Terrence E. Cook

Department of Political Science

Washington State University

The maintenance of law and order by the use of punishment is the science of government.

Kautilya, The Arthashastra, I, ii

India's classic realist understood the centrality of punishment in politics, especially for autocratic, imperial states. After all, the Mauryas in India were like the Qins in China in regulating much of economic life. I will primarily look at the L egalist ideas behind Qin policy regarding crime and punishment. In doing so, I will argue that criminal deterrence has two distinctive variants in social thought, the one arising in premodern or agrarian society, the other emerging in later modern, libera l society. Both want punishment to be swift, sure, and certain, focused on prevention of future crime. Most moderns, however, advocate milder punishments of merely optimal severity, worried about too much more than too little, while most ancients wanted d eterrent punishments to be very severe, fearing too little more than too much. As the best elaborated and actually practiced case of the latter, I will address Chinese Legalism, then return to other comparisons with the criminal deterrence theory found in modern, classically liberal social thought.

In premodern deterrence theory, punishments were often severe, at least as severe as favored by classic retribution theories, but with a view to deterrent effects rather than to just desert. Before turning to the Qin, I may briefly describe rel evant aspects of the ancient Assyrian empire: Assyria was a semitic state of northern Mesopotamia, with their later capital at Nineveh (near modern Mosul, Iraq). Severe corporal punishments were normal: King Shalmaneser I (d. 1245 B.C.) at least partially blinded 14,400 deportees to incapacitate future rebellion yet permit their economic exploitation. In later imperial expansion, the Assyrians not only plundered captured cities but displayed outside the walls severed hands of executed rebel fighters, ofte n forcibly relocating remaining populations. Leading rebels could be flayed until their skins separated from their bodies, which were then peeled off and left on public display. Otherwise, their most brutal punishment was impaling, apparently by the anus, with the bodies thus left on public display. But later empire required onerous taxes and tributes to sustain the system, especially with grandiose construction projects. In an 80-year expansion, the empire overextended its resources in occupying Babyloni a, which had long pushed its boundaries northward every time Assyria was weakened. Assyrian power had waxed and waned over the centuries, but the brutality of the last empire (which peaked with the murder of Sennacherib by his sons in 681 B.C.) partly exp lains its final defeat at the hands of an alliance of Medes and Babylonians in 612 B.C. (Munn-Rankin, 1967; Saggs, 1984; Kings II: 17-19).

While one could perhaps infer thinking of the Assyrians from their punitive practices, study of the Qin empire in China has advantages in that we have texts of the school of Chinese Legalists, which arose with the Qin (Ch'in) dynasty and receded wi th its collapse. The Warring States period (481-221 B.C.) had been a protracted dissolution of the semi-feudal Chou dynasty, ending when China's first Emperor, Qin Shihuangdi (Ch'in Shih-huang ti) unified China by conquest in 221 B.C. before dying in 210 B.C. During the warring, contending princes had fielded increasingly massive armies, which probably encouraged commanders to attempt standardization, while also displacing hereditary rights of nobles with commanders of battle-tested competence, even if of non-aristocratic social origins. The Legalist theory of punishment seems essentially the view of such military officers applying to the domestic domain their approach to organization (interchangeable units, standard operating procedures, no autonomous th inking in the ranks) and to preventing war (threaten credible retaliation with preponderant force). Everything is made uniform, and the first recourse for order is force. As one analyst notes, the Chinese term for penal law (Fa-chia) has dual conno tations of reference to one of several kinds of carpenter's norms (a plumb, a square, etc.) and to method (Scharfstein, 1995, 31-32). To the Legalists, law was but the promulgated will of the ruler, a will which could change at the ruler's discretion. To them, power (shi) meant potential of the position, and they had no separable concept of the legitimacy of the authority (Fu, 1996, 37-38).

While most would regard the later Han Fei Zi (c. 280-233 B.C.) as the foremost legalist theory of international relations, Legalist theory as focused on domestic security is best articulated in The Book of Lord Shang. While many think it a c ompilation by different hands at different times, the book's conceptual unity argues that it closely follows the written and oral traditions of one man, Wei Yang (sometimes also called Kung-sun Yang), who, upon eventually being given an estate by Duke Hsi ao, is renamed Shang Yang or Lord Shang. Wei Yang (d. 338 B.C.), a rough contemporary of Aristotle (d. 322 B.C.), was descended from a ruler of Wei, but only through a concubine, and his family had lost aristocratic privileges even if of the lower landlor d status, permitting his good education (Li, 1977, 20, 32 n.1). He had sought a consular position in his native northwest China kingdom of Wei, but he was spurned in favor of Confucians better trained in the arts of flattery and gift-giving. In anger he e migrated westward to the kingdom of Qin (Ch'in), where he became counselor to its ruler, Duke Hsiao, while also commanding military forces. Among other services, Lord Shang sets up a banquet with the far too trusting old friend, Prince Ang of Wei, who fin ds himself entrapped and his army soon destroyed. This permitted the Qin to begin their expansion into the eastern plains. But after turning down an offer of the royal succession, Lord Shang suffers a reversal with the illness and death of Duke Hsiao, for Lord Shang had once offended the crown Prince Hui, vicariously punishing him for a youthful violation of law by cutting off the nose of his guardian and tattooing the face of his tutor, as then done to slaves. Ascent of an unfriendly ruler forced Shang t o flee and to then attempt a failed armed rising, which led to his capture. He is then torn apart by chariots and all of his family members were also executed.

Before his death, Lord Shang had advanced a new teaching, possibly influenced by the law codifier Li Kui before him. In part Legalism was influenced by Daoism (Taoism), especially the Huang-Lao school, as when having the ruler seem to be in "inacti on" (wu wei) but really be in total control (see Fu, 1996, 21-27). In a far more consistent commitment, Legalism took dead aim against Confucianism. In crime control theory, Legalists denigrated the Confucian approach of inculcated morality, stress ing rather external punishment and fear as the basis for good order. From the start, Legalism's main goal was to build an internally strong state for military prowess amid warring states. Classically realist in all key assumptions, including a pessimistic outlook on human nature, Lord Shang held: "A sage...in organizing a country causes the people in home affairs, to adhere to agriculture, and in foreign affairs, to scheme for war" (Shang, 1928, 219). A regime of farmers and fighters was best suited for c onquest.

In this like the later James Harrington in the West, Lord Shang saw productive agriculture as linked to war, since public granary reserves are needed not only against harvest failures but also to supply armies marching off to war. Although his tax policies had the unexpected consequence of causing peasants to lose their alienable land to speculators and eventually turn against the Qin, Shang had the land taken from nobles' control under the well-field system and distributed directly to peasants, sa ying that they would better fight if attached to their homes. Instead of laboring on the noble's central plot in the old nine field system (the broad, dividing pathways were also privatized for cultivation now), the peasants were taxed only 1/15th of thei r harvest of grain in a centralized in-kind revenue collection. Even that tax was waived for any farmer-fighter who captured an enemy head when at war. But Shang also instituted a male adult capitation tax, which was doubled for any household where a fath er kept any second grown son at home, which pressured smaller farmers to divide their land, send extra sons to newly opened lands, or even sell sons into slavery to avoid going into debt (Li, 1977, esp. 61).

While this last policy was not onerous for larger landowners, and while middling landowners may have supported Legalism for opening new land-acquisition opportunities, in both Wei and Qin the aristocrats hated Lord Shang, since he not only reduced their power and revenues but openly denounced their luxury consumptions and failure to help prepare the state for war. Shang also denounced the parasitical nature of their idle sons, servants, and the artisans who made their luxury goods.

Lord Shang also alienated the merchants, who were to be kept as few as possible. While he encouraged the uniform weights and measures innovated by the Qin, this was aimed less at protecting the merchants themselves than at safeguarding everyone els e from being cheated by merchants, while also facilitating centralized administration. The Legalist Qin regime banned use of money, debarred the merchants from trading in grain, forbade the inns needed by traveling merchants (as well as itinerant thieves) , and doubly taxed the merchants at their market stalls and at toll gates along the roads. If any merchants nevertheless accumulated material wealth, they were forced to buy costly titles with it.

Lord Shang also attacked the bloated bureaucracy. He inveighs against the alleged parasitism of the Confucian scholar bureaucracy which neither grows food nor competently fights wars. He disdains their idleness, pretensions to knowledge, and gift-g iving and flattery in quest of promotion: "A country that loves talking is dismembered" (Shang, 1928, 188). He would eliminate these "professional talkers," favoring a lean, centralized bureaucracy of a military-meritocratic mode, which instead of words, keeps careful numerical data on all state resources relevant to warmaking: "Statistics is the true method of ministers and rulers and the essential of a state" (Shang, 1928, 218). Swift implementation of all centrally determined policies leaves no time fo r bureaucratic corruption to arise (Shang, 1928, 175). One rises in the state hierarchy solely by proven prowess in war, and the state gives out few direct rewards at all except for rewards for informing of criminal activity and for battlefield merit, eve n down to the lowliest soldier as previously noted regarding tax exemption. Informing on an important criminal counted equally with capturing one head in war.

Thinking that people are virtually rewarded when not being punished, Lord Shang suggests that the ratio of rewards given to punishments administered should be about one to nine. Sparse state rewards, he argues, will heighten the fearsomeness of sta te administered punishments. While the Legalists for some crimes may have favored forced labor beyond the normal corvee, perhaps in bringing wilderness under tillage, there is no record of fines. The usual recommendation of Lord Shang and his successors i s capital punishment. Lord Shang does not elaborate about modes of execution. But the usual practice of simple beheading was kind when compared to the more gruesome Chinese forms of that time, which included ripping out ribs, boring holes in skulls, boili ng in cauldrons of oil, or pulling head and limbs off by chariots driven in opposed directions, as in Shang's own execution.

Severity of punishment is reflected in collective responsibility, even within the military: If one man in a five-man squad is killed in battle, the other four who presumably could have prevented it are executed, too. If a state official or even ord inary soldier disobeys commands, the whole family of that person is executed. Even non-Legalist rulers had such collective family responsibility up to "three levels," including the perpetrator's generation (including spouse), his father's generation, and his children. Obviously enough, this was meant to promote deterrence by strong family vigilance against criminal action by any member (even if in a blood feud between families) while also minimizing the possibility of a disgruntled relative seeking vengea nce against the state. But it also shows Lord Shang's disinterest in individual justice as we have to come to know it. Perhaps the last vestige of the Chinese family responsibility model was the occasionally reported Maoist practice of making the family r epay the state for the bullet executing one of their family members. Collective family accountability seems to be in tension with the Qin campaign (begun under Duke Hsiao) to penalize redundant men within peasant households. This would have weakened the t ypically Qin father-son households of the uplands as well as the extended grandfather-father-son households more typical on the conquered plains. However, according to Han Fei, Lord Shang also innovated an auxiliary model by using "five and ten households " (at least the five grouping had been administratively used by the Qin earlier) of often unrelated peasants not only to prevent migration but to pay collectively for a crime done by any member. As Zhengyuan Fu hints, this parallels the mentioned military practice of collective responsibility in the five-man squad (Fu, 1996, 41-42).

In addition to such collective responsibility patterns, even small offenses are capitally punished, so there can be no retributivist or just deserts notion of just proportionality of crime and punishment: "If light offenses are heavily punished, pu nishments will disappear..." (Shang, 1928, 212). Since almost all punishment is capital, there is also no place for rehabilitation, or reform within. Nor could there be any specific deterrence, which is dissuasion of a particular malefactor from repeating crime by fear of another punishment. Other than total incapacitation, virtually the sole rationale for punishment must be general deterrence, or use of the example to dissuade others than the malefactor from committing crime. Lord Shang assumes what psyc hologists call stimulus generalization, such that capital punishment for even normally more numerous light offenses will more strongly dissuade people from more serious offenses. Note also his assumption that with initially severe punishments, the need fo r punishment would wane, a view like that of Machiavelli's Prince.

As in other deterrence theorists, Lord Shang in administering punishments looks more toward the future than to the past. Preventing future crime supposed utter clarity in codified, published laws, which the Qin inscribed on bamboo as well as publi c wooden pillars. Again like Machiavelli much later, Lord Shang held that any past merits should have no relevance to the punishment presently administered. Lord Shang also emphasizes deterrence aims in punishing those who would conceal crime, especially by the collective responsibility forms addressed above.

As already apparent by refusal to allow for past merits, Lord Shang wants perfect uniformity in administration, allowing no preferences by administrative district, social class or rank, or by kinship. As noted, Lord Shang apparently counseled Duke Hsiao to undermine extended family households. He expressly opposed Confucian training in "virtue" which had encouraged deference for elders and a loyalty to the family unit. Strong family bonds interfered with effective centralization of power for milita ry conquest. Relatives would be barred from bringing food to public prisoners or petitioning officials for their release, since Legalists disliked the granting of pardons or clemency.

Lord Shang believed that the Confucian approach to crime control muddied the waters with alternative standards (old texts, family loyalties, etc.) other than the published will of the ruler, which is identical to law. The Confucian approach require d too many parasitic scholars, who practice flattery when in office and fan seditious notions if out of office: "The capital of scholars, who are full of empty words, lies in their mouths; that of scholars, who are out of office, lies in their ideas..." ( Shang, 1928, 220). Besides, Confucians wrongly held that one had to remain fixed in the old ways, whereas a good legal code should adapt to changing conditions. But Legalists maintained that the effort to inculcate virtue is simply ineffective in steering behavior, while use of many and severe punishments and a few rewards really works.

The following passage from The Book of Lord Shang nicely captures the whole spirit of his theory of punishment:

In the next century, Legalism is led by Han Fei Zi (Tzu) (d. 233 B.C.) and Li Xi (Ssu) (d. 208 B.C.). The harshly predictable Legalist Qin rule eventually facilitates unification of China by 221 B.C. under its first Emperor, Qin Shihuangdi (d. 210 B.C.). Typically Legalist in his efforts to weaken hereditary nobles and centralize rule for military strength, Shihuangdi through his chief general linked shorter sections of prior walls into the Great Wall of China, and he conducted a census and complet ed standardizations, including a uniform script of Chinese and fixed axle width of carts so that they could all use the same roads. Under the guidance of his chief counselor, Li Xi, the Emperor ordered the burning of Confucian books in 213-12 B.C. and the burying alive of 460 Confucian scholars. Li Xi was blunt: "Those who use the past to criticize the present should be put to death together with their relatives" (Bodde, 1938, 83). Qin rule brought rivers of blood in warring. The serried ranks of the vast terra cotta army uncovered near the Emperor's tomb give us a visual sense of this militarist spirit of Legalism. All stand firmly prepared for war against any new invader from the western borders. While the army was mass produced with assembly line proce dures, the soldiers are yet depicted with very individual faces. Even the artists are so depicted, with pained expressions as if anticipating their eventual execution to prevent revelation of the location of the tomb.

Application of Qin Legalism involved much mutual accusation of crime and vast numbers of executions. Conceding this, Han Fei asserted that it became effective once all criminals came to expect death: "At least the people came to know that men guilt y of crimes would infallibly be censured and informers against culprits became many. Hence the people dared not violate the law and penalty could be inflicted on nobody. Therefore, the state became orderly, the army strong, the territory extensive, and th e sovereign honorable. The cause of all these was nothing other than heavy punishment for sheltering criminals and big rewards for denouncing culprits." Amid a later discussion which anticipates Machiavelli's recommendation not to seek the love of subjects through charity and mercy, he adds: "...severe penalty is what the people fear, heavy punishment is what the people hate. Accordingly, the wise man promulgates what they fear in order to forbid the practice of wickedness and establish what they h ate in order to prevent villainous acts. For this reason the state is safe and no outrage happens" (Han Fei Tzu, 1939, 123, 128).

Sima Qian, the more critical Confucian historian of the Han dynasty, has Fan Kuai remark: "The king of Qin had the heart of a tiger and a wolf. He killed men as though he thought he could never finish, he punished men as though he were afraid he wo uld never get around to them all, and the whole world revolted against him" (Qian, 1961, Vol. 1, 31). Beyond excesses in executions were excesses in warrings with the Xiongnu to the north and Yue to the south, and the onerous labors on the Great Wall: "Th e young men were obliged to don armor and the girls to transport provisions; hardship made life no longer worth living and the people hanged themselves from the roadside trees in such number that one corpse dangled within sight of another." Sima adds that the dynasty "was destroyed because its laws were harsh and its ambitions grandiose and endless." (Sima, 1961, Vol. 2, 201-2) The fall was precipitous. The peasant rebel Chen She was like the precipitating pebble which started the "landslide" fall of the Qin empire: "He rose up from the lanes and alleys, grasped his spear ... and raised a mighty cry, and the whole empire followed after him as though driven by the wind. Why did it happen thus? Because the people were in misery and their rulers had no mercy , because the governed were full of hatred and the governors lost in ignorance, and because the customs of the nation had been thrown into complete confusion and the rule of the dynasty had ceased to be effective. These three conditions were the material out of which Chen She fashioned success" (Sima, 1961, Vol. 2, 197).

Ironically, overuse of capital punishment was a key condition of this rebellion against the third and last Qin Emperor: Some 900 peasants ordered to a military rendezvous, unable to reach the place on time due to heavy rains and quagmire roads, kne w that they would be beheaded. Having nothing to lose, and encouraged by some manipulated, auspicious omens, they followed Chen She in rebellion. Soon more orderly forces attached to the old kingdoms also rebelled, eventually permitting the Han monarch to begin a new imperial dynasty, which, like many later ones, tended to veil some Legalist influences behind its Confucian orthodoxy.

Far more ephemeral than the brutal Assyrian empire, the Qin dynasty left China with the memory of unity but also of fear. Although many Chinese rulers liked some Legalist admixture, Confucianism was revived as the favored outward social philosophy of China. Notwithstanding Max Weber's emphasis on incompatibilities, eventually Confucianism adapted to capitalism, while Mao's brand of Communism was openly sympathetic to Legalism, especially for unifying China and maintaining order by force. But if als o reunifying a divided mainland China, Mao's militant Communism would itself take millions of Chinese lives, being fully dominant on the mainland only 1949-1976. The Qin Legalist orthodoxy dominated imperially over China about 15 years, while the Maoist o rthodoxy prevailed about 27 years. Put these in perspective: the Chinese measure some other imperial dynasties by centuries.

Now I will sketch some similarities and differences of the ancient Chinese Legalists and the modern liberal theorists of deterrence. While I have named some key Legalists above, in the latter I am thinking of such figures as Benedict Spinoza (d . 1677), Cesare Beccaria (d. 1794), and Jeremy Bentham (d. 1832).

First, what are some similarities? Both groups distrusted the nobility, opposed preferential treatment in law for nobles, and opposed massive expenditure on luxuries (including palace ornamentation, ceremonies, etc.) reflecting noble tastes. Both distrusted men of words too close to the nobles, whether the Confucian scholar gentry or the Roman Catholic prelates. Both wanted to displace hereditary inequalities with those ordered by merit, by measures of competence concerning effective perform ance of important skills (if Han Fei thought counting heads taken in war was too narrow) rather than the alleged sterility of traditionalist bookish learning. Further, both groups looked for more efficient and effective governance, and they shared an inte rest in standardizations of weights and measures as well as the gathering of extensive census data. Both distrusted extended family loyalties. As noted, the Legalists favored nuclear family households, which was also to be favored by liberals, when not fa voring individuals even more.

Both Legalists and liberals presupposed self-interestedness behind effective deterrence: Neither those who hate themselves nor those too much in love with others could be deterred by the prospect of punishment administered on themselves. Both assum e a calculating, instrumental rationality in such individuals subject to law. Behavior is hence best guided by clear, published laws which are predictably enforced. Spinoza held that even denial of free will would not make punishment irrational, since det errence aims remain: "Wicked men are not less to be feared, and are not less harmful, when they are wicked from necessity" (Spinoza, 1951, 392). While both Legalists and liberals recognize rewards as well as punishments in the guidance of behavior, both a ssume that state use of punishments would far outweigh any use of rewards. For Lord Shang or Bentham, the state rewards almost enough by leaving most law-abiding people at their productive work, not punishing them or unduly taxing them.

Both groups de-emphasize retributivism because it looks backward rather than forward, and if liberals are not consistent, both doubt rehabilitation as well. Any past merits would not extenuate punishment, just as prospective future merits are to be discounted. Both reject class bias as well as any other particularistic traits in administration of the law.

However, both deterrence theories can fail whenever criminal action is not a self-interested, carefully calculated quest for advantage, as when a juvenile engages in mindless vandalism or an adult murders in a blind rage. These theories are focused on preventing an aggressive gain, a shift to a more advantageous reference point, and deterrence works poorly when the reference is rather to avoid a loss from the status quo, as illustrated in the Chen She's rebellion against the Qi n. As noted by many (from Jeremy Bentham to Amos Tversky), anticipated loss brings more pain than an expected equivalent gain brings pleasure. Hence actions to avoid loss accept higher risk than actions to attain an otherwise equivalent gain. Yet not ever y deprivation is taken as loss: For those expecting and even wanting death (as when estranged spouses or fired workers intend to kill themselves if the police do not do it), deterrence also fails.

Turning to differences, Legalism is inspired by a military model, whereas the thinking of Spinoza, Beccaria, and Bentham is, I believe, inspired by mercantile (i.e., of merchants and their marketplaces) experience. As noted , Lord Shang despised merchants and wanted them maximally regulated to prevent their cheating of soldiers, peasants, and even state officials. If Legalists opposed noble and palace luxury, it was because of a wish to devote funds to military rather than d omestic productive purposes. Indeed, the Great Wall of China was to require severe taxation and corvee labor, and it contributed to withdrawal of peasant acquiescence in Qin rule. On other matters, the Legalists sometimes favored traditionalist moderation of desire, as articulated by the semi-Taoist Legalist Han Fei as one key to good rule. But no Legalist could accept the complicated rules of propriety by which the Confucians would guide desire. As if anticipating complaints against later Pharisees and P uritans (whom Spinoza thought similar), Han Fei puts it bluntly: "Whoever observes complicated rules of propriety is rotten in his innermost heart" (Han Fei Tzu, 1939, 174).

The classical liberals thought of military officers and not merchants as parasitic on society, seeing waste in most military expenditure beyond that necessary to protect borders and high seas shipping. In the emporia of Amsterdam, Milan and London, Spinoza, Beccaria and Bentham all had personal experience with trading houses, and Beccaria even taught money and banking. Classical liberals, I think, rethought deterrence on the model of investment behavior, both of the criminal who must not expect to "profit" (Bentham's term) from the crime, and of the state, which invests in the crime control apparatus to minimize its own pains while not neglecting those of the punished. Bentham, whose pleasures and pains look like credits and debits in double-entry bookkeeping, conceded that his felicific calculus would be impossible without conversion into money terms.

Legalism was decidedly unliberal, placing no value on civil associations and pluralism or on freedoms of movement, assembly, speech, and the like. Ordinary people were not to read books, even Legalist ones. Hence one commentator with experience of Mao's prisons accuses Legalists of pioneering "totalitarianism" (Fu, 1996). Civil libertarian commitments of liberal theorists make deterrence of crime a somewhat different and more difficult problem, especially given an economic order which encourages hi gher consumption strivings as the way to stimulate risk and effort in more production.

Legalism is inconsistent in both wanting a very predictable rule of law and in exempting the ruler from its sway. Han Fei holds that the wise ruler will firmly require that ministers keep to their assigned duties, give only accurate information to him, and not conspire with each other (a censor should be attached to each). While counselors are under close surveillance, the ruler himself is to be mysterious and even unpredictable in his administration of rewards and punishments, especially when curb ing anyone else in his entourage from gaining too much power. Ironically, Shihuangdi's entrenched set of counselors (upset by Han Fei's incautious denigrations of them) follow Li Xi (who could have seen a rival for his job) in plotting to put Han Fei unde r suspicion of espionage for the Han state. The Emperor orders Han Fei to drink poison. Slandered in turn by a cunning palace eunuch before the second Qin Emperor, Li Xi himself undergoes the ancient "five punishments," culminating with being cut in half in a public marketplace.

The Chinese could conceive of a dynasty as having or losing a Mandate of Heaven, an emphasis of the non-Legalist thinker Mo Di (Mo Tzu). But unlike all other ancient civilizations, Chinese traditions never viewed this worldly law itself as divine. Nor did they regard humanly made law as limited by some higher law (Fu, 1996, 57). Classical liberals such as Locke (who would run for his life) or Sydney (who eventually lost his life in a horrid execution) held that tyranny is more dangerous than anarchy, so insisted on subjecting the national chief executive to the rule of law. While sharing with the Legalists the idea of law as will, it is the will less of a monarch than of parliamentary representative body, but even that will for liberals had normative limits. While often more individualistic than those adhering to the Roman Catholic tradition of natural law, classical liberals at least held that human law must respect natural rights (e.g., Locke, Sydney, Paine) or the equivalent in social uti lity (Hume, Beccaria, Bentham). Liberals distrusted executive discretion, allowed wide scope only in emergency powers and in occasionally moderating the severity of punishment, as in pardoning or clemency judgments rejected by Legalists.

Most obviously, Legalists and liberals differ in that the former emphasized very severe, normally capital punishment, often with collective accountability of military squad, extended family, or group of neighbors. While the Legalists valued not onl y military exertion but agrarian labor, they apparently saw even peasants as abundant and replaceable. But from the 17th century onward, liberal theorists saw great productive possibilities in labor guided by better science and less regulated economic arr angements. This is one reason why they not only reject collective punishment but eventually turn against severity of individual punishment, especially for small crimes, replacing much capital punishment with incarceration or fines. William Petty (d. 1687) , viewed by both Marx and Keynes as the founder of political economy, complained that excessive hangings or even idle imprisonment wasted human labor potential, which could often be saved or even spurred by substituting "pecuniary mulcts" (Capital, III, xlvii, 1; Keynes, 1971, 7).

As Foucault notes, the later modern approach to crime control turns away from assaults on bodies to imprisonments, fines, and psychological manipulation (Foucault, 1979). Minimizing executions forces attention to specific deterrence as well as poss ible rehabilitation of offenders, both wholly irrelevant to Legalist rule. In thinking deterrence, the moderns keep to the idea of calculating self-interestedness, but shift their emphasis to increasing the probability of getting caught and punished, whic h, when multiplied by the milder punishment, may even weigh more in a would-be malefactor's calculation. In any case, excessively severe punishments in the modern world have often led judges and jurors to refuse to condemn the guilty. Summary. Influenced by the military model, Qin Legalist criminal deterrence understood autocratic command, regimentation, and massive coercion. By executing almost all malefactors and those close to them, Legalism was costly in human lives, with doubtful success i n crime control. Even if Legalism helped effective warring without, it did not prevent insurrection within.

Influenced by markets, liberal deterrence offered a first recourse of consultation, freely chosen leadership and uncoerced consent. Liberal regimes avoid major warring with each other, and at least the economically more advanced states face no succ essful domestic rebellions. But they attempt to deter any individual malefactors less by severity of punishment than probability of being caught and punished. After two centuries of its practice, liberal deterrence becomes costly in tax bills for bolstere d law enforcement and extensive incarceration. Also, by scanting justice in the city and in our souls, we only manage a crime problem we do not really solve.

Kautilya (Vishnugupta Chanakya). 1987. The Arthashastra. L. N. Rangaragjan, tr. New Delhi: Penguin Books. The epigraph is from p. 108.