[P]reventive justice is upon every principle, of reason, of humanity and of sound policy, preferable in all respects to punishing justice. . . .¹

Preventive justice which consists in restraining a man from committing a crime which he may commit but has not yet committed is common to all systems of jurisprudence.²

We all have heard the proverbs “An ounce of prevention is worth a pound of cure” and “A stitch in time saves nine.” These quaint expressions of folk wisdom reflect both the human instinct to anticipate and prevent harms and the human limitation on constructing a precise formula for which sorts of preventive actions are justified under what circumstances. In this chapter, I shall explore briefly the history of preventive actions, particularly with regard to efforts at predicting and preventing serious crimes. My focus will be on the Anglo-American legal system because I am most familiar with it and because it is becoming increasingly influential throughout the world.³

Consider the following hypothetical case, variations of which have recurred throughout the history of humankind. A man riding his horse accidentally runs down a ten-year-old child, killing him. The father of the dead child swears revenge on his child’s killer, threatening to cause “an accident” against the killer’s child. What should the society—whether a small, primitive tribe or a large, complex metropolis—do to prevent further bloodshed? Should the threatening blood avenger be preventively confined until his blood cools? Should the family of the accidental killer be placed in protective custody? Should human nature simply be allowed to run its course? If it were to be decided to confine anyone against his will in this situation, what would be the legal or moral basis for doing so? There is no existing jurisprudence that would authorize the involuntary confining of a person on the basis of what we think he might do. Today we do have laws criminalizing overt threats, but in the old days there were no such laws, and even today the threat has to be explicit. This, in a nutshell, is the problem of individualized prevention that, despite the absence of an explicit jurisprudence, has always been practiced and has always generated controversy.

Oliver Wendell Holmes, Jr., in a celebrated passage from The Common Law, argued that “prevention” is the “chief and only universal purpose of punishment” and that “probably most English-speaking lawyers would accept the preventive theory without hesitation.”⁴ William Blackstone, the eighteenth-century British jurist who had considerable influence on the development of American law, in a chapter entitled “Of the Means of Preventing Offenses,” observed that “if we consider all human punishments in a large and extended view, we shall find them all rather calculated to prevent future crimes, than to expiate the past.”⁵

Other legal authorities have argued in equally categorical terms that prevention has no proper role in the Anglo-American system of criminal justice. Francis Wharton, in his influential nineteenth-century A Treatise on the Criminal Law of the United States, dismissed “prevention” as a “proper theoretical justification” for criminal punishment: “If the [prevention] theory be correct, and be logically pursued, then punishment should precede and not follow crime. The state must explore for guilty tendencies, and make a trial consist of the psychological investigation of such tendencies. This contradicts one of the fundamental maxims of English common law, by which not a tendency to crime, but simply crime itself, can be made the subject of a criminal issue.”⁶

The debate over the proper role, if any, of prevention in criminal punishment has not been limited to Anglo-American legal writers. The Marchese di Beccaria, one of the founders of modern-day criminology, in his classic eighteenth-century Essay on Crimes and Punishments, put forth an essentially preventive justification for the criminal sanction: “It is better to prevent crimes than to punish them. This is the fundamental principle of good legislation. . . . [T]he intent of punishment is not . . . to undo a crime already committed. . . . [It is] no other than to prevent the criminal from doing further injury to society, and to prevent others from committing the like offence.”⁷

Immanuel Kant, the eighteenth-century German philosopher, in his Metaphysical Elements of Justice, took issue with Beccaria. For Kant, it was intolerable to impose punishment for any future-looking purpose: “Judicial punishment can never be used merely as a means to promote some other good for the criminal himself or for civil society, but instead it must in all cases be imposed on him only on the ground that he has committed a crime. . . .”⁸

Some of the disagreement over the proper role of prevention in a system of criminal justice results from a failure to define precisely what was being considered. “Prevention” obviously meant something very different to Holmes, for example, from what it did to Wharton. The furthest thing from Holmes’s mind was any kind of “psychological investigation” for “guilty tendencies” or a system of justice under which “punishment should precede and not follow crime.”⁹ What Holmes meant by “preventive” was simply a forward-looking approach designed to reduce the frequency of harmful events in the future: “There can be no case in which the law-maker makes certain conduct criminal without his thereby showing a wish and purpose to prevent that conduct.”¹⁰

Blackstone also defined “preventive” in a general way: “[A]ll punishments inflicted by temporal laws may be classed under three heads; such as tend to the amendment of the offender himself, or to deprive him of any power to do future mischief, or to deter others by his example: all of which conduce to one and the same end, of preventing future crimes. . . .”¹¹

When prevention is defined as broadly as Holmes and Black-stone defined it, then most authorities, though not Kant, would agree that curtailing future crimes is one permissible function of any legal system.¹²

The issue therefore is not whether prevention should play any role in a system of criminal justice—virtually all commentators (with the exception of strict Kantians) would agree that it should—but how much and what kind of role it should play. There is a considerable difference between a system that confines a youngster who has never committed a criminal act but who is predicted to be a future criminal and one that authorizes confinement only after a consummated harm has been committed, though both may have as their purpose the prevention of future crimes. Prevention, as an element of criminal justice, is best seen as a continuum; some systems authorize preventive intervention of varying sorts relatively early in the process from dangerousness to ultimate harm, while others authorize preventive intervention later.

For the purpose of this historical analysis, it is enough to distinguish three distinct but overlapping approaches to the control of crime that have, with varying emphasis, always been employed in the Anglo-American system of justice.

The Injury or Harm Approach

The first, which has characterized most primitive and simple societies, may be called the injury or harm approach. Cain kills Abel, and God punishes the killer (though perhaps not harshly enough because it was his first offense, indeed the world’s first murder according to the Bible). A serious physical injury, such as a death or maiming, is thought to require a response in kind. “Harm is harm and should be paid for. On the other hand, where there is no harm done, no crime is committed; an attempt to commit a crime is no crime.”¹³ In primitive societies, it is neither the law nor its agents that respond directly to harmful acts; the law simply authorizes those closest to the victim to seek blood revenge. Thus the father of the accidentally killed child was expected to seek revenge against the killer or his family.¹⁴ An “advance [was] marked” when the law moved “towards the suppression of blood-feuds” and private vengeance. This advance, called a “bot,” took the form of setting a price on the limbs and life of various victims. This system of compensation was at least in part preventive since its avowed purpose was “the suppression of blood-feuds.”¹⁵ The bot did, however, leave many dangerous criminals at liberty since its payment ended the “case.”

Another preventive device, employed from the earliest times against criminals who were deemed especially dangerous, was total exclusion from the community and its protection. Exclusion could take the relatively benign form of banishment or the extreme form of outlawry, which was characterized as the capital punishment of a rude age.¹⁶ Indeed, capital punishment (widely viewed today as the vestige of a retributive theory or as an arguable general deterrent) had an important preventive component during an age when the long-term confinement of dangerous offenders was not feasible. As imprisonment came into greater use and dangerous wrongdoers could be locked away, the preventive component of the harm-injury approach became more obvious. But since only those people who had already committed harmful crimes could be locked up, the preventive component of imprisonment was limited by the requirement, inherent in its definition, that no intervention can be authorized until an actual injury has been sustained. Dangerous people, even those who had committed dangerous acts, such as throwing an ax at, but missing, someone, were allowed, at least in theory, to remain free to try again, perhaps this time with better aim. Thus the father of a victim would not be punished until and unless he actually carried out his threat (if it was unlawful to kill the accidental killer of one’s child).

The Dangerous Act Approach

This limitation is not inherent in the second approach, which may be called the dangerous or inchoate act approach. It is similar to the harm-injury approach in that it conditions intervention on the commission of a past criminal act. The essential difference is that under this approach the criminal act need not actually have caused any injury or harm; it is enough that the act is deemed dangerous. Thus, throwing an ax, carrying a concealed weapon, driving above the speed limit, creating a fire hazard, and issuing a credible threat all may be made criminal acts without regard to whether an injury actually followed the act in any particular case. The acts themselves, and the situations they create, are considered dangerous enough to warrant preventive intervention. The purpose of the criminal punishment is to reduce the frequency of these acts because it is assumed that the more these acts are committed or the more these situations are permitted to exist, the more likely it will be that injuries will occur.

Included in this approach are the so-called inchoate crimes, such as attempted murder, solicitation to murder, conspiracy to murder, and incitement to kill. Again, no actual injury need be proved; it is enough that the act was culpable and that it evidenced dangerousness. If the father tried to kill his enemy but failed, he could be punished for attempted murder. Thus, the father of the accidentally killed child could be punished for threatening to kill the child’s murderer. This approach is considerably more preventive than the harm-injury approach since it authorizes intervention, indeed serious punishment, at a considerably earlier point on the danger-harm continuum.

The Dangerous Person Approach

Finally, the most obviously preventive approach may be called the dangerous person approach. It does not require the commission of any past criminal act as a condition to intervention. A person may be confined because it has been predicted that he may commit a dangerous or harmful act at some future time. Most such predictions will in fact be based upon suspicion that the person committed certain past acts, but these acts generally need not be proved, nor need they have been prohibited by law. Some obvious examples of the dangerous person approach are the confinement of predicted saboteurs or spies during wartime, the commitment of mentally ill persons thought to be dangerous, the pretrial preventive detention of criminal defendants on the basis of likely future criminality, the confinement of material witnesses, and the imprisonment of suspected terrorists. Under this approach, the father might be confined in order to prevent him from taking revenge, even if he has not threatened or tried to kill, so long as it is very likely that he would try.

A related phenomenon is the age-old right of self-defense. John Adams, in his closing argument on behalf of the British soldiers accused of perpetrating the Boston Massacre, invoked what he called “the first and strongest principle in our nature,” preventing our own deaths by killing those about to attack us. This too requires a prediction of sorts and an assessment of the dangerousness of the attacker and the imminence of his attack. (As we shall see subsequently, analogies are often drawn to self-defense by those who would justify preemptive military strikes and other preventive measures.)

It is the predictive approach to dangerous persons and the lack of a jurisprudence governing it that is the focus of this chapter.

One important reason why we lack a jurisprudence or philosophy of preventive intervention is that so many intellectual, judicial, and political leaders have denied the legitimacy, even the existence of such intervention throughout our history. If we don’t believe in or practice a particular mechanism of social control, then there is no need to construct a jurisprudence or philosophy that rationalizes and regulates it. Indeed, the very act of articulating such a jurisprudence is sometimes believed to lend legitimacy to an otherwise illegitimate mechanism.¹⁷

Moreover, Anglo-American law has generally evolved pragmatically over long periods of time; a given practice develops over time, is gradually recognized by the courts and commentators, and only then does a jurisprudence emerge. As Roscoe Pound, the former dean of Harvard Law School, put it, “It is true that in Anglo-American law, more than in other systems, juristic theories come after lawyer and judge have dealt with concrete cases and have in some measure learned how to dispose of them.”¹⁸ This is partly at least because the Anglo-American legal system is primarily a common law system that relies on cases, with different factual settings, being decided by the courts over the centuries. The Continental system, on the other hand, relies primarily on codification—that is, the enactment of statutes that speak in general terms and must be applied to specific cases as they arise.

Because lawyers and judges have not often “dealt” with preventive intervention—at least not overtly and systematically—and because they have not really “learned how to dispose”¹⁹ of cases involving such intervention, there has been little occasion to develop juristic theories regarding this important and pervasive mechanism of social control. To illustrate this point, let us look briefly at the Anglo-American history of one retail, or micro, mechanism of preemptive intervention, the preventive detention of individuals believed to be likely to cause harm because of their criminal propensities, status, past behavior, profiles, mental illness, or other assumed markers of dangerousness.

Preventive Confinement: A Catalog of Denials

It has been widely assumed and often dogmatically asserted that the very idea of confining someone preventively is antithetical to the principles of the Anglo-American system of criminal justice and virtually unprecedented. As the great American Supreme Court justice Robert Jackson put it more than half a century ago, “[T]he jailing of persons by the courts because of anticipated but as yet uncommitted crimes [could not be reconciled with] traditional American law. . . . [I]mprisonment to protect society from predicted unconsummated offenses [is] unprecedented in this country . . . [and] is fraught with danger of excess. . . .”²⁰ He was wrong about its being unprecedented (recall the preventive confinement of 110,000 Japanese-Americans during World War II), but he was right about its being “fraught with danger.”

Those who would deny the existence and certainly the legitimacy of preemptive intervention as a mechanism to prevent crimes and other harms point to history and its insistence that specific harm must have occurred before the law (or at least the criminal law) is invoked.

Ancient law has as a general rule no punishment for those who have tried to do harm but have not done it. The idea of punishment is but slowly severed from that of reparation, and where no harm is done there is none to be repaired.²¹ “[English law] had started from the principle that an attempt to do harm is no offense.” There was, of course, one striking exception to the general rule that a harm must already have been done. That exception involved any danger to the King. Not only was any plot against the King a high crime; it was even a crime ‘to compass’—that is, to imagine—the death of the King. When it came to defending the King, all bets were off. Prevention was the key, but for all others the principle remained: “The thought of man shall not be tried.”²²

According to the distinguished historians Frederick Pollock and Frederic Maitland, these words “might well be the motto for the early history of criminal law. . . . Harm is harm and should be paid for. On the other hand, where there is no harm done, no crime is committed. . . .”²³ Even attempts to commit violent harms were not generally punished at common law. Professor Hall emphasizes the fact that “Criminal attempt is conspicuous for its absence in early English law. There is not the slightest suggestion of theory or general doctrine. . . . Apparently in those forthright days, a miss was as good as a mile.”²⁴

These renditions of legal history, ancient and medieval, seem unrealistically wooden to the modern mind. They suggest a society that was willing to tolerate extremely dangerous persons, acts, and conditions in its midst without any effort to prevent future harms—no matter how imminent or certain these harms appeared to be. It is unlikely that “a miss” was ever “as good as a mile,” even in “those forthright days.” If a man’s enemy hurled an ax at him and narrowly missed, it defies common sense to assume that the near victim’s attitude was simply to forget it: “For what harm did the attempt cause, since the injury took no effect?”²⁵ If the father of the accidentally killed child threatened and then tried to kill the person who had killed the child, you can be sure that something would have been done if not by the community then by the potential victim himself or his family. Human nature, even animal nature, reacts to imminent dangers as well as to consummated harms. It is difficult to accept as entirely accurate, therefore, a picture of a society, even a primitive and decentralized society, in which near misses, serious threats, conspiratorial plots, and obvious preparation for harm doing were totally ignored. What of the violently crazy person who had as yet caused no injury? What of the sinister-looking stranger aggressively bearing arms? What of the person suspected of past crimes but not subject to conviction for lack of sufficient evidence? Can it really be the case, as the commentators suggest, that such “dangerous” persons and “inchoate” acts were ignored unless they involved a plot against the king’s life?

It may well be true, as Hall argues, that there “is not the slightest suggestion of theory or general doctrine” justifying liability for wrongdoing that “fell short” of actual physical injury. But the history of humankind—even beyond Anglo-American law—is one of pragmatic actions, followed, often only centuries later, by doctrinal or theoretical justifications. Even the Bible begins with stories of crimes, sins, and primitive efforts to prevent them in the Book of Genesis, before moving to the codification of laws in the Book of Exodus. This has certainly been the history with regard to preventive or preemptive warfare, as we shall see in the next chapter. Pragmatically, it is difficult to conceive of a society that ignored serious threats of future harm and always waited for past harms before it intervened.

From the beginning of recorded history, societies have worried about dangerous people who had not “yet” done the harm it was believed—or predicted—they would do. Preventive imprisonment, even in times when places of confinement were primitive, was employed in some systems. For example, the Bible requires two witnesses to convict a murderer. If only one credible witness testified, the defendant had to be acquitted,²⁶ but that did not necessarily mean that the dangerous killer would be allowed to go free and perhaps murder again. Instead he was placed in a locked room, where he was sometimes fed a lethal concoction of water and grain calculated to cause his stomach to burst.²⁷ The rabbis justified this extrabiblical punishment as necessary to prevent lawlessness.²⁸

I suspect that most societies developed similar informal or extrajudicial mechanisms for dealing with obviously dangerous people who could not be convicted of past crimes.

This ancient example suggests a general rule about the need for preventive mechanisms. The need for preventive intervention becomes more obvious in a society that has circumscribed the formal process of criminal justice with safeguards that make conviction more difficult. Where conviction based upon suspicion (or “bad fame”) is relatively easy, then the formal system can effectively play a preventive role: It can convict those who are feared likely to commit future crimes. But as procedural and substantive safeguards are added over time, then one obvious result is that more and more dangerous people will evade the formal criminal process and remain free in society. This will create a perceived need for preventive devices capable of incapacitating the person feared as dangerous but not subject to conviction for past crimes.²⁹ A close look at the history of the actual administration of justice, both private and public, will corroborate the commonsense conclusion that crime prevention has always played a significant, though largely unarticulated, role in the Anglo-American legal system.

Dual Systems of Criminal Justice: Retrospective and Preventive

Throughout Anglo-American history two criminal justice systems have operated side by side. The more formal system (the one we are most familiar with) has been characterized by high visibility, by progressive common law development through appellate decisions, by frequent legislative revision, and by bookshelves of treatises and other scholarly descriptions and discussions of its philosophy, substance, and procedure. Perhaps most important, it has been characterized by a well-developed jurisprudence imposing principled limits on its employment. Among the most fundamental principles of this system is that criminal punishment must always be based upon a past act or omission, not upon predicted conduct likely to occur at some future time.

Proof beyond a reasonable doubt of a specific criminal act, such as murder or robbery, was the switch that triggered the on-off criminal process that authorized the government to punish the accused, generally by execution, since imprisonment was not yet in wide usage. As more calibrated responses became available, more nuanced uses of the criminal law became feasible. When “moderate” punishments could be imposed, it became morally acceptable to use the criminal law somewhat more preventively—for example, to encourage people to take precautions and not to engage in dangerous conduct that had not yet caused any harm.

Thus, to the extent that those historians and judges were purporting to describe the early formal system, they probably came close to the mark. With a few special and limited exceptions, the formal Anglo-American legal system has not authorized the “jailing of persons by the courts because of anticipated but as yet uncommitted crimes.”³⁰

There always has been, however, a parallel system that has played a significant role in the Anglo-American legal process. It was less formal and less visible. That is why we know so little about it today, though it was probably well known in its own day. It was characterized by the absence of published opinions and appellate review. It has always been less principled or at least less willing to articulate governing principles. Few treatises or commentaries have been written about it. Most important, it has never developed an articulated jurisprudence that imposes principled limits on its employment. Its primary function has been to fill the gaps that inevitably occur in any formal system of criminal justice. This less visible system has taken on renewed importance because it has been widely employed during emergencies, including the post-9/11 period.

In the earliest times the two systems were essentially merged because the earliest law in general lacked formality, appellate decisions, an articulated jurisprudence, and principled limitations. Thus there was little tension between principle and practice requiring the creation of separate systems.³¹

A more formalized system of criminal justice began to develop in England after the Norman Conquest. By the end of Henry II’s reign, near the end of the twelfth century, “there [was] a permanent central tribunal of persons expert in the administration of justice—of sworn judges.”³² There was also the establishment, in primitive form, of trial by jury, formal writs and pleas, the “king’s peace,” felonies, and the beginnings of a system of fundamental rights (to be solemnized in the Magna Carta in 1215).³³ A time traveler going back to the first millennium would hardly recognize our current legal system, but if that traveler then zipped forward to the thirteenth century, he would see some familiar attributes of the common law system of justice.

It is thus not surprising that at about the same time that the formal law began to develop, serious gaps began to be noted in the increasingly formalized, but “exceedingly inefficient,” system of justice.³⁴ It was necessary therefore to develop a less formal system that could operate with an eye more to efficiency than principle and with the goal of “preventing the breach of the peace (wisely foreseeing and repressing the beginnings therof)”³⁵ rather than wait until crimes actually were committed.³⁶ This preventive system, residues of which can still be seen, centered on the office of the conservator of the peace, later called by the more familiar title justice of the peace, a cross between a policeman on the beat and local magistrate.

The Development of a System of Preventive Justice

The development of the special office of conservator of the peace, accomplished by the time of Richard I in the twelfth century, was in large part a measure designed to fill the gaps in the formal system of criminal justice and to check incipient crime.³⁷ Knights, who were specially assigned to this task, were instructed to summon before them all males over the age of fifteen and “cause them to swear that they would not be outlaws, robbers, or thieves. . . . ”³⁸ These custodians—or keepers—of the peace were instructed to prevent homicide, incendiarism, robbery, and extortion. They were also authorized to prevent the bearing of arms without license, certainly an inchoate offense reminiscent of later gun control laws. As one might expect, among their functions was “to take every precaution to repress” uprisings against the royal authority at the earliest stages before any harm had yet occurred.³⁹

The office of conservator of the peace changed names a number of times: guardian of the peace, keeper of the peace, and, eventually, justice of the peace. But the functions of the office, though constantly expanding and taking on more administrative and judicial duties, retained a common core. And that core always included a large preventive component.

In the fourteenth century lawlessness was rampant, and Parliament was convened a number of times to concern itself with problems of law and order. In 1360 a statute empowered the justices of the peace “to enquire of all those that have been pillors [sic] and robbers in the parts beyond the sea, and be now come again, and go wandering, and will not labor as they were wont in times past. . . .”⁴⁰ This provision was explicitly preventive in that it authorized inquiry of past malefactors who had been punished but who had returned and were still feared. The justices were authorized “to take and arrest all those that they may find . . . by suspicion, and to put them in prison. . . .” They were also authorized to “take all of them that be [not] of good fame” and to require them to post “sufficient surety and mainprise of their good behavior”⁴¹ (i.e., trusted persons willing to vouch for him and to underwrite a recognizance bond).

By the time of the Tudors, in the sixteenth century, systematic legislation had been enacted against the poor and vagrant classes, which were thought to include a large number of dangerous persons. Poverty, idleness, drunkenness, vagrancy, and the like were seen as early markers of potential criminality.⁴² The preamble to a statute enacted in the early sixteenth century declared that idleness among vagabonds and beggars was the “mother and root of all vices, whereby hath insurged and sprung and daily insurgeth and springeth continual thefts, murders, and other heinous offences and great enormities . . . to the marvelous disturbance of the common weal of this realm. . . .”⁴³

Later statutes required the justices “to direct night searches for rufflers, vagabonds, and other suspects, and punish all such offenders.”⁴⁴ If the captured “rogue was a dangerous character,” then “two of the justices could commit him to gaol or the house of correction to await quarter sessions.” If he was convicted there, banishment from the realm would follow.⁴⁵ Vagrancy laws of this kind were in large part preventive in origin and function.⁴⁶ These statutes—indeed, the entire jurisdiction of the justice of the peace—filled the interstices of the formal criminal law, which was in large part, though never entirely, retrospective.⁴⁷

The need for preventive law throughout England was apparently great. Crime and violence were rampant, the formal criminal law was cumbersome, the “regular” criminal courts met in many locations at very long intervals, and the penalty structure was relatively inflexible. Thus the need was obvious for a less formal, more flexible, less principled, and more effective mechanism for dealing with those dangerous people who could not adequately, or swiftly, be dealt with under the formal system of criminal justice. The justice of the peace embodied that role.

The Justice of the Peace

Michael Dalton was one of those justices of the peace during the first half of the seventeenth century. He wrote a remarkable diary-treatise of the day-to-day operations of his office that remains the best contemporaneous account of this important job.⁴⁸ Dalton outlined the functions of his office as follows:

The conservation of this peace (and therin the care of the Just[ice] of Peace) consisteth in three things, viz.

1. In preventing the breach of the peace (wisely foreseeing and repressing the beginnings therof) by taking suretie for the keeping of it, or for the good behaviour of the offendors, as the case shall require.

2. In pacifying such as are in breaking of the peace. . . .

3. In punishing (according to Law) such as have broken the peace.⁴⁹

These important officials combined the functions of today’s police, prosecutors, judges, juries, and correctional officials. It is significant to note which of these functions Dalton deemed most specifically allocated to the justices of the peace: “But of the three, the first, the preventing Justice, is most worthy to be commended to the care of the Justices of peace.”⁵⁰

Dalton detailed the kinds of people against whom preventive action could be taken by the justices of the peace. It is evident from his listing that the justices exercised a preventive jurisdiction that did not require, as a condition for their intervention, that a crime already had been committed. It was enough if a person “is minded to break the peace”⁵¹ or if he “be in a fury ready to break the peace.”⁵² Witchcraft trials may also have had a significant preventive component. The contemporaneous descriptions of the personality types associated with witches would seem to bear this out: “Those who were boastful, illiterate, miserable, lustful, and leading a ‘lewd and naughty kind of life,’ melancholy—all were likely to be witches. Above all, they were thought to be the type of person who went round begging and those who had vicious tongues. Witches were people of ‘ill natures, of a wicked disposition, and spitefully malicious’; ‘malicious people, full of revenge, having hearts swolne with rancour.’ ”

There did not seem to be any requirement of a specific overt act (though it is likely that one was suspected in most cases). It was sufficient if “in the justices conscience [he or she] was a dangerous person”⁵³ or persons “such as are like to commit murder, homicide or other grievance to any of the Kings subjects in their bodies” or one “of evil fame or report generally” or any “suspicious person in the night”⁵⁴ or any person “suspected to be inclined to the breach of the peace.”⁵⁵ The underlying requirement seems to have been that “there is a fear of some present or future danger, and not merely for a [crime] that is past.”⁵⁶ Thus, the mere fact that an accused person was “acquitted of a felony” did not eliminate the authority of the justice of the peace: “[I]f he be of evil fame, or of evil behaviour, it seemeth the Justices of Peace upon their discretion, may binde him to his good behaviour.”⁵⁷ If the “defendant” failed or refused to produce satisfactory sureties, “then the Justice of Peace may commit him to the Gaol.”⁵⁸

In today’s world this wide discretion to apprehend anyone who looked “suspicious,” “dangerous,” or “inclined” to commit a crime would be a police officer’s dream and a civil libertarian’s nightmare. But as we shall see, when the formal and informal systems began to merge, especially during the last half of the twentieth century, the protections of the formal law were extended to preventive police activities, and arrests on “mere suspicion” became impermissible. However, until relatively recently the justices of the peace retained considerable discretion to respond to complaints about dangerous characters wandering about.

Following the attacks of 9/11, this discretion was once again invoked in an effort to prevent terrorism, by casting a wide net around people of Middle Eastern extraction who were suspected of association with terrorism. Some were held as “material witnesses,” others were arrested for minor offenses and held without bail, while still others were confined for violating immigration laws. Pursuant to the mandate of the U.S. Justice Department, the law was becoming more proactive and preemptive. It remains to be seen whether these “emergency” measures will outlive the emergencies that generated them.

Even in the days of the justice of peace it was not enough to sit back and wait for complaints to be made against rogues, vagabonds, and others of this sort; the justice’s job was to search them out, as illustrated by a document entitled “A warrant for a general search for Rogues.”⁵⁹ This warrant commanded the constables to “make a general privie search within every of the said several towns . . . upon (a specified date) at night . . . for the finding out and apprehending of all Rogues, Vagabonds, and wandering and idle persons. . . . And if any of the said rogues shall appear to be dangerous or incorrigible (they shall be tried and committed to the house of correction, or gaole).”⁶⁰

One area of the justice of the peace’s role that has been neglected by historians but that warrants special attention in the context of crime and terrorism prevention is the control of weapons. This type of control reflects a kind of prevention different from that exercised over dangerous persons; this is prevention directed at dangerous situations (or the combination of dangerous persons and situations).

The justices of the peace were authorized to take direct action against “any person [who] shall ride or go armed . . . in Fairs, Markets, or elsewhere (by night or by day) in affray of the Kings people. . . .” The justices “may cause them to be staid and arrested, and may bind all such to the Peace or good behaviour (or, for want of sureties, may commit them to the Gaol).” They were also authorized to “seize and take away their armor, and other weapons, and shall cause them to be prised, and answered to the King as forfeited.”⁶¹ It was not a crime to carry weapons of various kinds, yet the justices of the peace, in their discretionary power to repress “the beginnings” of crime, were empowered to “take away” any “gun, dag[ger], or pistols charged” or any other “armor worn in terrorism” or weapons that appeared likely to be used against the peace.⁶² There were absolute prohibitions against possession of certain kinds of weapons or by certain classes of persons: “No person may shoot in, carry, keep, use or have any hand-gun under one whole yard in length, nor any other Gun (Dag, or Pistoll) that shall be under three quarters of a yard in length.”⁶³ (This proscription against concealable weapons seems to be a direct predecessor of modern concealed handgun legislation.) Poor people—anyone who “hath [not] per annum 100. li”—were forbidden to have “any Gun, Dagge, Pistoll, Crossbow, or Stonebow. . . .” And any justice of the peace could take such weapons from a poor person found in possession of them. Indeed, every person “having in lands an hundred pounds by the year” was made a “private” justice of the peace for purposes of enforcing the weapons law. Any such person “may from such Malefactors, and to his own use for ever keep” any weapon found in the possession of an ineligible poor person.⁶⁴

Another “dangerous situation” about which the justices of the peace seem to have had special concern was dangerous or unlawful “assemblies.” It was believed that in the past such assemblies “began upon very small occasion, yet not being repressed in time, grew to such greatness and height, that they afterwards put in hazard the state and government of this Land.” Accordingly, “it is behovefull and good wisdome for all Justices of P. to indeavour by all good meanes to quench the beginnings and first sparks of such assemblies. . . .”⁶⁵

Today, of course, the right to assemble is protected by the Constitution of the United States and the common law of Great Britain, but some “assemblies”—lynch mobs, gatherings in certain sensitive locations—can be stopped even today if they are deemed too dangerous to the peace.

Unlawful assemblies were defined by the justices of the peace to include what is today covered by conspiracy laws (except that three persons or more rather than two persons or more had to be involved). These were inchoate crimes that did not require any harm (or substantive crime) to have been committed: “[A]lthough they shall after depart of their own accord, without doing anything, yet this is an unlawfull assembly.”⁶⁶ Nor was any complaint needed: “[E]very Justice of peace, hearing of any Rout, or of any intention of a Rout (without . . . tarrying . . .) shall do well to go himself . . . to the place where such persons be so assembled, and to suppress them . . . and to force them to put in surety for the peace . . . and also he may take away their weapons and armour. . . . ”⁶⁷

The justices of the peace had a similar preventive role in controlling “affrays. If the affray be ‘dangerous,’ then the affrayors” may be commanded to “prison for a small time, till their heat be over.” If “men shall contend only in hot words, this is no affray,” and the constable would have no jurisdiction to intervene. But the justice could still take preventive action, “for from (hot words) oftimes doe ensue Affrays and batteries, and sometimes maimes, yea manslaughters and murders.” This approach to “hot words” presages our contemporary approach to what the Supreme Court has called fighting words.⁶⁸

Though the jurisdiction of the justices was predominantly “preventive” and they were authorized to act against persons (even to incarcerate them) who had not “yet” committed any harmful or even criminal act, it seems likely, from what Dalton wrote, that “suspicion” of past misbehavior was often taken into account in deciding if one was to be regarded as a dangerous person. Indeed, the rules of evidence set out by Dalton make it clear that no sharp distinctions were drawn between evidence of past guilt and evidence of future dangerousness.

Among the “circumstances” that were to be “considered” in the “examination of Felons” were the following:

His parents, if they were wicked, and given to the same kind of fault.

His ability of body, if strong and swift, or weak or sickly, not likely to doe the act.

His nature . . . a quarreller, pilferer or bloody-minded. . . .

His trade; for if a man liveth idely or vagrant . . . it is a good cause to arrest him upon suspicion, if there has been any felony committed.

His company. . . .

His course of life. . . .

Whether he be of evill fame or report.

Whether he hath committed the like offense before or if he hath a pardon, or been acquitted of felony before . . . or beene outlawed for felony. . . .⁶⁹

It is obvious that these items of “evidence,” some of which are descriptive of character and not specifically related to any particular past offense, were as probative of future dangerousness as past guilt. They constitute an early form of profiling. Today we employ ethnicity, religion, and race, but in those days virtually every Englishman shared those characteristics and so poverty, heredity, and reputation became the markers. The essential difference between convicting someone of a past crime and preventing him from committing a future crime lay therefore not in the nature of the evidence that could be used but rather in the burden of proof and in the acceptable ratio of false positives to false negatives. There had to be greater certainty that the correct person was convicted of a past crime.⁷⁰ Hence, the more stringent requirements of trial by jury, two justices, heavier burdens of proof. Preventive justice, on the other hand, could afford to be overinclusive, to include more “false positives.” Dalton candidly acknowledged this and said that it “is no new thing”: “There is such a precedent in the old book of justices of peace . . . yea, it is the common practice at this day, and it seemeth to be very serviceable; and of two evils the lesser is to be chosen, that an offendor, or suspected person, should be imprisoned for a time (though sometimes wrongfully) than that one which hath committed felony should escape unpunished.”⁷¹

If the credo that it is better for an innocent person to be wrongfully imprisoned than for a guilty one to escape reflected the actual “practice” of the justices in administering preventive justice during the early seventeenth century in England, it seems strikingly at variance with the famous credo of the formal common law “that it is better that ten guilty persons escape, than that one innocent suffer.”⁷² This credo refers of course specifically to the burden of proof for a conviction for felonies already committed. Even Blackstone suggested a different standard of certainty with regard to preventing future crimes. In cataloging the kind of person who may be subjected to what he called preventive justice, one who may be required to produce sureties or who “may immediately be committed till he does,” Blackstone acknowledged that there is “so great a latitude, as leaves much to be determined by the discretion of the magistrate himself.”⁷³ But he did not seem to object to the possibility that numerous innocent persons may suffer confinement as a consequence of this preventive procedure: “The caution, which we speak of at present, is such as is intended merely for prevention, without any crime actually committed by the party, but arising only from a probable suspicion, that some crime is intended or likely to happen; and consequently it is not meant as any degree of punishment, unless perhaps for a man’s imprudence in giving just ground of apprehension.”⁷⁴

Blackstone regarded the peace bond (and the analogous good behavior bond) as an innovation unique to English jurisprudence: “And really it is an honour, and almost a singular one, to our English laws, that they furnish a title of this sort [the title being ‘of the means of preventing offenses’]: since preventive justice is upon every principle, of reason, of humanity, and of sound policy, preferable in all respects to punishing justice;⁷⁵ the execution of which, though necessary . . . is always attended with many harsh and disagreeable circumstances.” Blackstone was reasonably careful in his use of the term “preventive justice.” He defined it as a restraint “intended merely for prevention, without any crime actually committed by the party, but arising only from a probable suspicion, that some crime is intended or likely to happen. . . .”⁷⁶

It is still quite remarkable that legal commentators seemed to accept so great a disparity between the preference for one type of error over another, depending on whether confinement was deemed punitive or preventive. It was, after all, still confinement. With regard to the former, the strong preference was for acquitting the possibly guilty over convicting the possibly innocent (“better that ten guilty persons escape than that one innocent suffer”). With regard to preventive confinement, on the other hand, the strong preference was for imprisoning the possibly innocent over freeing the possibly dangerous: “of the two evils, the lesser is . . . that [a] suspected person, should be imprisoned for a time (though sometimes wrongfully) than that one which hath committed felony should escape unpunished.”⁷⁷ One difference of course was that the punishment for past felonies was often death, whereas the preventive sanction was imprisonment “for a time.” But that difference alone would not seem to justify so great a disparity between the highly formalized jurisprudence of punishment and the largely discretionary and ad hoc practice of prevention, especially since imprisonment in English jails carried with it a significant risk of death.⁷⁸

Early English jails were generally designed for short-term,

temporary detention rather than for long-term incapacitation and punishment. The erection of institutions suitable for long-term confinement was probably regarded as needless expense and burden in a land that was still sparsely enough settled that its wandering undesirables could be shunted from isolated community to isolated community and to a people who were still callous enough to tolerate the inexpensive sanction of execution for a wide range of criminal offenses.

Imprisonment was apparently employed, even in America during colonial times, against some persons who could not secure sureties to keep the peace or for good behavior. This institution—the preventive justice administered by the justices of the peace—was transplanted from across the ocean. Justices of the peace were appointed in the colonies and their jurisdiction and powers paralleled those exercised in the mother country. As in Britain, much was left to the discretion of the justices. Their mission was broad: “[T]he prevention of . . . mischief which [persons of evil fame] may be chiefly suspected to be likely to do; and . . . to sever the public from that danger which may probably be apprehended from their future behavior, whether any actual crime can be proved upon them or not. . . .”⁷⁹

The restrictions on their power were few. As an early American commentator said about one of the original statutes empowering justices of the peace to keep the peace, “[I]t is certain it hath been carried much further [than its original purpose] by construction, and the purport of it hath been extended by degrees, until at length there is scarcely any other statute, which has received such a largeness of interpretation.”⁸⁰

The available records provide little information on how frequently the peace or good behavior bond was actually employed in the colonies or how often imprisonment was actually ordered for failure to obtain a surety. There is evidence that these devices were employed in certain of the colonies, but it seems likely that actual confinement was not relied on as extensively as in England during that same period.”⁸¹

The authority of the justice of the peace to confine dangerous people on preventive grounds persisted even after the adoption of the Constitution and the Bill of Rights, though the duration of the confinement that could be ordered without a jury trial was generally limited, except in the case of the dangerously mentally ill.⁸² This authority lasted until well into the twentieth century, when the courts began to apply constitutional restrictions more aggressively to the police. The events of September 11 and the fear of terrorism may, however, restore some preventive authority to the police, the FBI, and other contemporary keepers of the peace.

Some Tentative Historical Conclusions

This brief history of one type of preventive intervention suggests that the confinement of dangerous persons who cannot be convicted of past criminality but are thought likely to cause serious injury in the future has always been practiced to some degree by every society in history regardless of the jurisprudential rhetoric it has employed. Moreover, it is likely that some forms of preventive confinement, as well as other preventive and preemptive measures, will continue to be practiced by every society.

It suggests that at least some of what has passed for preventive confinement has really been mechanisms for detaining persons who are thought to have committed past crimes but who, for one reason or another, cannot be convicted. Thus the original jurisdiction of the justices of the peace, though preventive, often required a “finding” (but a far less rigorous finding) of past criminality.⁸³

This historical material suggests an interesting hypothesis: that all other factors being equal, the necessity for and emergence of informal preventive mechanisms will increase as it becomes more difficult to secure convictions against dangerous persons thought to be guilty of past serious crime. Mechanisms of social control, more specifically of confinement and isolation, frequently operate on a balloon principle: If you squeeze the air out of one end, the other will become more inflated. Thus it is a fairly constant phenomenon in most societies that dangerous people will be isolated by one means or another: If banishment is available, imprisonment may be employed less frequently; if there is no law of criminal attempts, then peace bonds may be used against those who attempt, but fail, to do harm; if insane asylums become available, then vagrancy and poor laws will be used less frequently against the insane.⁸⁴ This principle is useful, at least, as a caveat: When a particular mechanism of confinement is rendered useless (or less useful), one should at least search to see if the slack has been picked up, in whole or in part, by other mechanisms.

The most disturbing conclusion—and most relevant to this book—is that although preventive confinement has always been and will always be practiced, no systematic and widely accepted jurisprudence of preventive intervention has ever been developed. There were some rules, to be sure, such as those cataloged by Dalton and others. But these rules seem to reflect the ad hoc practices of the justices rather than a fully articulated legal or moral philosophy. It may sound surprising, even arrogant to say that no jurisprudence governing preventive confinement has ever been articulated, but it appears to be true. No philosopher, legal writer, or political theorist has ever, to this writer’s knowledge, attempted to construct a systematic theory of when it is appropriate for the state to confine preventively. This is so for a number of reasons. The mechanisms of prevention have been, for the most part, informal; accordingly, they have not required articulate defense or justification. Moreover, there are many scholars who simply deny that preventive intervention, especially preventive confinement, really exists, or if they acknowledge the existence of these mechanisms, they deny their legitimacy, thus obviating the need for a theory or jurisprudence. Finally, it is extremely difficult to construct a theory of preventive confinement that neatly fits into existing theories of criminal law and democracy.

The upshot, however, is that there has always existed a widespread series of practices, involving significant restraints on human liberty, without an articulated jurisprudence circumscribing and limiting its application. People are confined to prevent predicted harms without any systematic effort to decide what kinds of harms warrant preventive confinement, or what degree of likelihood should be required, or what duration of preventive confinement should be permitted; or what relationship should exist among the harm, the likelihood, and the duration.⁸⁵ This is not to say that there currently exists a completely satisfactory jurisprudence or theory justifying the imposition of punishment for past acts. (Recall Pound’s observation that “in Anglo-American law, more than in other systems, juristic theories come after lawyer and judge have dealt with concrete cases and have in some measure learned to dispose of them.”)⁸⁶ But at least many of the right questions have been asked, and some interesting answers have been attempted. Even the primitive statement that “it is better that ten guilty persons escape than that one innocent suffer” tells us something important about how to devise rules of evidence and procedure. There is no comparable aphorism for preventive confinement: Is it better for X number of false positives to be erroneously confined (and for how long?) than for Y number of preventable harms (and of what kind?) to occur? What relationship between X and Y does justice require? We have not even begun to ask these questions in a systematic way, or to develop modes of analysis for answering them, beyond clichés extolling the virtues of “an ounce of prevention” or “a stitch in time.” We must come to understand and acknowledge the harms that could be caused by ounces of preventive warfare or by stitches of preventive confinement.⁸⁷

This brief historical account of one paradigmatic genre of preemption—preventive confinement of individuals believed to be dangerous—is intended to illustrate a more general phenomenon, namely, that preemptive actions of many kinds have been far more common in practice than they appear in theory. Similar historical accounts, though with varying content and rationalizations, could be provided for other preemptive mechanisms as well. For example, as we shall now see, preemptive military attacks on enemies who may themselves be contemplating attack are at least as old as the Bible.