien (145-86 B.C.), forty-six notable types of commodities of great value-major sources of industrial and commercial wealth-in market towns and commercial metropolises, and twenty-two of these were produced through industrial processes of varying degrees of sophistication. Regional specialization emerged and with it the rise of prosperous interregional trade. Major industrial enterprises were those of iron, salt, and textiles. The centers of the textile and clothing industries (silks, silken fabrics, textiles made of vegetable fibers, and so on) were mainly in Ch'i and Lu (modern Shantung), Ch'en-liu (in modern Honan), and Shu (in modern Szechwan). Lin-tzu (Lin-tse) of Ch'i and Hsiang-i of Ch'en-liu were the two largest centers. Lin-tse was famous for garments and Hsiang-i for fine embroideries. The speciality of Shu was hemp cloth. Information on the size of these industries in early Former Han times is not available. But in 48 B.C. each of the three government garments offices (San-fu) in Ch'i alone generally employed two to three thousand men. Together with the fact that textiles were a major source of industrial and commercial wealth, this leads us to assume that a large preexpansion textile factory could easily have employed over a thousand men.

The iron and salt industries were spread over various regions of the empire, with the largest centers located in what are modern Shantung, Szechwan,' Kiangsu, Chekiang, Anhui, and Hopei. From 120 to 110 B.C., the Han government reorganized its system of salt and iron monopolies. It established special offices for supervision and management in areas where salt and iron production and profits were concentrated. In early Han times, there were forty iron offices in forty chun (provinces) and kuo (feudal kingdoms), and thirty-eight salt offices in thirty chun and kuo. Excluding regions that were acquired during Wu-ti's expansion, forty-five iron offices and thirty-one salt offices were in Han regions of the preexpansion period. These regions most likely were centers-or, in some cases potential centers-of salt and iron production in the pre-expansion period. The chun and kuo where these Han offices were located are listed in table I, together with their modern geographical locations. The wide geographical range of salt and iron production is clearly shown in the table.
Information on the total work force of salt and iron laborers in the early Former Han period is not available. But certain sources indicate that salt and iron magnates often employed more than a thousand men to manufacture salt and process iron. Kung Yii (123-43 B.C.) observed in 44 B.C. that the various iron offices employed an annual work force of over one hundred thousand men, mainly slaves, to gather iron and copper. It seems that the salt and iron monopoly implemented in the years 120-110 B.C. was mainly a new attempt to control these businesses, not a rapid expansion of the existing enterprises. Judging from our examination of salt and iron offices, about 88 percent of the production facilities of salt and iron probably existed in the preexpansion period. If this is the case, then very likely the total annual work force of salt and iron laborers in the early Former Han was around eighty-eight thousand men.

Salt and iron businesses were evidently the most profitable enterprises in early Former Han times. They produced such wealthy and powerful families as the Cho and Cheng of Shu, the K'ung of Wan (in modern Honan), and the Ping of Ts'ao (in modern Shantung), all in the iron enterprise; and the Tiao Hsien of Ch'i in the salt enterprise. The Chos and Chengs grew so rich that they each owned a thousand young slaves. Their pleasures in possessing lands and in fishing, archery, and hunting were comparable to those of great feudal lords. The K'ung family's fortune reached several thousand catties of gold, and its head resembled the young men of princely rank in his behavior, disposition, and activities. The wealth of the Ping family amounted to 100 million in cash. The Tiao Hsien's wealth grew to several tens of millions in cash. Needless to say, all of these families engaged in diverse trading and other commercial activities and employed every possible means-including lending money, skillful use of slaves, and political contacts with feudal lords, provincial governors, and prime ministers of feudal kingdoms-to increase their fortunes. Their wealth and power exerted great influence on the lifestyles and thinking of people in various walks of life. The traders in Nan-yang (modem Hopei) all imitated the K'ung family's lordly and openhanded ways. In Tsou and Lu (in modem Shantung), many people abandoned scholarly pursuits and, following the model of the Ping family, turned to the quest for profits. Various feudal states, particularly Wu in the south and Chao (in Hopei) in the north, engaged in the production of iron and salt for huge profits and at times became the largest producers of these commodities. In fact, the tremendous financial strength derived from iron and salt production enabled these states to threaten the central government and invited it to take them over.

Copper, the source of coins, was another profitable industry. It usually went with iron manufacture. A considerable number of the iron manufacturers also engaged in copper mining and casting. Shu and Tan-yang (in mqdern Anhui) and part of the Wu kingdom (the lower Yangtze Valley), among others, were the well-known copper-producing regions. At the time of the emperors Wen and Ching (180-141 B.C.), the two most productive copper mines were located in mountains in Yen-tao of Shu (modem lung-ching of Szechwan) and Ch'angshan of Y ii-chang (modem An-chi of Chekiang). The former was granted to the high official Teng Tung by Emperor Wen; the latter was in the territory of the king of Wu (Uu P'i, 213-154 B.C.). Both Teng and the king of Wu minted coins from copper mined from the two mountains by tremendous numbers of workers. The result was that the coins of Wu and of Teng spread all over the empire. Teng accumulated wealth that exceeded that of a vassal king. For the king of Wu, the mintage of coins, together with his salt enterprise, produced so much revenue that he not only dispensed with taxation but was economically confident enough to start a rebellion against the imperial government.

But Han expansionism would also become their downfall, because the Han needed nomads to join the army, yet they were never fully incorporated into the military hierarchy. Instead, the Han government relied on the standing frontier commands to keep them under control. As more and more tribes moved inside the frontiers, this burden proved too great for the relatively small armies in the frontier camps. Loyalty was also weak among the convicts and professionals who spent their lives at the frontier and were linked to the Han state only through the person of their commander. Another reason for the failure of the Eastern Han army in the second century was its success in the first. Just as the armies of the Warring States and early Western Han had been designed to fight Sinitic rivals, so those of the Eastern Han had been aimed at the northern Xiongnu. With their defeat, many of the "inner barbarians" who had helped the Han in the first century turned against it. The Southern Xiongnu, Wuhuan, and Xianbei lost their chief motive for submission to the Han, as well as their chief source of bonuses for military service. So the Wuhuan and Southern Xiongnu turned increasingly to internal pillage for income, while the Xianbei replaced the Xiongnu as the chief external threat. To the west, the problem was even more severe, for this area suffered through the disastrous Qiang wars.

Every army is intended to fight a certain type of war or counter a particular kind of threat. The entire Eastern Han defense faced north, providing a screen against small-scale raids and a warning in case of invasions. Its large cavalry forces were assembled for offensive expeditions against a united foe with substantial armies. Such dispositions were of little use against the Qiang, located to the west beyond the Han's border defenses. These nomads lacked any over arching political order and did not form large confederacies. The consequences of any defeat were thus limited, and victories, however small, soon led to major rebellions as scattered groups assembled under a successful leader. For the same reasons, peace agreements with the Qiang could not last for long. Moreover, scattered groups of Qiang lived throughout the western and northwestern territories, as well as beyond the frontier. There was no clear geographic boundary between the Qiang and the Han, and under the Eastern Han the Qiang were resettled in the old capital region. The only defense against such an adversary was to move Han farmers and soldiers into the provinces, so that no settlements were left exposed to low-level attacks and the Qiang could be absorbed into the Han economy and polity.

But whenever the Han attempted such a policy, it ended in failure. In 61 B.C. Zhao Chongguo propose founding military colonies in the west (Honan), Ho-nei (Honan), Chi-nan (Shantung), T'ai-shan (Shantung), and Shu and Kuang-Han (Szechwan). In addition, shipbuilding, weaponry, and lumber were profitable industries, particularly in regions such as Lu-chiang, Nan (Hupei), and Shu. Animal husbandry was an important enterprise in northern and northwestern border territories. Pottery and lacquerware also were prosperous industries in certain regions.

With these commercial and industrial developments, the cities rapidly expanded. In Wu-ti's time, there were twenty Han cities with populations ranging from 50,000 to 650,000 people, and sixty cities of 20,000 to 56,000. The two largest cities in population were Ch'ang-an, the imperial capital built only in Hui-ti's reign (especially in 192-189 B.C.), and Lin-tzu (in Shantung); the former had a population over 500,000 in a walled city of 13.5 square miles, and the latter had 650,000 in a walled city of 9.31 square miles. The next five largest were Yuan (in Nan-yang, Honan), with 4°0,000; Ch'eng-tu (in Chengtu, Szechwan), with 380,000; Han-tan (Han-tan, Hopei), with 27°,000; La-yang (Loyang, Honan), with 260,000; and Lu (Ch'ii-fu, southern Shantung), with 23°,000. These seven cities were the major Han commercial and industrial hubs and were located in the key economic regions in the west, central, northeast, east, and southwest. They were the distribution centers of special regional products such as iron, gold, copper, textiles, lacquerware, and agricultural goods. In essence, they were the nerve centers of the Han economic and business world. The cities were naturally the centers of political command and economic and military mobilization to support longtime war efforts. Considering the fact that at the beginning of the Han dynasty these major urban centers had only 20 to 30 percent of the surviving Ch'in population, their tremendous growth and size certainly informed the stupendous increase of the Han population in the sixty some years before Wu-ti's reign. At the same time, it was also recorded that all earlier Han reigns had made special and aggressive efforts to promote population growth. Emperor Kao, for example, decreed in 200 B.C. that all taxes be forgiven for a family with a newborn baby. Under Emperor Hui, the court even ordered that a woman's whole family be levied taxes five times higher than normal if she was not married by the age of twenty-nine (thirty sui). Under these aggressive population policies and favorable economic conditions, it is reasonable to assume that the population would have increased over time. In fact, the Han population is estimated, in different primary sources and later references, to have reached the range of 40 to 50 million before 150 B.C. and increased to the range of 50 million by Wu-ti's early reign, almost five times the early Han total.

By examining the availability of the Han economic sources and resources and the strength of the Han military, modem scholars have estimated that in Wu-ti's time the Han government revenue was about 12 billion in cash. However, as soon as the colonists had pacified the were allowed to return to their former homes. A few years I was another attempt to settle a permanent agricultural popula1 region, and others between 101 and 104 A.D. But when the Q exploded on a large scale in IIO A.D., the government pulled back to Chang' an. Loc;al officials sent out from the interior, I knowledge of the region, ordered the abandonment of three (eries, the confiscation of crops, and the leveling of homes so d would return. By 111 the population of the entire former cap] in Guanzhong was in flight. An attempt was made between 12 to restore the abandoned commanderies and establish militaf) but when Qiang uprisings resumed in 137, no significant ref had taken place.

Throughout the Eastern Han, particularly in the second cen the population of Guanzhong and the old capital region dras dined under the continuous pressure of Qiang onslaughts . Even in the early decades of the first century, the northwe regions had been seriously depopulated. The policies of barbarians inside China and sending convicts to the frontier may have been in part an attempt to repopulate these regions. However, these measures did little to check the demographic decline of the frontier. Census evidence shows that, with one exception, all commanderies in the west and northwest suffered significant losses, many of them more than 80 or 90 percent. While the figures are unreliable, a change of this magnitude, especially when contrasted with the relative stability and even some increases in inner provinces, probably indicates an actual decline in the Han population in the border regions.

Contemporary observations support these statistics. Wang Fu (ca. 90-165 A.D.) noted: "Now in the border commanderies for every thousand li there are two districts, and these have only a few hundred households. The Grand Administrator travels about for ten thousand li, and it is empty. Fine soil is abandoned and not cultivated. In the central provinces and inner commanderies cultivated land fills the borders to bursting and one cannot be alone. The population is in the millions and the land is completely used. People are numerous and land scarce, and there is not even room to set down one's foot." Writing several decades later, Cui Shi described a situation that was virtually identical.

The Eastern Han government made futile attempts to prevent people from leaving the frontier regions and to encourage those who had left to return. The Book of the Later Han (Hou Han shu) states, "Under the old system [under the Han] frontier people were not allowed to move inward." In 62 A.D. Emperor Ming offered a payment of twenty thousand cash to any refugee from the frontier who returned to his old home. As clear evidence for this ban on inward movement, Zhang Huan, who came from Dunhuang in the far northwest, was allowed to move to an inner commandery in r67 A.D. only as a special reward for meritorious service.

But these attempts to stabilize the frontier population failed. Between 92 and 94 A.D., Emperor He proclaimed geographic quotas to correlate the number of people recommended as "filially pious and incorrupt" (the primary route to office) with the population of a region. For every twenty thousand registered people, a commandery would be allowed to recommend one man per year. For a population between ten thousand and twenty thousand, a commandery was granted one recommendation every three years. But in ror A.D. frontier commanderies with a population of between ten thousand and twenty thousand were allowed to recommend one man every year. Those with a population of between five thousand and ten thousand could recommend one every other year. Aad those with fewer than five thousand were granted one every three years. This change shows that the populations of frontier districts were low and declining. Even the reduced limits were too high for many commanderies. Wang Fu observed that because of low population, the commanderies in his region had been unable to recommend even a single man for more than a decade. An examination of the geographic origins of the "filially pious and incorrupt" recorded in the Book of the Later Han and on stone inscriptions bears out his complaint.

The conduct of the Eastern Han government in the Qiang wars demonstrates a fundamental weakness of the regime: its single-minded focus on the Guandong region. The scale of the Qiang disasters and the collapse of Han civilization in the west and northwest were direct consequences of the eastern government's ultimate decision to leave the frontier commanderies defenseless and to remove population from the region. This lack of interest in the security of the west and northwest, which can be observed throughout the Eastern Han, stems from the shift of power to the new capital in the east.

When the Western Han capital was based in Guanzhong, the government pursued a policy of forcibly resettling population into new towns for the maintenance of imperial tombs. Through resettlement, powerful provincial families lost their local basis of influence and fell under the sway of the imperial court. Grain and other foodstuffs were eventually imported from the more productive Guandong region to maintain the demographic and economic well-being of Guanzhong.
The Western court regarded the area "east of the passes" with a mixture of suspicion and contempt. Jia Yi (201-169 B.C.) observed to the emperor: "The reason for which you have established the Wu, Hangu, and Jin passes is largely to guard against the enfeoffed nobles east of the mountains." In the Discourses on Salt and Iron (Yan tie lun) Sang Hongyang (executed 80 B.C.) remarked: "People have a saying, 'A provincial pedant is not as good as a capital official.' These literati all come from east of the mountains and seldom participate in the great discussions of state affairs." Although men from Guandong played a larger role in Western Han government after Emperor Wu's death, only when the capital moved to Luoyang did the situation truly change.

The Eastern Han founder Guangwu and most of his followers came from just south of Luoyang, and the rest of his closest adherents came from the great families of Guandong. Moving the capital from Guanzhong to Guandong transferred political power to their region.~ This break with the past was made self-consciously and deliberately, without regard to strategic considerations, particularly the fact that the newly reunited Xiongnu were drawing near Luoyang. Throughout the Eastern Han the court entertained many proposals to abandon territory in the north or west, leaving the old capital region vulnerable. In 35 A.D. officials urged that everything to the west of the Gansu corridor be abandoned, but this was blocked by Ma Yuan, a man from the northwest. In 11O A.D., in the wake of the Qiang uprisings, a proposal called for the abandonment of all of Liang province (from the western end of the Gansu corridor at Dunhuang east to the borders of the capital region around Chang'ang and even some of the old imperial tombs. Opponents of this idea argued that the warrior traditions of the western people were essential to the security of the empire, and that moving them toward the interior would incite rebellion. By the end of the Western Han the office of provincial governor ha grown from a mere inspector into the chief local administrator.

As the governors' power increased under the Eastern Han, they were able to appoint and dismiss officials within their provinces without the approval of the court. Holders of the office thus became autonomous regional lords who, though subject to dismissal, held sway within their own jurisdiction. Their powers included military duties, and in the second century A.D., when barbarian incursions and banditry led to constant combat, the governor replaced the grand administrator as the person in command of the state's emergency levies . As civil order decayed and provincial forces spent more time in the field, they took on the characteristics of semiprivate standing armies.

This development was a major change in Han local administration and an important step in the fall of the Eastern Han. The Han dynasty had based its administration on the commandery and the district-a two-tier structure that fragmented local power into small units to avoid threats to the central government. The provincial governor, however, became a third tier, with command of large populations, great wealth, and significant armed forces-resources that could challenge the authority of the imperial government. In the second half of the second century A.D., governors became semi-independent warlords. When Liu Yan took office as governor of Yi province, he massacred important local families, gave his own sons major positions, recruited personal followers from among refugees, and defied imperial commands. In similar fashion, Liu Yu established his own little kingdom as governor of You province. He pacified local barbarians, sheltered refugees, encouraged crafts, and gathered armies. Liu Biao pursued an identical course of action in Jing province.
By the late Eastern Han, governors had obtained the power to recruit troops on their own initiative. This in effect recognized their right to command private armies. In 178 A.D. when the provinces of Jiaozhi and Nanhai (in southern Guangdong and Vietnam) rebelled, Zhu Jun was sent out as governor and empowered to recruit "household troops" (one of the earliest uses of the term) to form an army. The commentary identifies these troops as his servants and slaves. In 189 A.D. He Jin sent Bao Xin to his home near Mount Tai to recruit troops for the purge of the eunuchs. By the time Bao Xin returned, He Jin had been slain. Bao Xin went back to Mount Tai, recruited twenty thousand men, and joined forces with Cao Cao, the warlord who ultimately conquered the Yellow River Valley and whose son formally brought the Han dynasty to an end. The delegation to individuals of the power to recruit private armies in their home regions shows that the central government had lost its ability to rule the population. Only through the personal networks of eminent families in their home regions could the state mobilize a military force.

Recruits in the provinces developed strong ties to those who recruited them. In 88 A.D. a certain Deng Xun had recruited Xiongnu soldiers to act as guards against the Qiang. Contrary to normal practice, he allowed these tribesmen and their families to live in his fortress, and he even let them into his own garden. They swore personal loyalty to Deng Xun and allowed him to raise several hundred of their children as his personal fol1owers. This was an exceptional case at the time, but by the end of the dynasty such ties between recruits and their commanders were common.

In r89 A.D., when Dong Zhuo declined to leave his army at the northwestern frontier and take up an appointment at court, he wrote: "The righteous followers from Huangzhong and the Han and barbarian troops under my command all came to me and said, 'Our rations and wages have not yet been completely paid, and now our provisions will be cut off, and our wives and children will die of hunger and cold.' Pulling back my carriage, they would not let me go." When the court attempted to have him yield his command to Huangfu Song, he replied: "Though I have no skill in planning and no great strength, I have without cause received your divine favor and for ten years have commanded the army. My soldiers both great and small have grown familiar with me over a long time, and cherishing my sustaining bounty they will lay down their lives for me. I ask to lead them to Beizhou, that I may render servi~e at the frontier."

This second passage points out another feature of the Eastern Han's collapse: the proliferation of long-term commands in the field. In the Western Han, generals had been appointed to command an expedition, after which the army was disbanded and the general returned to his regular post. The "Monograph on Officials" of the Book of the Later Han states, "Generals are not permanently established."20 However, the Eastern Han created permanent armies stationed at fixed camps. Although in the first century A.D. the size of armies was kept small and commanders were regularly rotated, prolonged crises on the frontier required generals to remain with their armies in the field for years. These armies-which now were composed of barbarians, convicts, and long-term recruits-became the loyal creatures of their commanders. Such men had no place in Han society and no home or family to which they could return. Instead, they formed families at the frontier, and their lives centered on the person who was, as Dong Zhuo observed, the source of their livelihood. TheHan court never acknowledged this shift. In Dong Zhuo's biography his title changed frequently in the ten years prior to r89 A.D., but his own testimony shows that he and his army stayed together for the entire period.

Another path leading to private armies was the development of a dependent tenantry. The absorption of the old category of "clients" into this new servile group meant that labor and military service were largely transferred from the state to great families. In the early Eastern Han, Ma Yuan commanded the services of several hundred families attached to him as clients.Military service was probably included in these obligations. Drawing from these service-providing dependents, the great families were able to assemble armies of hundreds or thousands of men. Such armies of tenants had overthrown Wang Mang at the end of the Western Han, and the military capacity of dependent populations existed, as a latent possibility, right up to the Eastern Han dynasty's collapse. Like the government's commandery troops, they could be raised in times of emergency. With the decay of internal order and the outbreak of civil war, these dependents began to form full-time private armies recruited from what was becoming a hereditary soldiery. At the same time the dwellings of the great families became fortified compounds with walls and watchtowers.

The Eastern Han government gave up all attempts to restrict the rise of a dependent tenantry, and in so doing abandoned direct administration of the countryside. Furthermore, as power shifted to the inner court of affines and eunuchs, the imperial house became separated from the great families who controlled the outer court. This steady implosion" of imperial power ruptured the ties that bound the court to the countryside. As social order steadily deteriorated in the second century A.D., the court discovered that it had lost the ability to mobilize armies and enforce its own rule.

To counter the threat of rebellious "inner barbarians" and ultimately the millenarian rebel movements, the imperial government required military resources that only those who had developed personal ties to the soldiery could muster: provincial governors, generals on the frontier, resettled tribal chieftains, great landlords, and, in a few cases, leaders of religious rebels. While each type of commander had secured support in a different way, all of them had one thing in common: in an age of general social breakdown, they could call upon their own armed followers for security. These various warlords were key political actors in the long centuries of disunion that would follow the demise of the Han.

As the Chinese are quick to point out, the Han leadership has almost exclusively focused on defense on the international front (even while it suppresses ethnic minorities in the buffer zones). When China did reach out, aside from during the time of Mongol domination, it was largely along the Silk Road through Central Asia and into the Middle East, where China sought to acquire luxury goods more than vital resources. Even the famed treasure fleets of Zeng He in the early 15th century were more an expression of China's confidence in its own defensive position and its desire for frivolities than a strategic imperative -- and as threats of invasion from the north increased, China quickly abandoned its oceangoing enterprises, considering them expensive and distracting from real priorities.  

Case study: The Application of Law in the Early Chinese Empires

Written codes first emerged in the Warring States period, when tax and service obligations were extended to lower levels of urban society and to peasants in the hinterland. Local officials responsible for enforcing these obligations required written laws and regulations stipulating procedures for keeping accounts, the penalties for crimes, and other aspects of ad¬ministration. But far from being merely the tools of rational administration or of brutal realpolitik, these codes were embedded within the religious and ritual practices of the societies from which they emerged.

Religious links

Stories in the Transmissions of Master Zuo (Zuo zhuan), set in the seventh through fifth centuries B.C., depict Zhou aristocrats ritually invoking powerful spirits with blood sacrifices and calling upon these spirits to enforce the terms of their oaths. Such covenants, sanctified through smearing the lips of participants with the blood of sacrificial animals and burial of the covenants in the ground to transmit them to the spirit world, were used to form alliances between states or lineages. They also dictated the agreed-upon rules to be observed by all who joined these leagues. The recent discovery of some of these buried covenant (meng) texts from Houma, Wenxian, and Qinyang, along with the rereading of received texts in light of these discoveries, shows how such texts provided a religious foundation for a new political authority based on writing.1

In addition to covenants, a second form of writing that sacralized the earliest legal codes was inscription on bronze vessels of the type used in the religious cults of the Shang and Zhou. These inscriptions had served, among other purposes, to communicate with ancestors and to render permanent any gifts of regalia or grants of political authority made by the ruler. In the late Western Zhou and Spring and Autumn periods, several inscriptions record decisions in legal cases, most commonly disputes over land. A vessel discovered in the cache at Dongjia village records the punishment of a cowherd, who was sentenced to whipping and tattooing. By the sixth century B.C., according to the Transmissions of Master Zuo, the states of Zheng and Jin used bronze vessels to consecrate their new legal codes. Thus, inscriptions on sacred vessels that had fixed power and privilege under the Zhou were adapted to codify the powers of the emerging territorial states.2

The sacralization of law in covenants and bronzes did not end with the development of more elaborate codes written on bamboo or wooden strips. Han texts narrate several occasions in the Qin-Han interregnum and the early Han when ceremonies accompanied by blood sacrifice were used to consecrate new laws. But by this time the emphasis had shifted to the text of the oath as the binding force-a recognition of the power of sanctified writing.3

The religious links of early legal codes are also indicated by the discovery of substantial samples of late Warring States law in tombs of officials at Yunmeng (Qin state) and Baoshan (Chu state). This shows that legal texts figured in funerary ritual. It is unclear whether the documents were buried because they were powerful, sacred texts that would protect the deceased in the afterlife or whether they were an element in the program of equipping the tomb with all the materials needed to continue the deceased's mode of living in the world beyond. In either case, in the still overlapping realms of funerary cult and political authority, these legal texts played a role reminiscent of the Zhou bronzes. The deceased held them through the gift of the ruler, and they were signs and tools of the holder's power over his subordinates. Both binding and empowering,the texts were carried into the afterlife to preserve the status of the deceased.

The legal texts at Yunmeng and Baoshan were successors not just of the Zhou bronzes but also of the covenants. Like these ritual oaths, the legal texts were buried in the earth for transfer to the spirit realm. But more important, the texts played a pivotal role in the creation of the state by transmitting the policies of the ruler directly to leading political actors, who in turn transmitted them to their own subordinates and kin. The names on the buried covenant texts were heads of locally powerful families who had come into the presence of the ruler of the emerging state and sworn loyalty to his person and lineage. These oaths bound not only the family heads but the lesser members of their lineages. Similarly, the laws of the Warring States were inscribed on ritual texts bestowed upon political actors, who were bound to the ruler through the receipt of these sacred objects and who in turn imposed the rules on their subjects. The physical bestowal of the written statutes and associated documents at or in the wake of the ceremony of appointment was central to the law's function, and this ritualization of the code was carried forward in funerary rites.

This focus of laws on the ruler's control of officials is clear in the legal texts from Yunmeng, where the common people appear only in a secondary role. In these documents, the first and longest section in the groupings used by modern editors ("Eighteen Statutes") deals almost entirely with rules for official conduct, guidelines for keeping accounts, and procedures for the inspection of officials. The second section ("Rules for Checking") dictates the maintenance of official stores and the records thereof. The contents of the third section ("Miscellaneous Statutes") closely resembles those of the first two. The fourth section ("Answers to Questions Concerning Qin Statutes") defines terms and stipulates procedures so that officials could interpret and execute items of the code in the manner intended by the court. The fifth section (" Models for Sealing and Investigating") instructs officials in the proper conduct of investigations and interrogations so as to secure accurate results and transmit them to the court.4

The emphasis on the control of local officials reappears in the text "On the Way of Being an Official" found in the same tomb. The official is charged to obey his superiors, limit his own wants, and build roads so that directives from the center can arrive rapidly and without modification. It praises loyalty, absence of bias, deference, and openness to the actual facts of cases as the highest of virtues. It attacks personal desires, acting on one's own initiative, resisting superiors, and concentrating on private business as the worst of faults. In short, it proclaims the new ideal of the official as a conduit who transmits the facts of his locality to the court and the decisions of the court to the countryside without interposing his own will or ideas.5 This is the sort of official that was to be created through the dictates of the legal documents in the same tomb.

Principles underlying the early legal codes are also linked to the ritual practices of the period. Two of particular significance are the idea of punishment as do ut des (the exchange of one thing for another) and the importance attributed to titles and names.6

Divination materials found with the legal texts in the fourth-century

B.C. Baoshan tomb reveal a system of curing/exorcism through sacrifice that follows the Shang pattern. The physician/diviner ascertains the identity of the spirit causing the disease, its relation to the patient, and the type and number of sacrifices necessary to assuage it. The ritual is a mechanical form of exchange with no moral dimension. A similar process of identification of the spirit culprit and mechanical ritual expulsion or appeasement informs the "Demonography" found in the Yunmeng tomb. This text's title jie is a technical term in legal documents meaning "interrogation" but also refers to the commanding of spirits through the use of written spells; in Zhou documents this term meant "to obligate oneself to the spirits by means of a written document." Here a term used for written communications with the spirits was applied to the legal practice of making written records of testimony by witnesses. This close connection between religious and legal language figures throughout the texts from Yunmeng.7

Texts on demon control share with legal documents not only a common vocabulary but also a common mode of practice. In both spheres religious and legal-order and control are maintained through the process of identifying malefactors and applying graded responses sufficient to counteract the threat or compensate for damage. Legal punishments incorporated the minute, mathematical gradations that had characterized sacrificial responses to threatening spirits since the Shang. This parallel between exorcism and punishment was noted in a passage from the text of political philosophy Master Han Fei (Han Feizi), written under the Qin or early Han: "Ghosts' curses causing people to fall ill means that ghosts harm people. People's exorcising the ghosts means that people harm ghosts. People's breaking the law means that people harm their superiors. Superiors' punishing the people means that the superiors harm the people.8

The Yunmeng documents include a "Book of Days" (ri shu)-a text for determining which days were favorable or unfavorable for certain activities-which contains a guide to thief catching (a legal concern) through divination. The guide describes how the physical appearance of the thief can be determined based on the day when the crime occurred. Other strips deal with appropriate days to take up a post and indicate the consequences of holding audiences at various times of the day. Since these mantic texts were buried together with the legal materials, it is likely that the deceased official or his subordinates employed them in their everyday administrative activities, further blurring the line between legal and religious practice.9

The link between law and religion in early imperial China also entailed bringing the actions of government into conformity with Heaven and nature. For example, executions could legally take place only in autumn and winter, the seasons of decay and death. If a man condemned to death survived the winter, due to a procedural delay or dilatory action, then he was no longer liable to execution. One story tells of Wang Wen shu, a harsh official in the time of Emperor Wu who had just executed several thousand households of locally powerful families. "When the beginning of spring came, Wang Wenshu stamped his foot, sighing, 'If only I could make the winter last one more month it would suffice to finish my work!"10

A related practice was the regular granting of general amnesties for all but the most serious criminal offenses. Such amnesties were bestowed on happy occasions pertaining to the imperial household, such as the birth of a son or establishment of an heir. They were also granted in response to natural calamities thought to derive from excessive severity. In his role as a giver of life, the emperor patterned his amnesties on the life-giving potency of Heaven, his spiritual father and divine equivalent.11

The idea that human misconduct disturbed the natural order led to certain legal precepts pertaining to punishment. Punishments were imposed to baa (requite or pay back) the deed, implying a restoration of the natural balance disturbed by the crime. To be efficacious, a punishment should not be either too severe or too lenient; if the punishment was not in balance with the crime, the natural order would not be restored. On the basis of this theory, scholars would in some cases argue that particular natural disasters or unusual phenomena resulted from inappropriate punishments, or failure to punish at all.12

Law and Administration

While law in the Warring States and early empires remained embedded in the religious practices and ideas of the period, it was above all a tool of administration used to preserve social order. In this role, law went beyond simply enjoining certain actions, prohibiting others, and enforcing both with penal sanctions. An entire model of the social structure was built into the patterns of rewards and punishments stipulated in legal codes. The judicial practices of the period trace the lineaments of a properly functioning society as understood by the ruler and his agents.

One of the clearest examples of this is the differentiation of rewards and punishments based on the status of those involved. The semi-divine status of the emperor found legal expression in the designation of offenses against his person or property as the most heinous of crimes. Inflicting damage on an imperial dwelling or tomb, for example, was punishable by death. If an artisan's error caused a wheel or axle on the imperial chariot to break, the artisan was executed. Even accidentally breaking an imperially bestowed gift, such as a dove staff, could lead to capital punishment for an official.

This ranking of misdeeds and punishments also applied to ordinary households. Crimes committed by kin were more serious than those committed by a stranger, and crimes of juniors against seniors were more serious than the reverse. Because the Qin code gave legal sanction to the authority of parents, a son's denunciation of his father could not be accepted as evidence, and the denouncer could be punished for making the accusation. A father could steal from his children with impunity, but if a grandchild beat his grandparents, he was tattooed and assigned to forced menial labor. The father could use the legal system to mete out punishments to his own family, even banishing or executing them. Routine, low¬level violence of fathers against children in the name of discipline appears to have been the norm. Wang Chong's first-century A.D. biography notes how remarkable it was that his father never flogged him. Later writings show that such beatings were a regular aspect of education in China for over two millennia.

One of the most striking features of Qin law to survive into the Han was "mutual implication" (lian zuo). Punishment for certain major crimes did not end with the individual but was extended to family, neighbors, and, in the case of an official, to superiors, subordinates, or the man who ,had recommended him for office. But the most important links were those among kin, and the range of collective punishments offers important evidence of the limits of kinship that were regarded as socially or legally significant by the state.

Collective punishment of kin was known by the technical term "destruction of the lineage" (mie zu). In the Spring and Autumn period "destruction of the lineage" had referred to a political event in which one noble clan eliminated another through killing or enslaving thousands of people. The phrase began to change meaning in the Warring States as indicated in the buried covenant texts, where it referred to the destruction of individual families who violated the terms of the oath. Shortly thereafter it came to include the punishing of family members for failures of their kin in military service. Thus, by the Qin and Han dynasties, "destruction of the lineage" had become a legal tool in the government's effort to delimit, register, and control the individual households that were the foundation of the state's power.13

This extended familial liability for crimes paralleled in significant ways a moral obligation to avenge kin, as prescribed in the classicist texts. First, the practice of vengeance entailed collective responsibility. A memorial written in the early Eastern Han described how reciprocal vengeance among the people led to the destruction of entire families. In some cases an avenger did not kill the actual offender but rather his wife and children or one of his other relatives. Sometimes a relative of the offender would voluntarily present himself to the avengers in the hope of rescuing his kinsman. Since the obligation to avenge was based on ties of kinship, the roles of avenger and victim were played by the collective kin unit rather than isolated individuals.

The state's use of collective punishment mirrored or inspired this social practice, and the groups delineated by punishments in law were those who acted for revenge in local feuds.14 Thus collective legal liability was to some degree the state's mode of participation in a vengeful society constituted through reciprocal blood debts. In a world where people's kin were obliged to avenge them, anyone who killed a person had best destroy the family too, and the state could not exempt itself from this kill¬or-be-killed rule. In many cases the targets of revenge were officials who had performed legal executions.

Collective responsibility was not just a means for the state to terrify people through the enormity of the punishment but also a method to se¬cure their participation in mutual surveillance. If kin or neighbors re¬ported the crimes of those to whom they were linked, they not only es¬caped punishment but could receive rewards. The early imperial state sought to control a large population with a small number of officials; so, as noted by the reforming Qin chief minister Shang Yang, governing the state depended on the participation of the entire population.15

Through collective responsibility and mutual surveillance the state sought to fashion a people who would actively enforce the legal dictates of their masters. In such a system the people would judge their fellows; or, more precisely, they would judge those to whom they were bound in kinship or membership in local units of collective responsibility. In the state advocated by Shang Yang, a man would stand in judgment of those whom, in the world of the classicists, he would be obliged to avenge.

The punishments in the legal texts also suggest a social hierarchy constituted through reciprocal obligations of gift and debt. Rather than stipulating fixed punishments, the law indicated that each crime produced an obligation to be redeemed through the performance of certain acts or certain payments. People of different status had different means for paying off the obligation created through violation of the legal codes. The status of the individual determined the appropriate punishment as much as did the nature of the crime. In the Qin code and in the Baoshan materials, the underlying principle was that the relation of the culprit to the ruler determined the form of punishment. This relationship was marked by the receipt of gifts from the ruler that entered into reciprocal exchanges built into the penal codes.16

The clearest example is the granting of titles of rank. These ranks were received from the ruler in exchange for service, at first for military service and then for the payment of sums to the state. Titles were the chief reward that common people could receive, and holders of titles could surrender them for a reduction of punishment.17 This meant that any crime committed by a title-holder would be punished less severely-or more precisely, that the title could be returned to the ruler in exchange for a reduction of the normal penalty. This exchange of status for a lesser punishment was worked out in a highly elaborated form in the Qin legal order.

Under the Qin code, violations by officials were usually punished with fines, measured as a certain number of suits of armor. These individuals, already protected by their titles, thus enjoyed the added privilege of redeeming any misconduct with forced contributions to the army. Most scholars assume that this was a financial penalty, since money had to be relinquished in order to purchase the suits of armor. So not only could those who received titles from the ruler use them to redeem their crimes, but those who received cash in the form of a salary from the ruler could return it to him in order to escape servitude. This principle is not absolute, however, for individuals without office accused of cash-related crimes or failures in military duties also paid fines in the form of armor. People with neither rank nor office could redeem crimes only through labor service for a fixed term or through state slavery. Once again, the punishment reflected the ties of the individual to the ruler, for labor service was the primary obligation of a peasant.18

Mutilating punishments, which dated back at least to the Shang and Zhou, fell outside the scheme of status-based reciprocity because they could not be avoided through payments. But even for these archaic punishments, graded degrees of mutilation were adapted to the position and privileges of the criminal. Mutilations, ranging from shaving beards and hair (a temporary stripping away of manhood) through tattooing and cutting off the nose or foot to castration or death, were used to make minute distinctions in the scale of punishments and balance them against past rewards.

In addition to receiving titles as a mark of their incorporation into the state, common people also received the gift of a family name. In the Zhou state a family name had been an aristocratic privilege, but in the Warring States this was extended to commoners. Consequently, the term "Hundred Surnames" (bai xing) that had previously indicated "the aristocracy" came to mean "commoners." Membership in society, as defined by law, entailed being registered within a lineage and household. The "Models for Sealing and Investigations" from the Yunmeng legal documents states that any testimony must begin with the name, status (rank), and legal residence of the witness. Even the earlier cases from Baoshan provide many examples of this practice.19

Once the common people were granted ranks and family names like those that had previously defined the nobility, the meaning of these attributes changed. Just as the process of universalizing military service had transformed a mark of nobility into a token of servitude, so the universalizing of naming, titling, and registration stripped these attributes of their authority. When the Zhou nobility received titles and surnames from the king, they inscribed these in reports to their ancestors, who granted them power and status not wholly dependent on the king. By contrast, the names and titles of the Warring State peasant, although showing status as a legally free member of the political order, were inscribed on registers whose ultimate recipient was the ruler of the state. To have one's name and title inscribed meant being a subject under the law.

Population registers, and the maps associated with them, came to represent authority in a way that transcended their power as legal documents. They magically embodied the people and territory they represented. For example, when Jing Ke attempted to assassinate the First Emperor, his pretext for appearing at court was to surrender territory to Qin through the formal presentation of the relevant population registers and maps. The force in law of such documents has been demonstrated by a tomb text entitled "Statute for forming arable plots" dated to 309 B.C. 20

By the late Warring States period, population registration had also become an element of religious practice. The clearest evidence is a story found in a third-century B.C. Qin tomb at Fangmatan, which tells of a man who committed suicide to avoid the disgrace of an unjust execution. Legal petitions addressed to the "Master of Life Spans" led to his body's being exhumed at the command of a scribe of the god and gradually re¬stored to life.21 This story depicts an underworld bureaucracy which keeps registers and communicates with mortals in texts based on terrestrial legal documents. It is also significant that the underworld administration was invoked to correct a death resulting from failings of earthly legal procedure.

"Spirit registrars" controlling the human life span also figure in the Warring States texts Master Mo and Words of the State, and the early Han philosophical compendium the Master of Huainanzi. The fourth¬century B.C. Chu silk manuscript likewise contains evidence of a bureaucratic pantheon.22 An even more elaborate underworld bureaucracy operating through codes and laws appears in Han tomb texts for the control of the spirit world.

Qin texts also discuss the handling of documents, the administration of granaries and storehouses, and the loaning of grain or oxen to peasants. Han wooden strips from the northwestern sites of Dunhuang and Juyan provide details of administrative concerns and procedures, including rules for keeping records and making reports, maintaining equipment, annually testing soldiers in archery (with rewards for good performance), obtaining a certificate for travel, granting leave so that a soldier could return home to bury a parent, paying taxes, and issuing circulars describing wanted criminals.23 All these documents show the pervasive role of law in early imperial administration.

Law and Language

As James Boyd White has elaborated, law is not only a system of rules and sanctions but also a specialized form of language and rhetoric. Legal systems generate their own technical vocabulary and usages, the mastery of which is essential for participation in the legal process.24 Because local Qin and Han officials were also the chief judges in their districts, they were required to learn the distinct linguistic usages of the legal code. Similarly in the central court, those in charge of drafting legal regulations had to become specialists in legal language. However, the links between law and language in early imperial China were not simply a matter of mastering technical language but of controlling society through the regulation of language.

The most elaborate accounts of the relations among law, administration, and language were those under the rubric of "forms and names" (xing ming), as discussed in the Qin text Master Han Fei.25 The ruler is advised to hold himself quiet and allow the ministers to describe the administrative tasks they will accomplish (that is, to "name" themselves). These declarations should be written down to provide a "tally" or "contract" against which performance will be measured. Later, if the claim and performance match, then the official will be rewarded; if not, he will be punished.26 This prescription for the use of contracts and annual verifications in the Master Han Fei corresponds to descriptions of actual administrative practice.

In both theory and practice, the use of rules and punishments is based on the matching of language to reality. Successful law and administration stemmed from the correction of words so that they matched actions. A related idea had appeared earlier in the theory attributed to Confucius of rule through the "rectification of names." In reply to a question of what policy he would advocate for the state of Wei, Confucius stated that he would begin by rectifying names. If names were not correct, then words would not correspond to the world, affairs would not be regulated, rites and music would not flourish, punishments would be incorrect, and the people would not know how to behave. As this list is clearly sequential, the correct use of names is the foundation of the system of rites that in turn makes possible the adequacy of legal punishments and social order.27

This idea developed in several ways, but the one most relevant to a theory of law and language was elaborated in the Gongyang Commentary to the Spring and Autumn Annals. It read the Annals as a coded text in which the choice of title or the mention of a person's name indicated judgments equivalent to a ruler's rewards and punishments. In this interpretation, the text was the blueprint for an imaginary kingdom of Lu, which Confucius ruled as the "uncrowned king" through a legal system consisting entirely of the correct application of names.28 This mode of reading the text explicitly identified the power of rectified language with that of legal judgments. Law, in this tradition of commentary, was the quintessential expression of the social powers of language.

By the first century of Han rule, most scholars at the imperial court had accepted this idea of the Annals as the textual expression of law as perfected language. When asked why Confucius wrote the Spring and Autumn Annals, the historian Sima Qian quotes the classicist scholar of the Annals Dong Zhongshu (ca. 179-104 B.C.): "When the Way of Zhou declined and was abandoned, Confucius was chief magistrate in Lu ... Confucius knew that his words would not be employed and his way not put into practice. He judged the rights and wrongs of 242 years [the span covered in the Annals] to create a standard for the world. He criticized the Son of Heaven, demoted the feudal lords, and punished the high officials, all in order to carry out the tasks of the king." By the Eastern Han, a text attributed to the same Dong Zhong shu taught how to use the coded judgments of the Annals to make actual legal decisions. To the ex¬tent that this and similar texts were used, the identification of the Annals as the textual form of the purified language of law had become a political reality.29

Linking the Annals to Confucius' supposed role as a judge is significant. Law, along with religion and government, is one of the spheres in which language most frequently plays a performative role. The declarations of the judge, like those of the official or the priest, produce social realities. His pronouncement of guilt creates the fact of guilt in a way that the words of ordinary people cannot. When the Annals assigns guilt or innocence, or grants or withholds a certain status, it acts as judge. Parables in the fourth-century B.C. Transmissions of Master Zuo also treated historians as judges. This notion that historians pass judgment on the past is commonplace in the West, but the Gongyang Commentary and its Han followers treat the figure of speech as a literal reality. Confucius' supposed role as a judge and his imposition of punishments was central to the myth that had developed around him by the early Han.30

Law as purified language figured not only in philosophy, commentary, and history but also in the legal texts. The clearest examples of this appear in the Qin legal documents from Yunmeng. Many passages on these strips consist of definitions for technical legal terms. In discussions of the family as a legal entity they gloss terms dealing with the household: "'Household' [hu] means 'those who dwell together' [tong jul. They implicate dependents [Ii, servants and slaves], but dependents do not implicate the household. What is meant by 'people of the house' [shi ren]? What is meant by 'dwelling together' [tong ju]? 'Dwelling together' means only those on the household register. 'People of the house' means the entire household, all those who would be mutually implicated with a criminal. " 31

From other Han sources we know that these terms were also part of conventional language. A decree of Emperor Hui (r. 195-188 B.C.) uses tong ju in semantic parallel with the graph jia "family." This shows that the terms were synonyms, both meaning "family" in a loose sense.32 Similarly, hu in Han texts has the sense of "household" or "family." What the legal documents do is impose a technical definition on words and phrases that possessed a range of meanings in conventional speech. In the example above, shared residence rather than blood ties defined the legal "family" for the Qin government, whatever might have been the conventional usages of the period. Similar passages from the strips give technical meanings to other graphs, or stipulate whether a given person or case fits into a particular legal category. These are practical manifestations of the philosophical vision of law as a form of rectified language that imposes social order.

Sima Qian's chapter on legal specialists at the imperial court points out the drawbacks of this vision. Speaking of "cruel officials" who use law to impose state authority on powerful families, he quotes philosophical pas¬sages on the limits of law and the counterproductive nature of relying on it. He invokes the by-then standard argument against the Qin empire that the severity of its laws and the ruthlessness with which they were imposed led to Qin's collapse.33 (His animus against rigorous use of the law was not unrelated to his being castrated for arguing that the letter of the law should not be imposed on a general who had been forced to surrender to the Xiongnu.)

Sima Qian then paints a largely negative portrait of officials who employed the full severity of the law. One of the recurring formulas by which he describes them is that they "applied the letter of the law and made no exceptions even for the emperor's in-laws." Others are "too severe in applying the letter of the law" or "manipulate the letter of the law to devise ways to convict people."34 In short, one of the bases of his critique was that law was a rigorous language which gave power to those who mastered its subtleties and permutations but did not always achieve justice as he or others perceived it. Like the texts cited above, Sima Qian's negative view nevertheless defines law as a distinct form of technically regulated and hence powerful language.

The central figure in Sima Qian's chapter is Zhang Tang. This man, who for a period dominated the court of Emperor Wu, was a specialist in the drafting of legal documents. Like others in the chapter, he is accused of excessive severity in imposing the letter of the law and of manipulating the technical language of legal documents to secure the results he desired. However, his biography also contains distinctive features. First, it includes a childhood narrative that identifies him as a prodigy in the use of legal language and procedure.

Once his father went out and left Tang, who was still a child, to mind the house. When he returned he found that a rat had stolen a piece of meat. He was enraged and beat Tang for his negligence. Tang dug up the rat's hole, caught the rat, and recovered the remains of the meat. He proceeded to indict the rat, beat it until it confessed, write out a record of its testimony, compare these with the evidence, and draw up a proposal for punishment. After this he took the rat and the meat out into the courtyard. There he held a trial, presented the charges, and executed the rat. When his father saw what he was doing and examined the documents, he found to his astonishment that the boy had carried out the entire procedure like a veteran prison official. 35

Such accounts of childhood deeds that reveal character and future career featured in many early Chinese biographies, notably that of Confucius, who as a child played at laying out ritual vessels. The story of the boy Zhang Tang reads almost as a parody of such accounts, but it also treats law as a distinctive, life-defining activity marked by the preparation of documents and the formal correctness of language. Moreover, it presents the standard forms of investigation and interrogation as described in the Qin legal documents (and deals with the thieving of rats, which similarly figures in Qin laws).

The second notable feature of Zhang Tang's biography is that once he recognized the emperor's fondness for classical studies and literature, he began to support his decisions with quotations from the classics. He even recruited scholars who were thoroughly versed in the Canon of Documents and the Spring and Autumn Annals to act as his secretaries so that he could cite these works to greatest effect. Sima Qian notes that many of these students of the classics became his fiercest "claws and teeth" in the rigorous imposition of penalties.36 Here the theory of Confucius' writing as a mode of legal judgment and the theory of law as a technical language of government converge. The same pattern also figures in other biographies, such as that of the leading classicist scholar of Emperor Wu's court, Gongsun Hong. This demonstrates the close ties between the technical languages of law and of commentary on the classics that developed under the Han.

Another passage that illuminates, from a negative angle, the importance of legal language is the Eastern Han historian and poet Ban Gu's (32-92 A.D.) complaint about the length and complexity of the Han code as it evolved across the centuries. Having supposedly been introduced as a radically simplified version of the Qin code in 200 B.C., by the late first century A.D. it had grown into tens of thousands of articles totaling more than seven million characters. Ban Gu asserted: "[Legal] documents filled tables and cupboards, so even knowledgeable officials could not examine them all. Therefore local officials disagreed, and sometimes the same crime elicited completely different judgments." 37 As in many cultures where elaborate legal systems have evolved, Ban Gu insisted that the search for a rigorous language to cover all cases inevitably broke down due to the multiplication of possibilities.

Law and Punishments

Early imperial law, in the narrowest sense, was a set of rules for behavior and of punishments imposed for their violation. Both Chinese and Western scholars have long noted that early Chinese writers regarded punishment as the defining aspect of law. The monograph on law in Ban Gu's Book of the Han is entitled "The Monograph on Punishments" and devotes its first half to the history of the military as the highest form of punishment.

Early Chinese law was largely a system of criminal law, but legal disputes over property and inheritance can be found in this period. The Han will discovered in a tomb in modern Jiangsu shows official participation in the inheritance of property, a fact already suggested by "grave purchase agreements" buried in tombs. Although these were religious documents intended to secure the position of the deceased in the afterlife, we can now confirm that they were patterned on actual practice. The legal texts also deal with such issues as the disposal of the dowry of a married woman convicted of a crime, the rights of slaves to marry, and the relations of parents and children. Nevertheless, the transmitted discussions of law and the bulk of our evidence deal with criminal law and its punish merits.38

The early imperial state used three primary types of punishments: the death penalty, physical mutilation, and hard labor. Imprisonment was not a punishment in itself but rather a means of detaining suspects and witnesses during the legal process. However, imprisonment could last for a long period if the legal process dragged on, and at least one of the legal specialists described by Sima Qian contrived to hold suspects indefinitely in prison if he thought that the emperor did not desire their execution. Ban Gu also notes that if officials found a case doubtful and could not come to a decision, they would hold people in prison indefinitely.39 Without possessing any legal grounding, protracted imprisonment was a fact of life.

The highest sanction was capital punishment, of which the most common form was beheading. This was often followed by exposing the head, the corpse, or both in the market. Much rarer was the practice of cutting the condemned in half at the waist by using a large blade hinged on a block. For particularly heinous crimes, which included the execution of three or nine (magical numbers indicating totality) sets of relatives, the primary culprit was subject to all punishments: first he was tattooed, then his nose cut off, then both feet, and finally he was beaten to death, after which his head was hung in the market and the corpse chopped to pieces in the same place. The number of crimes punishable by execution in the Han was very high, though amnesties or commutation into military service at the frontier were also frequent outcomes of capital crimes.40

The second category of punishments, mutilation, was much debated and modified over the centuries of Han rule. Originally these punishments had consisted of tattooing the criminal's face, cutting off his nose, amputating one or both feet, and castrating him. In 167 B.C. all of these except castration were formally abolished and replaced by punishments that would not leave the convicted party marked and maimed for life. Tattooing was replaced by shaving the head, wearing an iron collar, and doing labor service. Cutting off the nose or the feet was replaced by a specified number of strokes on the back or buttocks with the bastinado (cane). However, the weight of the bastinado and number of blows usually led to death through abundant bleeding, so what had been proclaimed as an act of mercy in fact represented a significant increase in severity.

The number of strokes and weight of the bastinado were consequently reduced in 156 B.C. and again in 151. But as these punishments grew bearable, officials complained that they could not deter violations. In response, the number of crimes punished by death was increased, so that there were more than a thousand capital crimes in the later Han. Castration may have been abolished as an independent punishment either be¬fore or shortly after the abolition of the other mutilating punishments, but it continued to be used sporadically as an exceptional form of deathsentence commutation (which was normally military service at the frontier). It was again abolished in the second decade of the second century A.D. and does not appear to have been used after that date.41

The most common punishment in early imperial China was hard labor ranging from one to five years. Each duration was identified by a particular task, such as "wall building and keeping guard," "gathering fire¬wood for sacrifice," or (for women) "pounding grain." These archaic categories did not describe the actual work performed, however, which included constructing roads and bridges, serving in government workshops, diking rivers, transporting grain, and producing iron. All periods of service were preceded by beating with the bastinado, and the more serious ones involved temporary mutilation. A man sentenced to a five-year term had his head shaved and wore an iron collar, while those sentenced to four years had only the beard and mustache shaved.42

Another type of punishment commonly used in the Qin was banishment. This punishment may have been frequently invoked because Qin had underpopulated and newly conquered regions that benefited from sending people to live there. A striking case appears in the "Models for Sealing [Houses] and Investigating": "A, a commoner of X village, said in his denunciation, 'I request to have the feet of my own son C, a commoner of the same village, fettered and to have him banished to a border prefecture in Shu [Sichuan], with the injunction that he must not be allowed to leave the place of banishment.'" The official approved the request and provided detailed instructions on how the son was to be escorted. Banishment became less common in the Han, except for those reprieved from death sentences and sent to the frontiers.43

In addition to amnesties, many punishments in Qin and Han China were eligible for redemption, a practice to which the Chinese applied the same term as a slave buying his or her freedom. The Qin legal documents frequently refer to the possibility of redemption from banishment, hard labor, mutilation, castration, and even the death penalty.44 In most cases redemption took the form of a surrender of titles or the payment of a fine. If the fine could not be paid, redemption might be imposed in the form of a period of labor service. Fines in the strict sense were levied primarily on officials for failures in the performance of their duties, with the fine being measured in terms of the purchase of suits of armor, as we have seen. In the Han, payments were conventionally made in gold.

It is difficult to generalize about punishments during this period. On the one hand, executions, mutilations, and even the treatment of hair and clothing that substituted for mutilations entailed a strong element of public display of the state's power over the bodies of its people, as well as the separation from the human community of those who committed major crimes. On the other hand, many punishments took the form of an ex¬change of services, money, or ranks that was carefully articulated in terms of differential relations. Using labor service as a form of punishment, as well as banishing convicts to frontiers or newly conquered regions, shows that punishments also provided human resources to the state. This is not unlike the use of galley slaves and chain gangs in early modern Europe or the United States. Although Han writers insisted on the disastrous consequences of Qin's excessively detailed regulations and cruel punishments, the records from their own period do not suggest that the Han made any significant improvement in these areas.

The essential elements of judicial procedure in early imperial China are fairly well documented. There was no sharp distinction between administrative and legal authority, and in general the chief administrator of a region was also the chief judge: the district magistrate was judge of his district, and the commander of his commandery. A general had the power to punish his troops, even to execute them. Stone inscriptions and textual references tell of specialist legal counselors assisting officials at the local level, but they provide no details.

At the central court, several officials handled legal cases. The superintendent of ceremonies, as chief administrator of the towns associated with the imperial tombs, was high judge for the area around the capital. The primary legal official, however, was the commandant of justice. He acted as judge in all cases pertaining to the imperial house, feudal kings or marquises, and high officials. He also resolved doubtful cases sent up by local officials. Finally, the emperor himself was highest judge, ultimate arbiter, and the source of all laws. While most emperors were happy to delegate to legal officials the actual work of rendering judgments, they could personally intervene or empower specific officials (such as the "cruel clerks" cited above) to act on their behalf.45

The central court also supervised the judicial activities of local officials.

One method was granting convicted parties or their families the right to appeal judgments, although we have no records of the procedure. The court also established traveling regional inspectors in 106 B.C. Among the matters that these men examined was the fairness and impartiality of the judgments in each administration. They were particularly charged to search for collusion between local officials and the great families of their districts or commanderies. Most of the legal specialists were ordered to attack these families and curb their influence. The court's worry about such collaboration of its own agents with regional powers was not unfounded. In the cities, the equivalent cases of corruption and collusion usually involved gangsters. Han histories tell of protracted collaboration between local officials and gangsters and of the destruction of these illicit leagues by agents from the court.46

From scattered references, of which the most detailed example is Zhang Tang's prosecution of the rat, scholars have reconstructed a rough picture of the conduct of legal investigations and trials. More detailed information has been obtained from Qin legal documents, particularly the "Models on Sealing and Investigation." These consist of boilerplate forms showing local officials how to write up reports of investigations and trials, which were sent to the central government. While these models may be based on actual cases, they were used as a procedural guide for local officials.

Many of the models insist on a forensic examination of the scene of the crime and the victims, specifying the details to be studied. A model case on a death by hanging gives a detailed description of the house, the location of the body, the type and size of the rope, the disposition and state of the different body parts, the clothes worn by the victim, the size of the beam from which he was hanged, and even the condition of the soil that prevented footprints. This is followed by a general injunction on methods of investigation:

When investigating, it is essential carefully to examine and consider the physical traces. One should go alone to the place where the corpse is and consider the knot of the rope. If at the place of the knot there are traces of a noose, then observe whether the victim's tongue protrudes or not, how far the head and feet are distant from the place of the knot and the ground, and whether he had discharged feces and urine or not. Then untie the rope and observe whether mouth and nose emit a sigh or not. Then observe the condition of the blood congealing along the trace of the rope. Try to free the head from the knot in the rope. If you can free it then [text missing] his clothes and completely observe his body, from underneath the hair on his head down to the perineum. If the tongue does not protrude, if mouth and nose do not emit a sigh, if there is no congealed blood along the trace of the rope, if the knot in the rope is so tight that it cannot be slipped off [text missing] Since he has been dead for a long time, the mouth and nose may not produce a sigh. People who kill themselves must first have had reasons. Question the members of his household so that they will reply as to the reasons.47

Other accounts include the examination of a tunnel used to break into a house, the study of a fetus and the body of the mother when a fight has induced a miscarriage, and the consultation of a medical specialist to determine which diseases could produce a given physical state.

In addition to investigations of the material and bodily traces of a crime, many cases involved accusations by neighbors or family members. Suspects were arrested by the posthouse chief or constabulary, often retired military men who served the local government. The local administrator then interrogated witnesses and accused:

In all cases of interrogating one should first listen fully to their words and note these down, letting each of them set out his or her state¬ment. Although the investigator knows that one is lying, there is no need to question pressingly at each point. When a statement has been completely taken down and does not cohere, then question pressingly on the required points. Having questioned pressingly, one listens to everything, notes down the explanatory statements, and then looks further at other unexplained points and questions pressingly on these. When one has questioned pressingly to the ultimate limits of the case, but the suspect has repeatedly lied, changed testimony, and not confessed, then bastinado those whom the stat¬utes allow to be bastinadoed. When you bastinado, be sure to note down: "Report-Because X repeatedly changed testimony and provided no explanatory statement, X has been interrogated with the bastinado.48

Physical evidence and testimony converged in a grand confrontation of all concerned parties and the investigating official. Each party gave an account, and the investigating official combined these with the physical evidence to provide a coherent and satisfactory version of events. At points where accounts or physical traces conflicted, further questioning was applied. Changes in testimony or internal contradictions provided the leads for further interrogation. Ultimately, the confession of the accused was necessary, and obdurate refusal to confess in the face of the evidence would lead to beatings and perhaps other forms of torture. However, beating was only a last resort, and its use had to be noted in the report.

In this procedure the official remained silent but kept an exact account of the witnesses' words and then matched the accounts for internal consistency and their relation to material evidence. This corresponded to the ruler's use of "forms and names" to control his ministers. Here, the local official stood in the position of the ruler, and the witness in the position of the minister. The witness named himself and gave an account of his actions, which were transformed into writing, just as the aspirant officer did in the presence of the ruler. Meanwhile, the local official sat silently

and measured everything against his knowledge of the facts. This model of administration through a hierarchical series of staged personal en¬counters moving from the local level to the central court was fundamental to early imperial government.

These texts also show the idea in Chinese legal theory and practice that the judge should be a detective. Through his ability to read the hidden meanings of physical traces and human speech, the ideal judge penetrated the veil of confusion and deceit to arrive at a truthful account of events and a just disposition of the concerned parties. This vision of the judge as a sagely reader of signs also figures in a case found in a Han tomb at Zhangjiashan, wherein an official assigned to discover who had contaminated a meal deduced the failings of an entire household.49 The judge as a detective figured prominently in later Chinese theater and fiction, where Judge Di (Dee) or Judge Bao became stock embodiments of the aspiration for truth and justice. As the new Qin and Han materials have revealed, this model of justice as the result of sagely ability to read the meaning of signs had already emerged at the beginning of the imperial period.

Law and Labor

Under the Qin, several categories of lawbreakers, along with other social delinquents such as merchants, were forcibly dispatched to the frontiers. Under the Han, as frontier garrisons were increasingly filled with reprieved convicts, physical transfer to the frontier for military service be¬came conflated with the banishment of criminals. Military action at the frontiers thus became a means of exporting internal violence and pacifying the interior. The army served not only to suppress external enemies of the imperial order but also to expel disruptive members of society to regions that in the Han imagination lay beyond the reach of civilization.50

The evidence for this comes from many sources. In 109 B.C. Emperor Wu recruited men condemned to death for use in an expeditionary army, and further uses of convicts in expeditions are recorded in 105, 104, 100, and 97 B.C., as well as under later emperors. Emperor Wu's decree of 100 B.C. stated that convicts were to establish a base at Wuyuan. Han wooden strips also show that, from the time of Emperor Wu, convicts were stationed in frontier garrisons, and they increased as a percentage of the total forces throughout the Western Han.51

Sending Convicts to the war front

Following the abolition of universal military service by the Eastern Han, the use of convicts dramatically increased. This was not simply an increase in quantity but also a qualitative shift in terms of organization. In 32 A.D., the year after the abolition, the Eastern Han founder Guangwu established the office of the left inspecting commandant of the black turban. Several biographies show that this officer's staff administered punishments. The title also indicates a military role, and in 91 A.D. its holder was indeed sent out on an expedition. A petition referred to a man placed under the commandant as a "reprieved convict," the formula applied to convicts sent to the frontier for military service. Similarly, a memorial requesting the pardon of several former officials called them "reprieved convicts of the left [inspecting] commandant." These show that those placed under this office were convicts who had their death sentences commuted and were sent to the frontier as garrison troops. Thus in addition to those reprieved in the decrees, convicts were routinely sent to the frontiers through this office.52

In addition to this evidence of an office devoted to sending convicts to the frontier, documents indicate that garrisons were manned largely by convicts. In 45 A.D. Guangwu established three camps with associated fields at the frontier and ordered that these be filled with reprieved convicts. When Ban Chao returned to court in 102 A.D. after thirty-one years campaigning in Central Asia, a friend apologized that the veteran commander had not received a higher office. In reply, Ban Chao observed that he was not fit for an office at court: "The officers and soldiers beyond the frontiers are not filial sons and obedient grandsons. They have all been transported to fill the frontier camps because of their crimes.53

A memorial by Yang Zhong in 76 A.D. indicates the scale of this practice. He wrote that, since the beginning of the reign of Emperor Ming in 58 A.D., officials had repeatedly scoured the prisons and condemned innocent men in order to obtain recruits for the frontier; those sent out numbered in the tens of thousands. The "Fundamental Chronicles" and biographies indicate an average of one decree every five years up to 154 A.D. sending pardoned convicts to the frontier. Although no decrees are listed after 154, biographies cite individuals sent as convicts to the frontier after that date. 54

We cannot calculate the exact number of men transported, but one can get an idea of the order of magnitude from figures in the "Monograph on Punishments" in the Book of the Han: "Now in the commanderies and princely states, those executed annually are numbered in the tens of thou¬sands. In the empire there are more than two thousand prisons, and the bodies of the unjustly slain pile one on the other." 55 The phrase "numbered in the tens of thousands" is a literary trope, although it echoes the statement of Yang Zhong. However, the figure of two thousand prisons looks more like a genuine number. If it is correct, and given the high number of crimes punishable by death under Han law, officials could easily have provided tens of thousands of convicts for frontier service. In 87 A.D. an official named Guo Geng sent up a memorial asking that those who committed a capital crime before an amnesty but not captured until afterward ought also to be sent to the frontier. Such men, he argued, numbered in the tens of thousands. He also gave the clearest explanation of the rationale of the system, arguing that it both "preserved life [through pardons] and benefited the frontiers."56 Ban Chao's remark that all the soldiers at the frontier were transported convicts might well have been no exaggeration.

Slave labor

Forced labor was the foundation of the Qin and Han states. Monumental public works such as palaces, temples, and imperial tombs-as well as more overtly utilitarian projects such as canals and roads-required all sorts of skills. Unskilled manual toil figured prominently in leveling ground, moving mountains, piling up earth, and so on. Some tasks were extremely dangerous and resulted in numerous fatalities, particularly in the iron casting foundries that were run by the state monopoly in certain periods. For such types of work, convict labor was essential to the state.57

The early empires employed four types of manual labor: peasant corvee, hired, convict, and slave. Each of these had different legal and social characteristics and was consequently suitable for different types of work. Adult males in free households owed one month's labor per year. Such work was devoted to diverse tasks, and the legal texts mention repairing walls of government buildings, mending roads and bridges, excavating ponds, and digging or dredging canals. Corvee labor was most frequently employed in local projects such as flood control, irrigation, or roads, but it was also used to build imperial tomb mounds, construct walls around the capital, and repair breaks in the dikes of the Yellow River. However, such work crews changed each month, and peasants were unavailable during times of crucial agricultural work. If peasants were forced to work away from their native area, the state provided food and tools. Consequently, the use of corvee labor could lead to costly delays in major projects.

Given these limitations, convict labor became crucial to the state.

These men and women could work throughout the year until a project was finished. They could be transported across great distances to projects that might take weeks to reach, because once there, they could remain indefinitely. Most important, they could perform the most arduous and dangerous of jobs which might entail the deaths of thousands of laborers. With so many types of crime, and the entire empire to draw upon, the convict population provided a bottomless supply of expendable labor. Several Han tombs contain wall tiles depicting convicts in neck collars, wooden fetters, and iron manacles. One scene even depicts a group of convicts having their heads shaved.

The labor of convicts was as diverse as that of corvee workers, and in some cases the two groups worked together, although the convicts were distinguished by their shaved heads, red caps, and physical restraints. Many convicts were employed in building the tombs of emperors, which often required years of labor by tens of thousands of men. Inscriptions also show that the Qin employed convicts, some of them skilled artisans who had fallen afoul of the law, in factories that produced weapons. Under the Han, skilled labor was done mostly by wage-labor artisans, while convict labor was largely devoted to state iron and copper mines and metal casting facilities. Estimates based on scattered textual references and archaeological remains indicate that somewhere between ten thou¬sand and fifty thousand convicts were employed in the iron processing facilities of the state monopoly. Such work included the mining and the smelting of iron into ingots at the site to facilitate its being shipped across the empire. These activities were extremely dangerous, and deaths were frequent.

The lives of convicts were hard. They received an adequate diet of about 3400 calories per day for hard physical labor, but it consisted almost entirely of grain. They were beaten for the smallest infraction, although officials faced a stiff fine if convicts died within twenty days of a beating. Any further criminal violation by a convict warranted the death penalty. The physical condition of these men and women can be reconstructed in part from skeletons found in three large convict cemeteries dating from the Qin, the reign of the Han emperor ling (r. 156-141 B.C.), and the late Eastern Han (ca. 86-170 A.D.). All three cemeteries seem to have been used for decades, largely for the remains of those who died while building imperial palaces and tombs. In all three cases more than 90 percent were young men. Seven percent had died from the trauma of sudden blows, almost always to the skull. The jawbones and teeth from the Eastern Han cemetery show a high frequency of advanced gum disease and dental abscesses, probably due to malnutrition. Many of the skeletons still wore their iron collars and leg fetters. The Western Han collars weighed from two and a half to three and a half pounds and had a long spike that would have made it impossible for the convict to bend over very far without impaling himself. Presumably this was removed for certain tasks.

Most of the skeletons were accompanied by notations on brick or some other material with their name and sometimes fuller information on their place of origin, crime, rank, and day of death. These show that convicts were transported from all over the empire to work on these imperial projects. At one major imperial construction site, from one to six people died per day. This seems to have been considered an acceptable rate of mortality, because only once is there a record of an emperor complaining that too many men were dying to build his tomb.58

In addition to corvee workers and convicts, the state also drew on a pool of slave labor created through the enslavement of family members of convicted criminals. Archaeologically recovered Qin documents indicate that the state enslaved the women and children of anyone sentenced to three years of hard labor or worse, as well as those castrated for rape. The children of government slaves were slaves from birth. Finally, thou¬sands of prisoners of war were enslaved to the state. However, slaves in this period were largely employed in domestic duties such as cleaning, cooking, mending, running errands, and caring for animals. References to slaves in agriculture or industry are quite rare. 59

While there is no evidence that slaves worked state-owned land, a Han tomb at Fenghuangshan indicates that about a dozen slaves, mostly female, had been used in private agriculture by the tomb's occupant, and a few texts mention the use of private slaves in largescale craft production.But not a single surviving imperially produced Oin or Han object mentions a slave as its manufacturer; and other than the First Emperor, there is no indication that any emperor employed slaves in building his tomb. This suggests that for wealthy private individuals, who did not have access to convict labor, employing slaves in certain kinds of work could be more efficient than hiring laborers. For the state, however, convicts provided a steady supply of expendable, cheap labor, whereas slaves were permanent property not to be lightly discarded.

This preference for the use of convicts in the greatest imperial construction projects and in the crucial state monopolies shows that while slaves were legally the lowest form of humanity, it was upon convicts' expendable backs that the material foundations of the early imperial state were built.


1. Lewis, Sanctioned Violence, pp. 43-50; Liu, Origins of Chinese Law, ch.5.

2. Lewis, Writing and Authority, p. 20.

3. Lewis, Sanctioned Violence, pp. 67-80.

4. See the translation of the legal documents in Hulseswe, Remnants of Ch'in Law.

5. Shuihudi Qin mu zhu jian, pp. 281-293.

6. Ibid., pp. 26-27, 181, 182, 183, 261-262, 263.

7. Harper, "Warring States Natural Philosophy and Occult Thought," pp. 854-856; Harper, "A Chinese Demonography," pp. 47°-498; Katrina McLeod and Robin Yates, "Forms of Ch'in Law," n. 57; Zuo zhuan zhu, Cheng year 5, pp. 822-823; Guo yu, pp. 405-406.

8. Han Feizi ji shi, ch. 6, p. 357.

9. Yunmeng Shuihudi Qin rnu, strips 827 verso-Sr a verso, 886-895; Hulsewe, "The Wide Scope of Tao 'Theft' in Ch'in-Han Law," pp. 182-183; Yates, "Some Notes on Ch'in Law," p. 245.

10. Shi ji 122, p. 3148.

11. Mcknight, The Quality of Mercy, ch. 2.

12. Han shu 56, pp. 25°0-25°2; Hulsewe, "Ch'in and Han Law," pp. 522-523.

13. Hulsewe, Remnants of Han Law, pp. 271-272. Lewis, Sanctioned Violence, pp. 80-94.

14. Lewis, Sanctioned Violence, pp. 49-50, 91-94.

15. Shang Jun shu zhu yi, ch. 5, pp. 140-141.

16. For what follows see Lewis, Writing and Authority, pp. 23-26.

17. Shuihudi Qin mu zhu jian, pp. 92, 93-94,101-102,102-1°3,103,136¬147; Lewis, Sanctioned Violence, pp. 61-64.

18. Shuihudi Qin mu zhu jian, pp. 97,113-148; Hulsewe, Remnants of Cb'in Law, pp. 14-18.

19. Shuihudi Qin mu zhu jian, pp. 247-249; Baoshan Chu jian, pp. 17-39.

20. Hulsewe, Remnants of Ch'in Law, pp. 211-215; Sage, Ancient Sichuan and the Unification of China, pp. 131-133.

21. Harper, "Resurrection in Warring States Popular Religion," pp. 13-28.

22. Riegel, "Kou-rnang and ju-shou," pp. 57-66; Barnard, The Ch'u Silk Manuscript, pp. 2°7-210; Huainanzi, ch. 8.

23. Loewe, Records of Han Administration.

24. White, Heracles' Bow, The Legal Imagination, Justice as Translation.

25. Yates, "New Light on Ancient Chinese Military Texts," pp. 220-222; Wang and Chang, The Philosophical Foundation of Han Fei's Political Theory, pp. 59-60; Makeham, Name and Actuality, pp. 69-75.

26. Lewis, Writing and Authority, p. 33.

27. Lun yu zheng yi, ch. 7, p. 129; ch. 15, p. 271; ch. 16, pp. 280-293; ch. 20, p. 364; Makeham, Name and Actuality, chs. 2-4.

28. Lewis, Writing and Authority, pp. 139-144.

29. Shi ji 130, p. 3297; Han shu 30, p. 1714; Shi ji 122, p. 3139; Queen, From Chronicle to Canon, chs. 6-7.

30. Zuo zhuan zhu, Xu an year 2, pp. 662-63; Xiang year 25, p. 1099; Lewis, Writing and Authority, pp. 13°-131,222-224.

31. Shuihudi Qin mu zhu jian, pp. 160, 238.


Bibliography: (also used in other related articles on this website)


For updates click homepage here