Dominated by Anglophone scholars interested in their own nations and engaged by an important political issue, criminology is focused largely on the present.

Much of the recent historiography reflects the American situation—its crime wave of the 1960s and 1970s and the attendant incarceration boom. It is preoccupied by the turn to retributive justice in the late twentieth century, with its adoption of long and harsh jail sentences and abandonment of earlier reha­bilitative ideals.¹ Only recently has the literature begun to address the decline of urban crime since the 1980s as many big cities have become safer.²

We gain perspective by stepping back from the ripples of imme­diate events to consider long-term currents.

If we instead examine how the state has grappled with crime over the longue duree of three millennia, recent events appear in a broad historical context, revealing some unexpected aspects. Two general observations on the deep history of crime merit mention. History deals with both ruptures and continuities. Taken over three thousand years, as here, both will inevitably figure. Almost every era has debated whether punishment's justification is retributive (delivering what offenders deserve) or utilitarian (aiming to dimin­ish crime).

Equally notable from a long perspective is how development pro­ceeds by accretion, not just by substitution. Layers of the old remain, a palimpsest of policies. The old is present though partly shunted aside, never fully obscured or covered by the new. Much of the old privatized approach to crime remains even now when the state pretends to have assumed this task wholly. Pardons remind us of our feudal past, when the king kindly dispensed justice regardless of what the law actually said. Churches remain refuges, an excep­tion by courtesy to the state's allegedly all-encompassing territorial domination. Parole is a modernized form of surety as the commu­nity to which the inmate is released fulfills a similar role to com­purgators. Bail is the obverse of the frankpledge, a guarantee given before rather than after the crime.⁷ Civil asset forfeitures continue the medieval idea of deodands, the state confiscating the tools of crime. The persistent acceptance of the concept of justifiable homi­cide suggests that the state cannot be bothered to—or perhaps just cannot—impose its will on all our actions, even extreme ones. So do the remnants of vigilante justice the state still tolerates, as does the private assistance it accepts (and indeed relies on) from civil society, whether in the form of vast nonofficial police forces or the cooperation of citizens in investigating crime. Also to be included here is the barely polite fiction that the state has imposed a thor­oughgoing monopoly on violence and its means. In the United States, Second Amendment discussions continually raise the primal heresy—sedition, really—of civil society's possible armed revolt against its own state, not just against outside enemies.

That brings us to more specific conclusions about the state's role in dealing with crime over the past several millennia. The first is how late the idea of crime as an offense against society emerged as distinct from the more immediate sense of wrongs committed between private parties and resolved among them independently. That broader idea of crime, in turn, was a function of how late the state came to what now seems one of its core competencies— enforcing law and punishing its violation. As treason's immediate victim, the state did, of course, punish it from the outset. But for centuries other offenses were left to be sorted by the implicated par­ties themselves. The ancient Greeks and Romans established rudi­mentary judicial and policing systems, taking some matters out of private hands. But not until the early modern era did the state once again penalize crime. That holds even for the world's oldest contin­uous statutory authority, in China. Across the world, most judicial and police work was outsourced to civil society, with kin groups and other intermediary organizations accountable for their members' transgressions. Today we think of law enforcement as one of the state's primary functions. But policing in the modern sense is in fact a very recent activity—arising long after defense, taxation, eco­nomic regulation, social services, and often even public education.⁸

Starting in the seventeenth century, as the European state became directly involved in enforcing law and penalizing crime, it moved to consolidate its power. Monopolizing violence, confiscat­ing and regulating weaponry, subordinating private military troops to its own armies, running one universally recognized judicial sys­tem, the state took in hand the business of formulating, promulgat­ing, and enforcing law.

From a long historical perspective, the state's need to demon­strate its power diminished as its actual strength grew. As the state multiplied its capabilities, it could better detect, deter, and pun­ish transgression. It no longer had to strong-arm and terrorize its subjects. The more powerful and self-confident a community, the more moderate its penal law, Friedrich Nietzsche pointed out.⁹ The Enlightenment philosophes were right: predictability deterred more than ferocity. Torture was the weak state's route to evidence. Better surveillance and detective work made torture redundant. Recali­brating the correspondence between offense and punishment also changed the nature of the evidence required. In the early Middle Ages, confession or two eyewitnesses were needed for capital crimes. Later, when lesser but factual evidence was admitted, in return pun­ishments were moderated.¹⁰

With the admittedly gaping exception of Europe's twentieth­century totalitarian regimes and arguably today in China, the state's power and its harshness have been inversely correlated. A stronger and more pervasive state could police more effectively and there­fore benignly. With data flowing evermore freely, the reciprocal transparency between authority and citizenry reduced the need for drastic interventions and in any case limited the extent to which they were tolerated. That historical trend continues today. DNA analysis has retrospectively spared innocents convicted on faulty testimony, electronic automobile tracking has dramatically reduced thefts, and closed-circuit TV evidence has raised detection and con­viction rates.

This inverse correlation between the state's power and its harsh­ness leaves a conceptual conundrum. Is the state becoming nicer, less intrusive, less draconian? Or is it becoming more pervasive, pos­sibly less visible, but simultaneously more widespread and embed­ded in its subjects' lives? In the early nineteenth century, prison reformers heralded solitary confinement as a sea change in puni­tive techniques. Hoping to resocialize offenders' souls, its propo­nents knew they were inflicting a different but no less drastic form of pain: “no longer mere animal pain, but a pain that affects the whole spirit” or a “slow and daily tampering with the mysteries of the brain” that was “immeasurably worse than any torture of the body.”¹¹ Was a lesser force being exerted? Or did solitary confine­ment herald something different but equally imposing? The new, apparently more moderate state might simply be more subtle, not less strong.

Legions of social control theorists have sought to rip the mask from the face of modern state power, revealing how the authorities— though more surreptitious—dominate citizens' lives more than ever. Herbert Marcuse's seeming paradox of “repressive tolerance” emblematized this approach. Even the exercise of democratic rights, he argued, merely legitimated suppression.¹² That explanation solved the dilemma by definition, not by empirical analysis.

Michel Foucault famously argued that as the modern state evolved from absolutism, it aimed to punish better, not less. That was a fence straddle. On the one hand, he spurned the whiggish view that growing enlightenment made the democratic state friend­lier, exerting less immediate sway over its subjects. On the other, prison certainly seems preferable to drawing and quartering. His solution was to reframe the concept of power, moving from the absolutist authorities' centralized somatic brutality to modernity's “discipline,” a form of coercion that was exerted from multiple sources, that was often self-imposed, and that held sway over both elites and subordinates. In this reframing, he was following the torch brandished by Norbert Elias in his concept of the civilizing process. Foucault and Elias share more in common than often real­ized. Elias focused on individuals learning to control themselves, the demand side. This they were spurred to accomplish by a mutual dependence arising from society's growing complexity and inter­flection as well as by states that encouraged such reciprocal self­limitation, as among the French nobles corralled at Versailles before the revolution.

Foucault focused more on the supply side, looking at the all- pervasive disciplining imposed on subjects in countless ways through networks of power.¹³ Yet Foucault's only incipiently formulated con­cept of governmentality followed Elias in looking at civil society's self-disciplining and the mutually reinforcing effects of internal and external constraints. Late in his life, he turned away from his earlier preoccupation with state power to examine how individuals took care of and cultivated themselves, a task more akin to Elias's con­cerns.¹⁴ His interest became “to show how the government of self is integrated with the government of others.”¹⁵ The concept of govern- mentality was also a move beyond his early concern with social dis­cipline. Though dispersed, power was still at the heart of discipline, exerted over citizens to mold their behavior to align with prevailing norms. Elias focused on individual self-discipline, and Foucault on the disciplining exerted through institutions, but both were con­cerned with the shaping of the individual psyche and its habits.

Governmentality was the strategy of managing risk by using nondisciplinary techniques. And that, in turn, was but one aspect of a broader governance approach to the state that saw it less as imposing top-down power and more as coordinating multiple ways of regulating independent social actors.¹⁶ The averaging of risk through insurance, for example, has long been put to use by indi­viduals, but in the past two centuries it has come to pervade society as one of the main technologies imposed by the modern welfare state. Businesspeople have shared risk almost as long as humans have traded.¹⁷ The broader state strategy of averaging the risk of social problems threw up issues of free riding and underinsurance. Those, in turn, were dealt with once the state marshaled its powers of enforcement, requiring evermore citizens to be insured against risks that could be quantified, averaged, predicted, and priced. In the late nineteenth century, the German social reformer Lujo Bren- tano advised workers to maintain at least six different insurance policies: life insurance for their children, pension insurance for old age, burial insurance to pay for their funerals, and coverage for dis­ability, illness, and unemployment.¹⁸ Not all of these risks were pri­vately insurable, however, and over time the welfare state socialized the cost of such eventualities.

The modern state has also effectively mandated public health, thus socializing individually beneficial behaviors. It prophylactically vaccinates infants, screens schoolchildren and army recruits, for­bids noxious habits, quarantines the infected, and determines what happens to our bodies after death. In its benign form, such risk management became the liberal welfare state, but similar impulses prompted other states to assess, categorize, manage, improve, mold, and ultimately to deport or kill unwanted subjects. These impulses, some scholars have argued, lay at the heart of Stalin's supposedly improving ambitions and even—formulated in racial terms—of Nazism.¹⁹ Those terror regimes take us beyond our con­cern with crime and its punishment. In nontotalitarian countries, the shift from social disciplining to governmentality was not just from meddling with the individual psyche to mustering citizens. Managing risks meant combining individual self-discipline and social engineering. Citizens were increasingly expected to antici­pate what needed to be done—adopting healthy habits, curbing excesses, training for the work to be done, insuring themselves. Only in the absence of or in addition to such individual prudential- ist approaches did the state then step in.²⁰

Both Elias and Foucault realized that a binary approach to power led astray. Both rejected the nineteenth century's philosophical dead end of viewing the state as unilaterally all-powerful. Hegel defined the state as a transcendent force above civil society. Marx thought he was exposing the state's faux neutrality by unmask­ing its role as the tool of the dominant classes. But both agreed in regarding the state as being the dominant force in the relationship between the two. A century later Elias and Foucault parted with this tradition, insisting instead on the interaction between state and civil society and on how the state's role was shaped by the tasks pre­sented to it by civil society. The modern democratic state could not impose itself unbidden on its citizens. Without at least some coop­eration of the masses, for whom all modern states claimed to rule, it would fail. Even Hitler was in no position to force his subjects to obey when they refused outright.²¹ Stalin, governing a much less industrialized and urbanized—thus less self-controlled—population than the Germans, used more overt violence against so-called anti­social elements (the unemployed, petty criminals, drifters, and the like) than did the Nazis.²²

Democracy required the state to work in tandem with its sub­jects. They in turn were expected to meet authority halfway, disci­plining themselves and thus requiring less attention from on high. The state's authority was no longer unilateral. It interacted with civil society. That citizens imposed limitations on themselves did not necessarily mean that these were self-willed or self-chosen— anymore than are the details of our childhood socialization.²³ Yet, by accomplishing much of what it would otherwise have had to do, self-discipline did reduce the state's remit. The psychic, emotional, and instinctual limitations citizens imposed on themselves were rewarded with physical freedom from statutory attention.

Both Foucauldian governmentality and Elias's civilizing process thus displaced attention from the unilateral imposition of state power to the interaction of authority and civil society.²⁴ Humans are hypersocial creatures, our communal existence wholly depen­dent on the culture that we learn from each other, passed down from our ancestors.²⁵ But we are not massified by instinct, like ants or bees, blindly slotting into our appointed social roles. Culture, not biology, socializes us. How then to ensure human behavior that is compatible with organized society?

Social contract theorists have imagined humans discussing the terms they would unite under, trading nature's autonomy—whether the bliss of noble savages or a war of all against all—for protection in organized society. But that reversed cart and horse. These alleg­edly presocial negotiators, adeptly parsing the fine points of con­tract law, had somehow already achieved what they were supposed to be accomplishing. Historically, no such ur-constitutional conven­tion of Robinson Crusoes ever assembled. Humans first organized themselves in small bands. Some of these bands eventually joined

together in larger associations with incipient social stratification and some semblance of durable authority. Finally, some five millen­nia ago they formed the first recognizable states.

Before states emerged, and in the societies that remain without them, small and homogenous groups have governed themselves without much formal rule making or vesting of authority beyond their clan head.²⁶ Kin alliances forged largely through marriage helped create broader unity among disparate members as a matter of (sometimes made-up) descent. Polite fictions allowed new fam­ily members to be incorporated—adoption among the Romans, for example—alongside other means of artificially extending con­sanguinity.²⁷ Existing and historically documented tribes typically used such devices.²⁸ Chiefdoms, in turn, represented a halfway step between tribes and formalized state structures. They were still based on descent from a common ancestor but now had some social strat­ification. Hereditary leaders ruled, but there was still no formal leg­islation or enforcement.²⁹

How tribal societies kept the peace by punishing aggressive or aso­cial members with the cold shoulder or worse is a leitmotiv of the anthropological literature.³⁰ With the hyperdense sociability of the igloo as their norm, Inuits regarded the Western anthropologists who visited them as emotionally incontinent in their petty outbursts and everyday irritable flare-ups. Having violated precept, offend­ers were ostracized into the cold. Shunning and exclusion were the most widespread sanctions, formalized in larger-scale societies as excommunication. Habitual bullies or other intolerables who violated the egalitarian premises of hunter-gatherer societies were killed by delegated executioners, the victim's immediate kin stand­ing aside.³¹ Extrapolating backward from such anthropological find­ings, historians agree that prehistoric societies of hunter-gatherers and early agricultural settlements were likely similar.³²

Yet such theories of purely informal regulation have also been questioned. Anthropologists have disputed the relative roles of custom and law to guide behavior in what used to be called primitive soci­ety. Bronislaw Malinowski rebelled against the anthropology of his day, which argued that group custom, obeyed unthinkingly, gov­erned tribes and clans, with the individual but a cog in a kinship machine. He claimed to discern both criminal and civil law in such societies, with autonomous individuals tied into to a web of mutual obligation, responding to strictures prescribed for neglecting their duties. Much hinged on definition.³³ Formal mechanisms of adju­dication, “codes, courts, and constables,” indicating the presence of law, were admittedly absent, but that scarcely meant no regulation. Quite the contrary. Early humans were likely governed by a dense network of stricture, with law being, as Malinowski put it, but one form of custom.³⁴

States, in turn, first arose some five millennia ago because they were better able to deal with challenges that stumped smaller, less-organized societies. Their organizational prowess delivered a competitive advantage, allowing them to prosper, grow, and domi­nate. Early states were adept at marshaling resources. Armies were assembled, economies organized, towering infrastructural projects stamped out of the ground—all by institutions with only basic tech­nology. Homo sapiens lived for three hundred thousand years in rudimentary circumstances. Having invented the state, our ances­tors took but a few thousand years to build pyramids, aqueducts, palaces, and roads, start to write, and worship universalist gods. The state is arguably the most important invention in human history. It is the ultimate organizing tool of our ultrasociable species, whose main competitive advantage has been its ability to muster itself col­lectively and to transmit accumulated knowledge over generations.

Historians of Europe are perhaps at a disadvantage in appreci­ating how ancient states are. After the fall of the western Roman Empire, nearly a millennium passed before European institutions again approximated what had been achieved under the Egyptians and Greeks, not to mention the Romans. Europe's early modern history is largely the story of rebuilding the state. We tend there­fore to think of the state as a recent development. Add Hegel's insis­tence that the state of his era was its culmination and Max Weber's precise, exclusive definitions that made the difference between the modern state and all earlier states one of kind, not just degree. The result has been a historically myopic view of the state as a recent and unprecedented invention.

But states have long been with us. Unlike small, informally gov­erned societies, they imposed the controls required to coordinate their many subjects from the top and from outside. Thanks to pains­taking historical work, we now know, for example, how insufficient Weber's idea is that real bureaucracy developed only in the nine­teenth century. It may be that the bureaucracy of the Third Dynasty of Ur, two millennia before Christ, answered to individual families and was thus not rational and impersonal in a Weberian sense.³⁵ But already the Zhou dynasty of China, in power for some eight hundred years starting a millennium before Christ, had developed a bureaucratic apparatus independent of the ruler's person, with meritocratic recruitment and advancement as well as specialized functions.³⁶

Such efforts were both cumbersome and costly. States with suf­ficient resources could police and tame their civil societies. Indepen­dent sources of revenue helped, such as state-owned mines, farms, or eventually factories. So did slaves or other forced labor. Early states were mechanisms to tame subject populations and extract resources. Karl Wittfogel's theory of hydraulic despotism argued that ancient Egypt and other early states organized agricultural water by harness­ing subjects for public irrigation works. He has been criticized for extending this theory too broadly, but it has the virtue of highlight­ing how states mustered resources and subjugated populations to solve collective problems. James Scott has also recently argued that early states were predatory, centralizing power to accumulate wealth stored as grain.³⁷

We can thus imagine at one extreme autocratic external control over everything and at the other voluntary conformity produced by autonomous self-regulation. At times, states have been overbearing— not just among the Egyptians, Aztecs, or Incas but also in early modern Europe's Polizei regulations, addressing every conceivable behavior, and in the twentieth century's totalitarian autocracies, where the state permeated civil society, though perhaps less thor­oughly than historians first imagined.³⁸ Of the organizational forms that governed through pervasive informal behavioral control, we find only faint echoes from distant and semidocumented societies and from short-lived experiments of the like-minded, withdrawing from established society to unite in voluntarily adopted uniformity: utopian associations, communes, kibbutzim, and religious com­munities. That is what Christian society was supposed to be like. Christ had only a staff, Luther explained, because Christians were sheep and he their shepherd. The wolves and lions of the mundane world, in contrast, required force. In historical fact, a few such com­munities have been successful, such as the Amish, Mennonites, and the Hasidic town of Kiryas Joel in New York State. Others went spectacularly bad: the Anabaptists in Munster in 1534, the People's Temple in Jonestown in 1978, and the Branch Davidians in Waco in 1993.³⁹

Complex societies have traded off between these two techniques of behavioral control— outside imposition and internal informal self­regulation. At times, civil society has functioned efficiently as the engine of socialization, with family and church acculturating com­munity members. If someone who had violated a deeply felt point of honor was shunned and ostracized or left to commit suicide—thus sparing the law the unpleasant necessity of an execution—that indi­cated that custom and code still reinforced each other.⁴⁰ In ancient Egypt, defendants condemned to death were sometimes allowed to kill themselves, as was Socrates in Athens. Convicted Japanese Samurai, too, were permitted to commit seppuku, or harakiri.⁴¹ Early twentieth-century Melanesia considered it a sign of cultural decline that those who had seriously violated norms were now jailed rather than, as earlier, committing suicide.⁴²

Yet at some point more was needed. Most states began formal­izing rules that had earlier been implicit. Whom one married, how one dressed, what one believed about the supernatural—all became matters of law. The state moved into new arenas of behavioral regu­lation, taking as its remit education, large swaths of childrearing, and more generally the socialization of citizens. Laws replaced or supplemented informal behavioral molding. Raising the next gen­eration became evermore a task for institutions beyond the family. As the church lost influence, religion ceded pride of place to secular morality and etiquette, the last-mile guide to appropriate behav­ior. And as the informal socialization into correct conduct that had been religion and morality's task dissolved in modernity's anomic acid bath, it was displaced in turn by the law as an immediate guide to behavior. Village communities once prodded, scolded, exhorted, and shamed their members to toe the line. In urban society, mat­ters became formalized. In the eighteenth century, laws were passed against vagrancy, prostitution, public drunkenness, sloth, and other behaviors that had earlier been only informally discouraged. Expanding its policing function, the state minutely regulated every conceivable behavior: from how children should address their par­ents to locking doors in the evenings. The state replaced kin as the enforcer.