← Back to 漢書

卷二十三 刑法志

Volume 23: Treatise on Punishment and Law

Chapter 25 of 漢書 · Book of Han
← Previous Chapter
Chapter 25
Next Chapter →
1
Human beings, bearing the form of Heaven and Earth and the nature of the Five Constants, are the most numinous of living creatures. Because humans lack natural weapons and coverings, they survive by organizing the world through intelligence, not brute strength; that is what makes them distinctive. Without humane solidarity, people cannot form society; without society, they cannot secure resources for life. When scarcity persists, conflict begins. The sages preempted this by modeling respect, deference, and universal concern, winning willing allegiance. Once people gathered under such leadership, that person became a ruler; When all under heaven gravitated to him, he became a true king. The Great Plan says: 'The Son of Heaven acts as father and mother of the people, becoming king for all under heaven.' Calling the ruler 'parent' emphasized that humane care and moral deference are the foundation of legitimate kingship. Affection alone is unstable; it must be paired with respect, and virtue must be backed by enforceable authority. Hence both ritual and penal law. Because sages understood both human nature and cosmic principle, they built ritual, education, law, and punishment around real human emotions while mirroring cosmic order. Therefore it is said former kings established rites by 'patterning Heaven's brightness and following Earth's nature.' Punitive law mirrors the coercive side of cosmic power—thunder, lightning, and lethal force; while benevolence and care imitate Heaven's life-giving and nurturing functions. The Documents says, 'Heaven's ordered ranks have rites,' and 'Heaven punishes the guilty.' So the sages derived both ritual classification and penal classification from Heaven's own ordering and sanction. At the highest level, punishment means military force; next comes execution by weapons; Middle grades involved mutilating instruments; Lighter ones used flogging. Major penalties were public in open ground, lesser penalties in urban spaces-a very old institutional pattern.
2
涿鹿
From the earliest legendary rulers, armed action was used to suppress catastrophic disorder, whether framed as fire or flood threats. Even under ideal sage rule, severe sanctions remained necessary against major disorder before universal compliance was achieved. Xia, Shang, and Zhou alike relied at key moments on military force to establish order. Even in peace they promoted civil culture while maintaining a structured military apparatus tied to land and taxation. Administrative-military units scaled up from the basic well-field: ten wells to a tong, ten tong to a cheng. Further aggregation produced larger territorial blocks up to one-hundred-li squares; and ultimately to the thousand-li royal domain. The system distinguished tax and rent obligations. Food revenue sustained population; military levy sustained defense. At lower levels, four wells formed a hamlet and four hamlets a larger unit. Each qiu of sixteen wells carried specified livestock obligations for war support. Four qiu constituted a dian. A sixty-four-well unit supported a full standardized military package of horses, chariot, draft oxen, officers, infantry, and weapons. In a hundred-li fief, after non-taxable terrain was removed, 6,400 productive wells supported 400 horses and 100 chariots-the upper scale of grand-official estates. At major feudal scale, assessed land supported a thousand-chariot military capacity. The royal domain's assessed capacity justified the classic title 'lord of ten thousand chariots.' Military readiness was seasonal and synchronized with agricultural downtime, preserving both farming and defense. Five states formed a lower cluster under a chief; Ten formed a larger command under a marshal; Thirty formed a corps under a chief officer; Two hundred ten constituted a regional circuit under a governor. The old system used regular inspection cycles to maintain force quality without abandoning institutional control.
3
As Zhou order eroded, Qi's reform under Guan Zhong restored wealth and stability. The duke asked about hegemony by military means. Guan Zhong said: 'If Your Grace wishes to fix ranks and repair armor and weapons, great states will also repair theirs, and small states will ready defenses, so quick success will be hard.' He embedded military organization within civil administration, creating readiness without provoking immediate counter-mobilization. By tightly integrating local units into shared daily life, he produced battlefield cohesion and mutual commitment under stress. Once internal discipline matured, Qi could defend frontiers, uphold royal legitimacy, and secure the central states. Jin's Duke Wen continued the model: domestic consolidation first, then interstate leadership. But hegemonic politics remained institutionally compromised and short-term, never achieving true kingly order. Post-hegemonic states distorted older military-ritual norms into ad hoc extraction and display. The Spring and Autumn Annals condemned these practices to preserve the normative ideal. Frequent campaigning wore down the populace and eroded sacrificial civic ethos. Confucius lamented it, saying: 'To send the people to war without teaching them is to abandon them.'
4
使 使
So he praised Zilu: 'You, You, in a thousand-chariot state, can be made to govern its levies.' And Zilu also said: 'In a thousand-chariot state hemmed between great states, pressed by armies and followed by famine, if I governed it, by three years I could make them brave and also know proper direction.' Military administration, in this view, must be inseparable from moral-political education.
5
使 使
In the Warring States era, military ritual became spectacle and competitive display. By Qin, old ritual forms had been recast as entertainments, their normative substance lost. Strategists such as Sun Wu, Sun Bin, Wu Qi, and Shang Yang turned military thought toward power and deception, with influential textual legacies. Alliance blocs constantly realigned, producing perpetual competitive warfare. Different states specialized in different military institutions to seek advantage. With utility dominant, military discourse canonized the Sun-Wu tradition. Only Xunzi was clear on the kingly way and criticized them, saying: 'Those Sun and Wu strategists esteem advantage and prize changing deception; Such methods work only in already disordered polities fractured by distrust and poor governance. Under genuine benevolent authority, social cohesion resembles familial protection; such a polity is not easily defeated. If we govern well, neighbors are drawn to us warmly, while resenting their own oppressive rulers. People do not willingly fight for what they detest against what they admire. Even tyrant-against-tyrant conflict leaves room for tactical skill; But trying to defeat true sage rule by trickery is like hurling an egg at rock. As the Odes describes King Wu's campaign, moral legitimacy generated unstoppable force. Rule grounded in benevolence and justice is strategically unmatched. Qi incentivized shock troops through cash-for-head rewards. Such troops may work against minor opponents; But against strong adversaries they quickly disintegrate. That is a military model fit for state collapse. Wei's elite infantry were heavily equipped and physically tested, then compensated through tax and property privileges. Such a force imposes heavy fiscal burdens and erodes demographic strength over time. That too is a dangerous military model. Qin organized society through strict coercive discipline and severe extraction. By combining pressure, reward conditioning, and punishment, Qin made military performance the sole path of advancement. Its quantified merit system produced sustained multi-generational battlefield success. But such forces were fundamentally transactional, lacking stable moral-institutional discipline. Hence even at peak strength, Qin lived in strategic fear of coalition encirclement. Huan and Wen improved discipline but still fell short of a fully benevolent-just political foundation. Thus Qi's technical fighters cannot meet Wei's martial graduates; Wei's martial graduates cannot directly face Qin's elites; Qin's elites cannot withstand Huan and Wen's disciplined system; Huan and Wen's discipline cannot rival Tang and Wu's benevolence and righteousness.'
6
使 滿
Therefore it is said: 'One good at commanding does not deploy; one good at deployment does not fight; one good at fighting does not lose; one good at losing does not perish.' When Shun ordered the hundred officials and Gao Yao served as judicial officer, tasked with 'barbarians troubling Xia, bandits and rebels,' yet punishment had no use—this is what is meant by one good at commanding not deploying. Tang and Wu's legitimacy made victory decisive once forces were arrayed, minimizing prolonged combat. Huan's record shows effective campaigning and interstate stabilization without strategic defeat. Even after catastrophic defeat under Wu pressure, King Zhao entered exile with popular support. The king said: 'Elders, return home! Why worry that there will be no ruler?' The elders replied, 'We have a ruler as worthy as this!' They all followed him together. Others appealed to Qin in tears, and Qin was moved to dispatch aid. With allied support Chu recovered, illustrating how political legitimacy can survive even major military defeat. Qin then pursued total conquest through accumulated advantage, strategic geography, and ruthless generals. By overreliance on coercion and deceit, Qin alienated society; its own lower ranks eventually joined the coalition that destroyed it. That is the lowest strategic outcome. The proper function of military force is restorative: preserving order, rescuing the endangered, and eliminating harm. Hence the great founders' strategists left enduring political lineages, their merit recognized alongside dynastic houses. In decadent times, rulers chased gain through fraud and violence, producing mass slaughter in both urban sieges and territorial wars. The lineages of Sunzi, Wuzi, Shang Yang, and Bai Qi ended in violent ruin, and their victories did not endure. Consequences answer in kind; that is how moral causation operates.
7
At Han's founding, Gaozu combined military charisma with political magnanimity, uniting talent to defeat Qin and Xiang Yu. By balancing bureaucrats, strategists, diplomats, and ritual specialists, Gaozu aligned civil and military governance into a coherent founding program. Once unified, Han retained Qin-style local military offices and permanent capital garrisons. After Emperor Wu's southern campaigns, Han expanded both land and naval forces and institutionalized regular training. Under Emperor Yuan, jiaodi spectacle was removed, but proper military-ritual discipline was still not fully reestablished.
8
An old saying says: 'Heaven produced five materials, and people use all of them; none can be discarded-who can remove arms?' Domestic discipline, state law, and military force each remain necessary in their domains; the issue is sequence and measure, not absolute presence or absence. Confucius said: 'If a craftsman wishes to do good work, he must first sharpen his tools.' For kingship, civil virtue is the primary instrument; military force serves as its supporting arm. The deeper moral influence runs, the broader military submission becomes; the wider virtue is applied, the wider coercive order can be stabilized. The Three Dynasties reached penal and military minimalism only after correctly ordering moral foundations and coercive means.
9
使使使使使
Zhou's tripartite penal doctrine prescribed lighter law for newly integrated polities; medium law for stable polities; and heavy law for chaotic ones. The so-called medium code enumerated graded corporal and capital penalties in standardized counts. Penalty types were tied to assigned labor posts, from gate and pass duties to granary and park custody. Male convicts became penal bond-laborers; females were assigned grain-processing labor. Rank holders, elders, and very young minors were exempted from enslavement.
10
As Zhou weakened, punitive legislation was expanded under King Mu through the code associated with Marquis Fu. The punitive taxonomy under this later code was far larger and harsher than earlier norms. This three-thousand-item structure exemplified heavy-law governance for disorder.
11
祿 使 使 使
As royal norms collapsed, Zheng's Zichan famously codified criminal law in public bronze text. Jin's Shuxiang criticized it: 'Former kings deliberated affairs and regulated, and did not make penal statutes public. To prevent litigious contention, governance had relied on layered moral-political discipline: righteousness, policy, rite, trust, and benevolence; Institutional incentives and decisive penalties worked together under a moral framework. Even so they reinforced order through constant ethical training, administrative harmony, and firm adjudication. They prioritized selecting high-caliber moral and administrative personnel. Under such conditions, the populace could be mobilized without social breakdown. Public legalism alone, he argued, breeds technical litigation and opportunistic evasion. Major penal codifications historically arose in periods of political disorder. Thus codified penal expansion marks late-decline politics, not high-order rule. Shuxiang warned Zichan that legal codification would not easily produce social pacification. As the Odes says, order comes by emulating King Wen's virtue. Likewise, moral modeling creates political trust across states. If virtue governs, why foreground penal text? Litigation culture will replace ritual culture. Petty technical disputes will explode, along with corruption in adjudication. By your generation's end, Zheng will likely fail!' Zichan replied: 'If it is as you say, I, Qiao, am unworthy and cannot benefit my descendants; I do this to save the present age.' Thereafter instrumental and morally thinned governance intensified. Confucius lamented it, saying: 'Guide with virtue and align with rites, and people will have shame and also rectify; guide with policy and align with punishments, and people avoid penalties but have no shame.' 'If rites and music do not flourish, punishments will not hit the mark; if punishments do not hit the mark, the people will have nowhere to place hands and feet.' When the Meng clan made Yang Fu judicial officer, he asked Zengzi, who likewise said: 'When superiors lose the Way, people have long been scattered. If one obtains the facts, then be sorrowful and compassionate, and do not rejoice.'
12
By Warring States times, collective-liability and extreme punitive innovations became standard in some states. Severe torture and execution forms proliferated.
13
After unification, Qin centralized legal bureaucracy under the emperor, suppressing ritual institutions and ruling through document-intensive penal administration. Hyper-penal governance produced mass criminalization, prison overexpansion, and ultimately revolt.
14
When Han arose, Gaozu upon first entering the passes made the Three Articles: 'Killers die; injuring others and theft are punished according to the crime.' Abolition of oppressive clauses won broad popular support. As security pressures rose, Han expanded from minimal law to Xiao He's Nine-Chapter code.
15
Early Han society sought recovery and household stability after Qin-era trauma. Under Huang-Lao style minimal interference, economic recovery accelerated and penal severity declined.
16
滿 滿
Chancellor Zhang Cang and Censor-in-chief Feng Jing memorialized: 'Corporal penalties forbidding wickedness have existed long. They acknowledged imperial concern that lifelong corporal penalties blocked moral rehabilitation. They proposed commutation schedules replacing some mutilations with hard labor terms; Tattoo penalties were converted to shaved-and-shackled labor; Nose-cutting became three hundred lashes; left-foot amputation became 500 lashes; A range of aggravated and official-corruption offenses remained capital, with public marketplace execution. Post-sentence labor moved through staged penal-service categories. One service stage transitioned convicts into state bond-servitude. Completion of term could restore commoner status. Extended service could shift them into judicial labor classification. Further service terms eventually led to formal release. Escapers and higher-grade offenders were excluded from commutation rules. Legacy penal terms were harmonized with the new commutation-release schedule. The ministers closed with the standard death-risk petition formula." The edict said: 'Approved.' In practice, nominal commutation often produced lethal outcomes. Some commuted categories remained effectively capital. Extreme flogging rates caused high mortality.
17
Under Emperor Wu, military expansion and fiscal extraction intensified social strain, while harsh enforcement failed to control rising crime. Legal technocrats expanded doctrinal categories and tightened accountability frameworks, especially around supervisory responsibility. Both officials and offenders exploited technical legal analogies, producing ever denser legal restrictions. Han law became massively complex in chapters, clauses, and precedent files. The legal archive grew beyond practical administrative mastery. Uneven implementation generated sentencing inconsistency across jurisdictions. Legal complexity enabled prosecutorial rent-seeking, with life/death outcomes manipulated through selective precedent.
18
使 祿 涿
Because Emperor Xuan knew grassroots legal abuse firsthand, memorial criticism of judicial bureaucracy resonated with him. The memorial text is preserved in Lu Wenshu's biography. The emperor deeply pitied this, and issued edict: 'Recently officials using law have grown deeply into clever wording-this is my lack of virtue. Misjudged cases both embolden true criminals and destroy innocents, causing profound social trauma. He created four 'Court Equalizer' posts to participate in local adjudication and improve balance. Their task is to level judgments, so as to accord with my intent.' He appointed capable judicial leadership and staffed equalizer roles with officials known for careful and lenient judgment. With direct imperial attention to adjudication, penal practice was for a time regarded as comparatively fair. At that time Zhuojun governor Zheng Chang memorialized: 'Sage kings establish remonstrating ministers not to ornament virtue, but to prevent the rise of indulgent ease; law is remedial for disorder, not the highest form of rule. Given active imperial oversight, he argued, adjudication could already be corrected; For the benefit of posterity, revising and fixing statutes and ordinances mattered more than temporary offices. A clear, stable code would both guide commoners and reduce discretionary abuse by officials. If fundamentals are left untouched and only procedural offices are added, those offices can eventually become new centers of abuse." Emperor Xuan never fully carried out structural reform.
19
便
When Emperor Yuan first ascended, he issued an edict: 'Laws and ordinances are meant to restrain the violent and support the weak, so they should be hard to violate and easy to avoid. The edict criticized legal overcomplexity that trapped ordinary people even when experts struggled to interpret it. Deliberate what can be remitted, removed, or reduced; submit by categories, with only the convenience and safety of the myriad people in view.'
20
便
In Chengdi's Heping years another edict said: 'The Fuxing says the Five Penalties have three thousand categories and capital punishment two hundred; now capital clauses are over a thousand, statutes and ordinances are cumbersome and many, over a million words, extraordinary petitions and analogies increase daily; even specialists do not know the basis-how can we explain this to the masses? The court explicitly linked legal complexity to wrongful death. A broad review group was convened to reduce capital punishment and simplify statutes. The edict cited scriptural authority: 'Be cautious in punishment. Examine and verify carefully, striving to align with ancient law; I will personally review with full attention.' Because officials lacked high-level statecraft, they answered reform edicts with piecemeal technicalities instead of systemic restructuring. As a result, major legal reform repeatedly stalled. The text rejects conservative legal inertia as a recurring obstacle to effective governance. It then summarizes instances where Han law moved toward both classical legitimacy and practical utility.
21
Classical Zhou jurisprudence included multi-layered evidentiary and mitigation frameworks. The Five Hearings evaluated testimony through words, expression, breath, auditory clues, and visual observation. The Eight Deliberations were status-based mitigating categories covering kinship, merit, office, and diplomatic standing. Judgment involved consultation across ministers, administrators, and populace. Three pardonable conditions were non-knowledge, mistake, and lapse of memory. Legal exemption categories included very young, very old, and intellectually limited persons. As for prisoners: 'Upper crimes cuffed and fettered; middle crimes cuffed and fettered; lower crimes cuffed; members of the king's own clan are only cuffed; those with rank only fettered, awaiting final outcome.' In Gaozu year seven, an edict to Censors said: 'In doubtful cases officials sometimes dare not decide; the guilty remain long untried and innocents remain long detained undecided. Local doubtful cases were escalated to higher prefectural review. Unresolved matters moved upward to central judicial review. What the Chief Justice cannot decide, carefully memorialize and attach the matching statutes and ordinances for hearing.' Despite clear directives, administrative implementation remained weak. Therefore in Jingdi middle year five another edict said: 'In all doubtful prisons, though textually fitting statute but not satisfying human feeling, immediately refer them for review.' Yet legal formalism reasserted itself through pedantic loopholing. In Houyuan year one another edict said: 'Prisons are weighty matters. Human and administrative capacities vary; Doubtful cases should be reviewed; when ordered review has already reported review but later proves improper, the reviewer is not counted at fault.' The system moved somewhat closer to classical evidentiary and mitigation principles. Three years later another edict said: 'The very old are those people honor and respect; Widowers and widows not subject to summons were also persons whom people pitied. Set this in statute: those over eighty, below eight, and pregnant women before weaning, as well as teachers and dwarf retainers who should be arrested, are to be held by summons rather than bound detention.' “By Emperor Xuan Yuankang year four, another edict said:
22
滿 便
He emphasized mercy for frail elders swept into penal machinery despite low social threat. The policy narrowed elder criminal liability to severe or malicious offenses. In Chengdi Hongjia year one, an order fixed: 'If not yet seven years old, and in robbery-fighting homicide or capital offenses, submit to Chief Justice for hearing and may receive commuted death.' The measure was framed as consistent with classical exemption doctrine. Such revisions are presented as successful examples of humane legal normalization.
23
滿
Confucius said: 'If there were a true king, it would take a generation before benevolence; A good man governing a state for a hundred years could overcome cruelty and remove killing.' Moral transformation is gradual and depends on long-term virtue-centered statecraft. Even second-best rulers, if persistent, can still greatly reduce penal violence over time. This sets the proper timeline expectation for political reform. Across the mid- and late Western Han reigns, capital sentences still averaged about one execution per thousand-odd persons yearly, and severe non-capital punishments ran at more than triple that rate. An old saying says: 'If a hall is drinking and one person in a corner weeps, then the whole hall is unhappy.' A ruler should treat each unjust case as a stain on the whole polity. Mass incarceration and high execution counts are cited as causes of unresolved social disharmony.
24
The text attributes penal overgrowth to moral-institutional failure, socioeconomic stress, elite predation, and uneven enforcement. The Documents says, 'Bo Yi descended with canons, wise people rely on punishment,' meaning rites are made to stop punishment, like dikes preventing flood overflow. With normative institutions eroded, legal pressure rises. Capital punishment is overused while minor liability is overly broad; Material deprivation pushes people into offense; elite shielding networks conceal crime, normalizing broader lawlessness and expanding punitive response. Confucius said: 'Those who understood law in antiquity could reduce punishments-this is root; Those who understand law now merely do not miss the guilty—this is branch.' He also said: 'Those who hear cases now seek how to kill; Those who heard cases in antiquity sought how to let live.' The preferred error is under-punishment over wrongful killing. Bureaucratic incentives reward harshness and punish moderation. A proverb says: 'The coffin seller desires epidemics each year.' The motive is profit structure, not explicit malice. The same perverse incentive logic applies to predatory legal officials. These five systemic pathologies drive penal overexpansion.
25
In early Eastern Han recovery, social conditions resembled early Han: pro-weak governance and reduced private coercive power. Per capita criminal adjudication fell dramatically versus late Western Han, indicating relative legal clarity. Even improved conditions fell short while structural causes and penal foundations remained uncorrected.
26
Well said! Xunzi's discussion of punishment says: 'Common talk says ancient good governance had no corporal penalties, only symbolic penalties like tattooing types and rough shoes/red clothes; this is incorrect. If crime truly vanished, all penalties-not just corporal ones-would be unnecessary. Simply lightening punishment without moral order would nullify deterrence for grave violence. Severe crime with weak sanction creates maximal disorder. The root purpose of punishment is to prohibit violence and evil, and to chastise their
27
downstream excesses. Excessive leniency toward violent offenders is itself injustice to social order. So-called symbolic punishment is not a pristine ancient ideal but a product of later disorder. Political order requires proportional correspondence among status, reward, and penalty. Any mismatch in these correspondences opens a path to institutional chaos. Nothing is more ruinous than systemic misalignment between merit, office, reward, and sanction. Coercive suppression of rebellion is an essential dimension of state authority. Across dynasties this principle remained constant and foundational. Xunzi paradoxically argues that robust order permits stricter sanction because normative standards are higher. This is the meaning of the classical phrase about historically variable penal weight. 'Symbolic punishment' should be read cosmologically, not as mere costume penalties.
28
穿忿 祿 祿
He reinterprets historical shifts as responses to declining social-moral conditions. Given post-Zhou/Qin social deterioration, applying ultra-lenient antiquarian models is seen as impractical and dangerous. Commutation policy can perversely increase death sentencing if replacement scales are misdesigned. That outcome betrays the reform's humanitarian purpose. High annual execution counts are attributed to excessive penal severity. Conversely, many offenses are under-deterred by too-light sanctions. Mass punishment without moral effect indicates miscalibration at the lenient end too. Administrative culture came to reward lethal zeal over lawful balance, corroding institutional norms. Legal expansion without legitimacy produces both more law and more defiance. Failure over long horizons is traced to absent ritual order and distorted penal structure. The text calls for foundational legal redrafting into a coherent, compressed code. It proposes commuting many currently capital crimes to calibrated corporal sanctions. For assault, theft, corrupt officials who pervert the law, and sexual misconduct, penalties should be reset to the older classical scale, yielding a code of three thousand provisions. All petty, manipulative, and overly technical legal provisions should be abolished entirely. If this is done, the law will be stern yet navigable, officials will no longer wield death arbitrarily, punishments will fit offenses, and lives will be preserved. It would restore balanced penal governance, deepen harmony between Heaven and human society, align with tested ancient institutions, and produce an age of social concord. We may not reach the ideal era of Cheng and Kang, when punishment was scarcely needed, but we can still aspire to the humane judicial standard associated with Emperor Wen. As the Odes says, 'When governance truly suits the people, Heaven grants its favor.' The Documents says: 'Establish achievement and establish institutions, and one can endure for long years.' In other words, governance that truly benefits the people creates lasting achievement and institutions. Such rule receives Heaven's favor and endures, fulfilling the saying: 'When one person is blessed, all the people depend on it.'
← Previous Chapter
Back to Chapters
Next Chapter →