ADVERSARY SYSTEM

The term adversary system sometimes characterizes an entire legal process, and sometimes it refers only to criminal procedure. In the latter instance it is often used interchangeably with "accusatorial procedure," and is juxtaposed to the.... View more

ADVERSARY SYSTEM

The term adversary system sometimes characterizes an entire legal process, and sometimes it refers only to criminal procedure. In the latter instance it is often used interchangeably with "accusatorial procedure," and is juxtaposed to the "inquisitorial," or "non-adversary," process. There is no precise understanding, however, of the institutions and arrangements denoted by these expressions.

Nevertheless, several characteristics are commonly associated by American lawyers with the adversary criminal process. These include a relatively passive tribunal that ideally comprises both judge and jury; the presentation of evidence by the parties through their lawyers, who proceed by direct questioning and cross-examination of witnesses; the representation of state interests by one of the parties, the prosecutor; a presumption that the defendant is innocent until proved guilty; and the principle that he cannot be forced to testify against himself. The contours of the adversary system remain uncertain because the phrase has been used to describe three distinctive, albeit related, meanings.

The traditional meaning

In Anglo-American jurisdictions the phrase evokes both the aspirations and the actual features of Anglo-American criminal justice. It is incorporated to some extent into American constitutional law through provisions dealing with assistance of counsel and due process of law.

The attributes of "adversariness" change according to context. When techniques of ascertaining facts and deciding legal issues are discussed, a central feature is seen as a confrontational style: prosecution and defense prepare and present their cases to the court, and a decision is reached on the basis of the two alternative versions of fact and law. In this variant, partisan advocates are an essential aspect of the system, with their partisanship supported by canons of legal ethics (Fuller, p. 32). There is some equivocation, however, in the case of the public prosecutor, who is recognized to have a public responsibility that imposes limits upon the allowable degree of partisanship.

When the position of the criminal defendant is at issue, the focus shifts. The mainstay of the adversary system resides in the privilege against self-incrimination (Malloy v. Hogan, 378 U.S. 1, 7 (1964)), which implies high obstacles to conviction and an opposition to unbridled crime control. Any lowering of the evidentiary barriers erected to protect the defendant, such as the requirement of a unanimous jury verdict of guilt, is treated as a step away from the adversary ideal.

Adversary features are found not only in the contested trial but also in appellate proceedings, where arguments by the parties must ordinarily precede the decision of the appellate court. Even the pretrial phase of the criminal process is increasingly seen as displaying adversary characteristics. The privilege against self-incrimination, for example, now radiates into the earliest police inquiries, according protection to the suspect. The right to pretrial release and the hostility to preventive detention are also linked to the adversary system, particularly its emphasis on the presumption of innocence. On the other hand, the widespread practice of negotiations between the prosecution and the defense (plea bargaining) is usually treated as subverting adversariness. Where the defendant pleads guilty after such negotiations, the core of the adversary system, the contested trial, does not take place; moreover, the pressures used to encourage guilty pleas threaten the adversary principle that the defendant may not be forced to incriminate himself. At the same time, however, plea bargaining is quite "adversary" in the sense that it is dominated by the parties and their lawyers, rather than the court.

The adversary system has its distinctive source in liberal ideology. Consider, for example, the image so often used by lawyers of "balancing advantages" (or maintaining an "equality of arms") between the prosecution and the defense; such a goal makes sense only in light of liberal theories that treat the state interest as analogous to—and not superior to—private interests. The presumption of innocence, the requirement of proving guilt beyond a reasonable doubt, and related notions are also suffused with liberal values. Moreover, the passive attitude of the decision-maker has an affinity with the passive laissez-faire ideology.

It is this linkage to ideological currents that has produced two versions of the adversary system in its traditional meaning. In the "classical" variant, the ideal judge is propelled into action only to resolve disputes between the contending parties. The emergence of welfare-state liberalism has generated changes in this version of the idea; just as modern liberal governments intervene in the economy to correct failures of competitive markets, so, according to this view, an adversary judge should intervene in the trial to redress the competition of the parties. Whereas the classical variant celebrates the parties' dominance over the process, a later variant would curb this dominance (Fuller, p. 41). But there is disagreement over the extent to which the judge can intervene without negatively affecting the incentives of the prosecution and the defense for the zealous action required by the adversary system. Some see a solution to failures of party competition not in making the judge more active, but rather in replacing "ineffective" advocates by more capable ones.

It is plain that the adversary system in both its traditional senses is inextricably linked to legal ideology. It is praised in many quarters as a palladium of liberty and contrasted with an antipodal "inquisitorial" criminal process, that term serving to convey the worst features of continental European criminal justice prior to its reform in the wake of the French Revolution. Any departure from adversary features is said to imply a lapse into a system where searches are unbridled, the accused is detained without limits, his confession is coerced, counsel is denied him, and he is not accorded the benefit of doubt. This overdrawn polarization is reflected in such important judicial decisions as Miranda v. Arizona, 384 U.S. 436, 460 (1966).

The adversary system is extolled not only because of the protection it accords the accused, but also because its competitive style of presenting evidence and argument is thought to produce a more accurate result than an "inquisitorial" alternative, where the judge monopolizes proof-taking. According to this view, the judge who conducts an apparently nonpartisan inquiry cannot truly keep an open mind and lacks sufficient incentives to do a proper job. The possibility of a tension between the goals of obtaining accurate results and maintaining high barriers to conviction is often denied. It is occasionally conceded, however, that such barriers, while they lessen the possibility of convicting an innocent person, also increase the possibility that the guilty may escape conviction. Hence, by keeping these barriers high, as mandated by the adversary system, the accuracy of outcomes in the total number of cases irrespective of the kind of error can well be decreased. Where this is recognized, proponents of the adversary system accord decisive weight to liberal values: it is better to let a larger number of the guilty go free than to convict a smaller number of innocent persons.

The traditional Anglo-American concept of the adversary system has often been criticized by lawyers from other legal cultures. It has been vigorously questioned whether the clash of two zealous partisans represents the best instrument of discovering the truth. Moreover, the ample opportunities for the defendant to escape conviction have been said to exist mainly for those able to retain high-powered counsel. Finally, the practical importance of the adversary system in America has been doubted in view of the fact that most criminal cases never reach the stage of a contested trial but are settled through negotiations between prosecution and defense in the course of plea bargaining.

The adversary system as traditionally understood has its domestic enemies as well. Early in the twentieth century an eminent American legal scholar attacked it as inspired by a "sporting theory of justice" that treats substantively correct outcomes as relatively unimportant (Pound, p. 404). It is testimony to the continuing vitality of the traditional concept, however, that most critics castigate the alleged excesses of the system but fail to formulate alternatives to it. Only occasionally is inspiration for fundamental change sought in the non-adversarial modern criminal justice systems of Western Europe (Weinreb, pp. 117–146; Schlesinger, pp. 382–385).

The traditional concept of the adversary system evokes both actual features of Anglo-American criminal process and its aspirations. Inevitably, therefore, it combines both descriptive and prescriptive elements and cannot be expected to achieve rigorous internal consistency and coherence. It is not so much analytically precise as it is hortatory and rhetorical, aimed at mobilizing consent and at winning points in legal argumentation.

A model of conflict-solving procedure

A second way to view the adversary system is as a theoretical model. Conflict resolution is posited as the goal of the process, and the adversary model is then understood to comprise those procedures that implement this goal most effectively. In this second sense, then, the adversary system is a blueprint designed to promote the choice of certain procedures. Elements of the blueprint and features traditionally classed as adversary do not coincide.

Two methods have been used to construct the theoretical model of the adversary process. One method begins from the initial state of conflict between two sides and conceives of the ideal conflict-solving process as a simulation of, and substitute for, the private war between them. This leads to the central image of proceedings as a contest of two sides before the conflict-resolver. The task is then to develop procedural arrangements logically following from this central image. For example, if the adversary judge were permitted to inquire into facts not in dispute between the parties, the proceedings to determine these facts would "logically" cease to be a party contest. Consequently, the adversary model denies to the judge any independent powers to inquire into facts.

The other method starts from the desired end, which is said to be the acceptance of the court's decision by the disputants. The task here is to identify those procedures most likely to produce such acceptance, beginning with the premise that the goal of acceptance is promoted where the parties are permitted to exercise control over procedural action. In contrast to the first method, which relies on logical analysis, the second relies on observation and experiment. For example, whether participation of lawyers is an integral feature of the model hinges on whether such participation contributes to the control of the parties over the process.

As a model of a conflict-solving process, the adversary system is known in both continental European and Anglo-American legal cultures. Under the label "accusatorial proceedings" the model has a long history on the Continent.

The continental legal culture. Efforts to construct an ideal conflict-solving process are to be found in twelfth-century Roman Catholic ecclesiastical scholarship. By the fourteenth century, Italian students of procedure included in accusatorial proceedings many features now incorporated in the adversary system. But the most inclusive models of the conflict-solving process are products of rationalist "natural law" scholarship at the turn of the nineteenth century.

These models appear extremely "adversarial" even from the perspective of Anglo-American legal culture. Termed "the party-dominated process" (Parteiverfahren) by German legal theorists, they deserve brief description. Under them, the judge cannot initiate or continue proceedings without an actual dispute. Parties control the factual and, to a great extent, the legal boundaries of the case. Pleadings and stipulations are necessary devices to define and narrow issues, and the judge is not permitted to overrule such mutual arrangements. The court is also denied the power to call witnesses on its own initiative. Even the court's powers of interrogation, otherwise very important on the Continent, are seriously curbed: only questions suggested by the litigants can be asked. Party "autonomy" is thought to be incompatible with the duty to testify, and thus a party can invoke a general "right to silence" if called to take the stand. Usually, minimal obligations are imposed on the litigants to disclose evidence or information. "Nobody is expected to supply weapons to his adversary" is the often-invoked maxim.

But this model, so rigorously designed as a contest of two sides before a passive judge, was recommended as a blueprint only for civil cases that were regarded as self-contained private controversies. Because no larger implications were perceived in such lawsuits, judicial passivity seemed appropriate, if not mandated by the requirement of judicial neutrality. Many continental European countries, therefore, enacted codes of civil procedure incorporating features of the recommended theoretical model. The rational implementation of policies toward crime was thought, however, to make the blueprint inapplicable in criminal cases. Though the logic of the party-dominated model might have permitted the prosecutor to represent the public interest in crime control, it was viewed as unacceptable to give the other party—the accused—mastery over defensive issues. If this were done, a substantively erroneous result might be imposed on the passive court. For example, an insane defendant could be convicted if, for some strategic reason of his own, he failed to raise the insanity defense.

European procedural theory thus developed a variety of modified blueprints for the criminal process, some of which were built on the "accusatorial principle" or on the "principle of contradiction" (Damaška, p.560). In their most radical form, they recommended a partial simulation of the party contest, with evidence collected mainly by a nonpartisan but active decision-maker. The facts alleged in the prosecutor's charge constitute the only limit on the court's inquiry.

The Anglo-American legal culture. In Anglo American countries, efforts to formulate organizing principles of procedure are mainly the product of the twentieth century. In civil procedure, for example, continental influences have led to the adoption of the twin principles of party prosecution (that the court will take no step in the case except on motion of a party) and of party presentation (that the scope and content of the controversy are to be defined by the parties). As a shorthand expression of the characteristics of the classical civil lawsuit, the two principles enjoy a certain currency in scholarly discourse.

In criminal procedure, theoretical study has been devoted principally to the discrepancy between the realities of law enforcement and the aspirations expressed in the traditional concept of the adversary system. But there was another factor that contributed to the emergence of theoretical models. A fascination with empirical science led to the desire to compare the efficiency of some features of the adversary system with the inquisitorial alternatives. Most of the empirical studies focused on alternative ways of developing factual and legal material for decision. For the narrow purposes of this research, an adversary "mini-model" was defined as an arrangement where proof and argument are presented to the decision-maker by two partisan advocates, whereas the inquisitorial mini-model was described as a unilateral official inquiry into facts and law. The two models were then used in laboratory experimentation to test their relative efficacy in counteracting the decision-maker's bias, producing reliable results, or attaining some other goal. For example, since in the adversary model the judge is required to listen passively to both sides of the case before making a decision, it was hypothesized that he or she would be less likely to become prematurely biased and draw a conclusion too early (Thibaut and Walker, 1975; Sheppard and Vidmar).

The models reviewed here are all based on the assumption that the goal of the process is the resolution of a conflict. They constitute useful guidelines for reform of procedural systems only insofar as these systems are directed toward the same goal. What then is the relation of theoretically posited goals to reality? Conflict resolution as a goal may be restricted to the contested trial in Anglo-American countries, and even there it may be a secondary or only a superficial aim. If the court refuses to accept the defendant's guilty plea, as it is empowered to do in the majority of common law jurisdictions, the case goes to trial despite the absence of a genuine controversy between the prosecution and the defense.

An archetype of Anglo-American process

In its third sense, the adversary system is a procedural type designed by comparative law scholars to capture characteristic features of the common law process, particularly when contrasted with continental systems. For some of these scholars the adversary type is the common denominator of all Anglo-American procedures, yet this conception is problematic. Consider, for example, the question whether the exclusion of illegally obtained evidence from the prosecution's case at trial represents a defining feature of the adversary type. Because the exclusionary rule has not been adopted by all Anglo-American jurisdictions, but has been adopted in several continental European countries, the answer is no (Hermann, p. 18). Under this approach the precise meaning of the adversary type remains hostage to changes in the law of a single common law country.

Other scholars conceive of the adversary type as an ideal of procedure that is not fully duplicated in any actual system. This second approach can best be exemplified by analogy with styles in art. To classify a work of art as pertaining to a particular style, it is thought sufficient that the work encompass some, though not all, elements of the stylistic ideal. Similarly, certain features can be viewed as typically adversarial, although they are found only in a small number of actual procedural systems. Of course, in order to be useful, the ideal type of the adversary process must provide a structure in which actual systems can be recognized, albeit in exaggerated or stylized form. This second approach is more widespread and will therefore be examined in some detail.

Most scholars describe the ideal type of the adversary process by focusing their attention on the trial stage of the criminal process and on the three-sided relation among the prosecution, the defense, and the court. This triadic relation is significantly different in continental and Anglo-American countries. In the former, the court tends to monopolize the courtroom activity; in the latter, the prosecution and the defense take the largest share of action. As a result, the ideal of the judge as a passive umpire, rather than an active seeker of the truth, is taken as the central ideal of the adversary system (Ploscowe, p. 433). But the focus on triadic relations leaves too much out of account. Both in Europe and in Anglo-American countries, important segments of the criminal process unfold in the absence of the judge and may involve other officials such as the police. Moreover, even if one considers only incourt proceedings, there are often four rather than three actors to consider—crime victims play an increasingly important role. Indeed, many European systems give the victim the rights to be heard as a party and be represented by counsel.

The contrast between Anglo-American and continental criminal procedure is best expressed in two basic notions. The first, underlying the inquisitorial type, regards the criminal process as an official inquiry. The second, underlying the adversary type, regards criminal procedure as a regulated contest between the prosecution and the defense. In discussing other meanings of the adversary system, the image of proceedings as a contest has already been encountered. But the comparative perspective highlights some aspects of this contest that are overlooked by a purely domestic vision.

First, under procedures of the adversary type the prosecution and the defense prepare two independent cases in advance of the trial (often with a view to possibly avoiding trial). Unlike the inquisitorial type, there is no nonpartisan agency preparing a single, or "integrative," case or case file. Problems of maintaining rough equality of the prosecution and the defense can thus arise long before the trial. Pretrial detention, for example, does not fit neatly into the adversary type, because it hampers the defendant in preparing his own independent case. Moreover, the resources and legal powers of investigation of the prosecutor are usually far greater than those of the defense. On the other hand, the exclusionary rule fits in smoothly. If in preparing its case the prosecution breached the law, it should not be permitted to reap advantages from such a "low blow."

Second, various forms of negotiation between the prosecution and the defense are a salient feature of the adversary type. Consider, for example, how easily the practice of plea bargaining fits the "style" of a process based on the notion of contest. It makes little sense to go on with such proceedings if the defendant refuses to oppose the demands of the prosecution. By contrast, in proceedings conceived of as an official inquiry, the defendant need not be asked how he pleads: the trial can go on irrespective of his attitude toward the prosecutorial charges. Inducements to facilitate the task of crime control agencies exist, of course, in both adversary and inquisitorial systems. But the two are characterized by the different loci of such inducements. In the adversary process, both sticks and carrots are used to persuade the defendant not to contest charges, so that the need for trial is obviated. In the inquisitorial process these inducements are used during the interrogation of the defendant: he is urged to reveal information facilitating the task of the officials conducting the inquiry.

So far we have dealt with the conventional position that attributes the same meaning to the words adversary and accusatorial. It has been suggested, however, that comparativists should draw a distinction between the two (Goldstein, p. 1016). Under this approach the adversary process is said to denote only a method of finding facts and deciding legal problems, and is characterized by two sides shaping issues before a relatively neutral judge. The accusatorial system, on the other hand, is a more encompassing concept, which includes the adversary method as its constituent element.

The meaning of this broader concept depends on the contrast with the inquisitorial system, and its non-adversary method of proof and trial. The contrast turns on the divergent attitudes of state officials. In the inquisitorial system, officials are self-propelling and affirmatively obligated to carry out state policies, but in the accusatorial system they step into action only when a controversy arises and they are requested by the participants to respond. Each attitude entails a variety of consequences and choices among procedural forms, the choice of the proper method of finding facts being only one of many. Ultimately, the contrast between the inquisitorial and the accusatorial modes of proceeding involves two polar views about the role of government in society; that is, whether government should be "reactive" or "proactive" (Goldstein, p. 1017).

The idea of linking the characteristics of the Anglo-American criminal process to political ideology is promising. Important features of the Anglo-American criminal process cannot be reduced to the abstract notion of contest, which is so central to the adversary type. Moreover, some features of Anglo-American justice are in conflict with procedures mandated by notions of a fair contest. For example, the right of the defendant to personally defend himself—a right unique to common law—follows from the tenets of the reactive liberal ideology, but it seriously strains notions of a fair contest (Faretta v. California, 422 U.S. 806 (1975)). If more common law characteristics are to be captured in procedural types, broader organizing principles are needed, and the ideology of reactive government provides one such principle. Consequently, it seems sensible to distinguish between the adversary type, which focuses on the contest design, and the accusatorial type, which centers on a political theory.

But even the broader concept of the accusatorial system fails to account for many striking characteristics of the Anglo-American criminal process when the latter is contrasted to continental systems. From the earliest known attempts to describe the peculiar nature of common law justice, the participation of the lay jury was regarded as its hallmark, and lay decision-making as one of its most characteristic elements. The law of evidence, for example, is the product of the interaction of the judge, the jury, and the lawyers (Langbein, p. 306). These features elude the adversary type organized around the notion of a contest; the latter can plainly take place with or without a jury. Nor does the accusatorial system, inspired by the reactive philosophy of government, require jury trials; lay adjudicators can be an arm of a totalitarian as well as of a laissez-faire government. Nevertheless, trial by jury reinforces the characteristic Anglo-American image of the criminal process as a contest of the accused and the state before outside arbiters. Where, as on the Continent, the apparatus of justice is dominated by hierarchically organized civil servants, this conception of the criminal process has little credibility—prosecutors and decision-makers are all too easily traceable to the center of state power. But the contest imagery has far greater plausibility in a procedural system where verdicts are reached by laypersons recruited to serve on the criminal court.

The difficulties involved in expressing the peculiar character of Anglo-American criminal procedure have given rise to increased skepticism as to whether any version of the adversary type can be useful. Those scholars of comparative law who subscribe to the common-denominator approach are clearly justified in their doubts: no single model can be set up to which all Anglo-American criminal procedures conform (Langbein and Weinreb, p. 1551). But even those scholars who are less demanding seem increasingly skeptical. Factors involved in describing the peculiar character of Anglo-American proceedings are too complex and heterogeneous to be captured in a single, internally consistent type of criminal justice. Moreover, as the world's criminal justice systems have become increasingly "hybridized," continental and other non–Anglo-American, "inquisitorial" systems have incorporated many adversary features traditionally seen as defining characteristics of common law systems.

... View less

ACTUS REUS

Actus reus is a term of art in criminal law. Literally the Latin phrase means bad act. The technical, legal use of the phrase denotes one of the elements that must be proven by the prosecution before anyone can be liable to criminal punishment. The actus reus element.... View more

ACTUS REUS

Actus reus is a term of art in criminal law. Literally the Latin phrase means bad act. The technical, legal use of the phrase denotes one of the elements that must be proven by the prosecution before anyone can be liable to criminal punishment. The actus reus element is the act made criminal by some statute or other valid source of criminal law. Thus, a defendant is said to have committed the actus reus of some offense if he has done some act that is an instance of the type of action prohibited by the offense in question. Murder statutes, for example, typically prohibit the "killing of a human being"; the actus reus of murder is satisfied by any act that is an instance of the type of act so described—that is, any act that is a killing of a human being.

Actus reus versus mens rea

There are two contrasts with other elements of criminal liability that help to clarify the nature of actus reus. The first is the contrast with mens rea. Mens rea literally translated from the Latin means guilty mind. The technical legal use of the phrase denotes that prerequisite of criminal liability having to do with the state of mind of the accused when he committed the actus reus of some offense. Thus, one of the mens reas sufficient for murder is general intent: such requirement is often stated as a prohibition on "intentionally killing another human being." The word "intentionally" tells us what kind of mental state an accused must have to be guilty of this kind of murder (either an intent or a belief, as it turns out). The phrase "killing another human being" tells us two things: first, what must be done by way of action to be guilty of murder; and second, what object an accused's intention or belief must take in order to be guilty of murder (Moore, 1993). The first is the actus reus requirement, whereas the second is part of the mens rea requirement. The accused must both actually kill someone, and intend (or believe) that he is killing someone, in order to be guilty of this kind of murder.

The relationship between actus reus and mens rea is not always this close in all offenses. In what are often called specific intent offenses, for example, the object of the prohibited mens rea will not coincide with the act prohibited by law. Thus, the actus reus of common law burglary is the breaking and entering of the dwelling house of another at night, whereas the mens rea includes the requirement that the accused do such breaking and entering with the intent to commit a felony once inside. The commission of such a further felony is no part of the actus reus of burglary, but the intent to commit such a further felony is part of the mens rea of burglary.

In its actus reus/mens rea distinction the criminal law has mirrored a deep divide in morality. This is the divide between wrongdoing and culpability. Although it is disputed, morality is most often thought to contain certain prohibitions and requirements, such as "Do not kill" and "Help others in distress." Morality generally permits us either to do or to refrain from doing most acts, but morality forbids certain actions and requires others. To do an act morality forbids, or to refrain from doing an act morality requires, is to breach one's moral obligations. This is moral wrongdoing.

Morality likewise concerns itself with the culpability with which a wrongful act is done. Overall moral blameworthiness includes culpability as well as wrongdoing. One is free from moral blame for causing a harm to another if one neither intended to cause such a harm, believed one's act could result in such a harm, or unreasonably risked such a harm coming about because of one's actions.

The legal distinction between actus reus and mens rea is best seen as a reflection of this underlying moral distinction. The parallel is one of form, with criminal law and morality dividing criminal liability and moral responsibility (respectively) into these two elements. The difference, of course, lies in the content of legal versus moral norms; in many legal systems much that morality prohibits or requires the law does not, and vice versa.

Actus reus versus justificatory defenses

The second distinction illuminating the nature of actus reus is the distinction between the prima facie case for criminal liability and the defenses. The distinction is a procedural one having to do with allocation of certain burdens in an adversary system. The burdens here pertinent are two: one party or the other is given the burden of producing evidence from which a reasonable fact-finder could find in their favor, and one party or the other is given the risk of not persuading the fact-finder with her evidence. Thus, if the prosecution in a criminal case has both burdens on a certain issue, it will have a verdict directed against it if it fails to produce evidence on that issue and it will also lose if the fact-finder is undecided about which direction the evidence points on a certain issue.

The prima facie case in a criminal trial is that part of the elements of liability on which the prosecution has both of these burdens. Actus reus is best conceived as being as much of the prohibited action as is part of the prima facie case, but no more. Specifically what is excluded by this way of conceptualizing actus reus are the justificatory defenses (Moore, 1993).

Consider homicide again by way of example. Criminal codes do not in fact prohibit simply the "killing of a human being." Rather, they prohibit the killing of a human being except in self-defense, defense of others, prevention of certain crimes, in cases of necessity, and so on. Built into the seemingly simple prohibitions of the criminal law are those exceptional circumstances where the act in question is permitted or encouraged by the law. The actus reus of murder nonetheless remains the exceptionless, simple prohibition against killing a human being, for it is only this much that the prosecution must sustain as part of its prima facie case. The defense has the burden of raising self-defense and the other justificatory defenses, so the absence of any justification of self-defense and the like is not part of the actus reus of murder.

A parallel limitation exists for mens rea. Certain defenses such as infancy, insanity, involuntary intoxication, duress, and provocation are not justificatory defenses; rather they only excuse prima facie illegal conduct. Some have urged an expansive definition of mens rea, so as to include absence of these excusing conditions as part of a "guilty mind." Preferable is the narrower conception of mens rea, paralleling the narrower conception of actus reus (Kadish). On this narrower conception, mens rea is present whenever the accused intends, believes, or unreasonably risks a prohibited action; such mens rea makes for a prima facie liability only, however, since such liability can be escaped by showing excusing circumstances in which the mental state arose.

It is controversial whether this second distinction in the criminal law reflects any underlying moral structure. On its face, the distinction is seemingly based only on the procedural convenience of dividing up the burdens of producing a trial between prosecution and defense in an adversarial system. On this view, the distinctions between actus reus and justification, and between mens rea and excuse, are morally arbitrary. That the actus reus of rape, for example, includes lack of victim consent, whereas the actus reus of criminal assault makes consent of the victim a defense (and thus not part of the actus reus), illustrates this apparent moral arbitrariness.

On the other hand, on some views of ethics morality consists of simple, exceptionless "absolutes" like "Thou shalt not kill." In this view of morality the justifications make actions permissible that are otherwise categorically prohibited. On this view one is morally permitted to kill in self-defense, for example, but it would be better if one did not take advantage of such permission (Moore, 1993). On this "stained permissions" view of the justifications like self-defense, the legal distinction between actus reus and the justificatory defenses reflects the underlying moral distinction between the categorical norms of obligation and the secondary norms of discretionary permissions. In such a view of morality the legal distinction between the actus reus of offenses and the justificatory defenses is not a morally arbitrary matter of procedural convenience.

Whatever may be the case about the moral basis for the two distinctions we have discussed, legally it is clear that actus reus is thus but one of four major elements in criminal liability. It joins mens rea, absence of justification, and absence of excuse as the four prerequisites for liability to punishment in the criminal law, and it joins mens rea in constituting the prima facie case for that liability.

The voluntary act principle

Having isolated actus reus within the overall requirements for criminal liability, it remains to examine its nature. The general nature of the requirement we have stated earlier: actus reus is the requirement that the accused have performed an action prohibited, at least prima facie, by the criminal law. We gain more insight into the nature of this requirement if we probe the nature of actions themselves. If the criminal law requires actions for liability, we would do well to understand what might be generally true of human actions.

This seemingly intuitive route for analysis runs into a long-existing, widely shared skepticism that denies the existence of any general truths about human actions as such. This skepticism admits that we can seek the nature of specific kinds of actions, such as killings, maimings, destroyings, and so on. Denied is that all such types of actions have any shared nature (Austin, 1956; Duff).

If such skepticism were true then the most we could say about the actus reus requirement of the criminal law is what we have said before: actus reus is the requirement that, before one is liable to punishment, one not do one of the many thousands of actions prohibited by the criminal law. Fortunately the metaphysics of human action is not as bleak as this skepticism would contend. There are two very general truths about human actions as such (and thus, about all those many human actions prohibited by Anglo-American criminal codes).

One of those truths is encapsulated within the criminal law's so-called voluntary act principle. The voluntary act principle states that there can be no actus reus (and thus, no criminal liability) unless the defendant performed a voluntary act. A voluntary act, in turn, is defined as a bodily movement caused by the "effort or determination of the actor, either conscious or habitual" (Model Penal Code § 2.01(2)(d)). The voluntary act principle thus requires willed bodily movement by a defendant before criminal liability may attach.

To understand the voluntary act principle, it is helpful to subdivide it into four subprinciples. The first is the idea that voluntary acts are events and are not one of those more enduring things we call states. My firing of a gun yesterday is an event that occurred over a relatively brief interval of time and that involved change in the world. My being a person who likes to fire guns is a more enduring state not involving change but rather stasis.

The U.S. Supreme Court for a time attempted to articulate this distinction in its holdings prohibiting criminal punishment for status rather than action. In Robinson v. California, 370 U.S. 660 (1962), the Court held it unconstitutional for California to punish someone for the status of being an addict, recognizing that it was constitutionally permissible to punish someone for the actions of using drugs. Similarly in Powell v. Texas, 392 U.S. 514 (1968), the Court allowed the punishment of someone for being drunk in public because implicit in the actus reus of the crime was the action of going into public while one was drunk.

The second subprinciple is that voluntary actions are physical events involving the only physical mechanism within our immediate control, our own bodies. While there are mental events like deciding or intending to do something, voluntary acts are not these kinds of events. Rather, a voluntary act is (at least in part) the physical event of our bodies moving in response to our intentions to move them. The insight motivating this second aspect of the voluntary act principle is that the criminal law cares about harms in the world. The only means persons have at their disposal to bring such harms about is by use of their bodies. None of us has telekinetic powers so that only through bodily movement do our evil thoughts produce evil consequences.

The third subprinciple is that only willed bodily movements count as voluntary actions. Our bodies often "act" in the same way that inanimate objects "act," which is to say without our direction or control. If my body is thrown through a window, I cannot be said to have performed the voluntary act of breaking the window; in such cases, my body is no different than a stone that I happen to own breaking the window—in neither case have I broken the window. Similarly, if I am in the midst of an epileptic seizure, a hypogly episode, a reflex or shock reaction, hypnosis, somnambulistic or fugue state, or the like, I am not the author of the harms my body may cause. It is only bodily movements caused by my intention (or "willing") to so move that constitute voluntary actions (Moore, 1993).

Fourth and last, the results of my willed bodily movements are not proper parts of my voluntary actions nor do such results themselves constitute separate voluntary actions. John Austin stated this thesis explicitly: "a voluntary movement of my body. . .is an act. . .bodily movements are the only objects to which the term 'acts' can be applied with perfect precision and propriety" (p. 415). Oliver Wendell Holmes put this point even more succinctly: An action "is a willed muscular contraction, nothing more" (pp. 73–74). Consider the actions of killing someone by way of example. The English language suggests that we cannot kill another without causing the other's death. Are we to infer that the death of the victim, or the causing of it by one's bodily movements, are parts of the voluntary act of killing? Perhaps surprisingly, the answer is no. The only voluntary acts we do are the willed bodily movements by which we kill. What happens after that is no part of our voluntary act, nor is the death resulting a separate voluntary act we do. To paraphrase Holmes and Austin, all we ever do is move our bodies, and the rest is up to nature.

On this view, the causing of death by some bodily movement is a property of that act just as being "the most talked about killing of the decade" can be a property of an act of killing. One way to refer to the act in idiomatic English is by use of these causal properties: "The killing of Nicole," or "the most talked about killing of the decade." Yet the death of Nicole, the causing of it, the talk generated by it, are no part of the voluntary act of her killer. These descriptions are simply ways of referring to that willed bodily movement by use of familiar properties. We do the same thing when we talk of "the teacher of Alexander," referring to Aristotle. It is not an essential part of Aristotle that he taught Alexander, but use of this nonessential but familiar property is a good way to refer to him.

Lawyers and legal theorists often present the voluntary act principle as a distinctively legal principle. They often defend it as a special invention of the law, serving law's unique needs. In fact the four subtheses of the voluntary act principle simply restate some well-worn truths about human actions in metaphysics. Everything lawyers say about voluntary acts many philosophers would say about human actions generically. Human actions—all of them, not just the ones used in criminal codes—are events; they are those physical events known as bodily movements; they are only a subclass of such physical events, namely, only those bodily movements caused by an intention to so move; and the only actions there are are willed bodily movements, however much we refer to those actions via their causal properties (Davidson; Moore, 1993). The voluntary act principle should be seen for what it is, an analysis of the nature of human action as such. So seen, it is one-half of the story of what it is the actus reus principle of criminal law requires: to be an action at all—and thus, an action prohibited by the criminal code—there must be a willed bodily movement.

Common criticisms of the voluntary act principle

Critics of the classic analysis of actus reus are legion. Such critics attack all four aspects of the voluntary act principle, sometimes construing it as a creature of legal doctrine and other times taking it to be a general metaphysical truth about human actions. First, it is urged, there are criminal prohibitions of states and not only of events. In Samuel Butler's fictional Erewhon one could be punished for having tuberculosis, but even in Anglo-American criminal codes one can be punished for vagrancy, possession of various items (drugs, firearms, burglary tools, etc.), being in a vehicle where marijuana is smoked, and so on.

It is sometimes said that crimes of status are compatible with the voluntary act principle because acts may "consist of a state of affairs, rather than an event" (Gross, p. 60). This, however, is to obliterate the voluntary act requirement. Voluntary acts are essentially events and if crimes of status truly exist they contradict the voluntary act principle. Better is the response of the late Herbert Packer, who noted that crimes of status "are in fact very much on the way out" (Packer, p. 78). Not only are such laws rarely enacted today, but in America a number of constitutional infirmities are regularly found to afflict such laws so that even where they do remain on the books they are not valid (Robinson v. California).

An exception to this last observation must be made for possession crimes, which are both numerous and constitutionally valid. Such crimes seemingly prohibit the state of possessing something (weapons, drugs, etc.) and thus seem to be incompatible with the voluntary act principle. Yet possession has become a term of art in Anglo-American criminal law. Although in ordinary English and in the law of property one might easily be said to possess something simply by being in the state of having it on one's person, criminal law requires more. "Possession" is defined in criminal law so that either an act of acquiring possession or an omission to rid oneself of possession are prerequisites to liability (Williams, 1961; Model Penal Code, § 2.01(4)). With "possession" so defined, possession crimes present no counter-examples to the voluntary act principle, or at least none greater than that presented by omissions generally (which we shall shortly discuss).

A second basis for denying that the voluntary act principle is part of the actus reus requirement stems from the supposed existence of criminal actions without any bodily movements on the part of the "actor." Sometimes this objection is cast as an observation about Anglo-American criminal law: certain crimes punish culpable mental acts alone without any execution into bodily movement. More often this objection is cast as a metaphysical observation about action: some actions can be done without any bodily movement.

The legal branch of the objection would be cogent if Anglo-American criminal law still punished thoughts alone, as in the ancient form of treason constituted by the mere "compassing the death of the king" (Williams, 1961; Fletcher, 1978). Yet modern statutes require execution of the most evil thought in bodily movement. This is true not only of treason, but also of attempt, solicitation, and conspiracy as well. Unlike the Romans, we have no crimes consisting only of mental events like dreaming of the death of the emperor (Scholz).

The metaphysical branch of the objection is more complicated. The objection is that one can literally do actions like killing without lifting a finger (Fletcher, 1995; Corrado; Annas; Brand). There are three sorts of examples here: (1) the actor ("A") pushes the victim ("V") into the water, and then stands still while V drowns; (2) A is attached to a device that will kill V if, but only if, A can stand on his head motionless for ten minutes, which A does, causing V's death; (3) A is driving when suddenly his old enemy, V, darts in front of his car, and A rather than swerving, remains motionless while his car runs over V, killing him.

In fact none of these sorts of cases present examples where A has killed V without a willed bodily movement by A. About (1), A's pushing V into the water is a voluntary act that caused V's death so that A did kill V, but not without moving his body. About (2), A again did kill V, but he again moved his body to do so. The trick is to see that A's activating the muscles needed to remain motionless are bodily movements too. For difficult routines where the "agent's body is about to be made to move by outside forces," to keep one's exterior body from moving by activating the appropriate muscles is to engage in willed bodily movement in the sense intended by the voluntary act principle (Vermazen, p. 95; see also Holmes; Moore, 1993, 1994). About (3), A does not kill V with his car. A will doubtlessly be liable for V's death, but not because A killed V; rather, A omitted to save V when A was duty-bound to do so because A's earlier acts of driving put V in danger (Moore, 1994). None of these examples disprove the voluntary act principle by producing instances of "motionless killings."

A third objection to the voluntary act principle stems from that principle's reliance on willings to mark the line between voluntary and involuntary bodily movements. The objection is that there are many voluntary actions where there is no datable mental state of willing. While this objection once had many adherents in both law (Hart) and philosophy (Ryle), more recent analyses have sustained the need for some state like willing, volition, endeavoring, intent to move, and so forth, to mark off voluntary action from mere involuntary movement (Moore, 1993, 1994; Bratman).

A fourth objection to the voluntary act principle disagrees with that principle's fourth thesis. Such an objection denies that the death I cause by shooting another is no proper part of my act. On this view my killing, my shooting, my pulling of the trigger, and my moving my trigger finger, are each distinct particular acts I did, not just four different descriptions of one act I did. On this view my act of killing is distinct from my act of moving my trigger finger, even though I did the former by doing the latter. The objection concludes that acts like killing others do not have at their core willed bodily movements or anything else. A killing is a killing, a burning is a burning, but they need share no feature universal to all actions, as is asserted to be the case by the voluntary act principle.

While there is a surprising amount to be said in favor of this objection (Goldman, 1970, 1994), common sense supports the voluntary act principle. When I move my trigger finger, when I move it slowly, when I move it smoothly, when I pull the trigger by moving it, and so on, I am doing one act, not as many acts as there are descriptions of it (Moore, 1993). Such a chain of descriptions of but a single act leaves open the possibility asserted to be true by the voluntary act principle: all actions are essentially willed bodily movements.

We have thus far deferred any discussion of omissions because they present the most serious objection to the view that the actus reus of all offenses includes a voluntary act. The objection also is a complicated one because those who voice it do not even agree what omissions are. The best conceptualization of omissions is that they are simply absent actions. An omission by actor A to save V from drowning is just the absence of any act by A of saving V. Such omission is not a ghostly act of saving or of anything else; rather, it is the absence of any such type of act. Such omissions are thus a kind of action no more than nonexistent elephants are a kind of elephant (Moore, 1993).

The voluntary act principle states that all actions are in essence willed bodily movements. An omission to save V at some time t thus might consist in A not moving his body at t. Yet motionless omitters are rare. Usually one who omits to save is busy doing something else at t—dancing a jig, buying a dishwasher, and so forth. What makes such persons omitters to save at time t is that none of their willed bodily movements at t has the causal property, saving-of-V's-life. One thus does not want to picture omitters as motionless statues because they need not be such (and they typically are not such).

Once we are clear as to what omissions are, we can see that Anglo-American law undeniably criminalizes some omissions. If we are the parent of a child who needs rescue, if we have undertaken such rescue even if we are not related to the child, if we have either innocently or culpably caused the condition of peril to the child, or if some statute specifically imposes a duty on us to rescue the child, we are under a positive legal duty to prevent the child's death. Despite numerous efforts to reconcile this liability with the voluntary act principle (Hughes; Gross; Epstein; Mack), the simple truth is that they are not reconcilable (Moore, 1993). Insofar as Anglo-American law criminalizes true omissions, it creates an exception to the principle that a willed bodily movement constitutes the essence of the actus reus of all criminal offenses. The voluntary act principle remains of great importance, however, because omission liability is rare in Anglo-American law and thus almost all the time it remains true that the actus reus requirement can be satisfied only by a willed bodily movement.

For the exceptional cases of omission liability we do need an account paralleling the voluntary act principle's account for act liability. If the essence of criminal omissions is not willed bodily movements, what might it be? Some have suggested that the essence of omissions is also to be found in willing. The analysis is that omissions are the willed absence of bodily movements (Fletcher, 1994). In this way one keeps as close a parallel to the voluntary act principle as possible. Yet willed absences of bodily movements is too narrow an analysis of omissions generically and it is even too narrow as an analysis of omissions made criminal by Anglo-American codes. If I negligently do not notice the child in distress, I negligently omit to save her. This is an omission, and if I am the child's parent, a criminal omission, yet I at no point willed the nonmovement of my limbs to refrain from saving her (Bentham).

The preferable line to take here is to see that the omissions we criminalize all have as their common element a capacity of the omitter not to have omitted. If I am to be held criminally liable for omitting to save my child, I must at a minimum have had the capacity to move my limbs in the relevant way—I was not asleep, in the middle of an epileptic seizure, under hypnosis, paralyzed, and so on, at the relevant times. Then I can be said to have voluntarily omitted to rescue the child.

This completes one-half of the analysis of actus reus in the criminal law. At a minimum, to satisfy the actus reus requirement of some offense one must satisfy the voluntary act principle (or in exceptional cases, the voluntary omission principle). We now need to see what else must be true in order to satisfy the actus reus requirement for criminal liability.

The properties common to complex types of actions

If criminal codes only prohibited actions like moving one's finger, then the actus reus requirement would be exhausted by the voluntary act principle. Yet for obvious reasons no criminal code consists exclusively (or even in part) of such prohibitions. We are morally and legally indifferent to such simple actions so no one has any reason to criminalize them. Rather, we criminalize more complex actions like killing another, destroying property, raping, maiming, and stealing. What else is true about all of these types of actions (beyond the fact that all are in essence willed bodily movements)?

What we seek here are useful generalizations about properties possessed by the thousands of actions prohibited by our criminal codes. It has been traditional to group all such properties into only two types: causal properties and noncausal properties of actions (Bentham; Austin; Williams; Model Penal Code § 1.13(9)). Killings of a human being, for example, are willed bodily movements having the causing of death of a human being as a property. The death is then said to be the result element of the actus reus of homicide. Death is so described because death of a human being must be the result of any willed bodily movement that is a killing of (i.e., a causing the death of) a human being. Killings of a police officer while in the performance of her official duties, by contrast, are willed bodily movements having not only the causal property of all killings but also having the noncausal properties that the person killed was a police officer and was on duty at the time. The facts that the victim was a police officer and that the victim was on duty at the time of the killing are then said to be the circumstance elements of the actus reus of cop-killing. These facts are described as "circumstances" because they are not caused by the defendant's willed bodily movement; they are simply facts ("circumstances") present at the time the defendant acted.

The criminal law's division of all properties of actions into these two kinds is uniquely legal. There is no corresponding division of the properties actions may possess in either philosophy or in ordinary thought. Philosophers of action often distinguish the properties actions may possess quite differently. Such philosophers often speak of causal properties, as do criminal lawyers, but noncausal properties are often divided up into conventional properties, mental properties, properties of the agent, properties of the victim, properties having to do with the manner, means, or instrumentality used, and so on (Rescher; Goldman; Bennet; Thalberg).

It is thus important to be clear why the criminal law is categorizing the properties actions may possess in order to assess the adequacy of its analysis. Perhaps surprisingly, the criminal law has little actus reus—oriented purpose in classifying the properties possessed by those actions criminal law prohibits. For we can determine whether a defendant satisfies the actus reus requirement for any crime without classifying the properties of action; we only need ascertain whether the act of the accused has the various properties each crime requires. Thus, the justification for classifying the properties of actions lies elsewhere, in the need of the criminal law to draw certain mens rea distinctions. These mens rea distinctions are between one who intends to cause a certain harm, one who knows to a practical certainty that his action will cause that harm, one who knows that his action will substantially and unjustifiably risk that harm, and one who unreasonably risks causing that harm even though he is unaware of that risk. These distinctions are used by the criminal law to grade the culpability with which a given wrongful act is done. The unaware but unreasonable risker is least culpable, and the intender is most culpable, with the knowing and reckless causers graded between these two extremes.

Such a grading scheme for culpable mens reas seemingly demands that the criminal law classify all properties of prohibited actions into causal or noncausal properties. The idea is that the grading scheme above described only makes moral sense with respect to the causal properties, but not the noncausal properties, of actions. Consider the crime of assault with intent to kill a police officer performing his official duties. Such a crime requires the most seriously culpable of the mental states, namely, an intent to kill; mere belief to a practical certainty that one's actions will result in death will not satisfy the mens rea requirement of this offense. Thus, a defendant who sets off a bomb against a prison wall in order to help some prisoners escape (while knowing that the guard next to the wall will be killed by the explosion) does not have the intent to kill; whereas another defendant who sets the bomb in order to kill the guard (in order that the guard cannot later identify the defendant) has the required intent to kill.

With regard to the causal property, causing-death-of-a-human-being, use of the intent/knowledge distinction seems to work well enough. The defendant who intends to kill is somewhat more culpable than the defendant who does not so intend but who only knows that his action will result in someone's death. But now imagine two more defendants, each of whom assault an on-duty police officer with the intent to kill him. The first of this pair of defendants knows that his intended victim is a police officer and knows that he is on duty; however, his reason for wanting to kill the officer has nothing to do with these facts, for this defendant hates the cop for personal reasons. By contrast the second defendant cares whether his intended victim is a police officer and whether that victim is on duty. We may suppose that this second defendant is engaged in a cop-killing contest between lifers in prison where there is no death penalty, and one "scores" in the contest only if one kills an on-duty policeman.

Defendant two is moved to kill the person he assaults by the fact that that person is an on-duty cop; defendant one is indifferent to these facts, although he knows that they exist. Both the common law and the Model Penal Code deny there to be any significant difference in culpability between these last two defendants. If one takes this view, then we do not want to distinguish between them when we grade culpability by mental states. We should thus lump those who literally intend to kill an on-duty cop with those who intend to kill a person who happens to be an on-duty cop (and who they know to be such), treating both as guilty of the most culpable grade of mental state.

We can define this most serious grade of culpability differently only if we can divide all criminal actions into two different aspects. This is where the causal versus noncausal property distinction is needed. If the property in question is causal, then the most serious grade of culpability requires intent as its mental state; if the property in question is noncausal, then the most serious grade of culpability allows belief to a practical certainty to suffice along with intent.

Other distinctions between the mental states that grade culpability are also thought to demand this distinction between causal and noncausal properties of action (Moore, 1993). Rather than pursue these, however, we should turn to three criticisms commonly made of this classification scheme.

Criticisms of the circumstance/consequence distinction

One is a moral criticism. The argument is that there is some difference in the culpability of the last pair of would-be cop-killers, and, indeed, as much difference as there is between the first pair of prison bombers. If this is so, then our reason for categorizing all properties of actions into two large clumps disappears.

It is hard not to have considerable sympathy for this moral criticism. Nonetheless, perhaps an enriched diet of examples can tip one back toward the orthodox criminal law categorization of actions. Consider this one. Two defendants each commit an assault with intent to have intercourse with a female who they know is not consenting. Defendant one is indifferent to the fact, seeking intercourse whether the victim consents or not; defendant two only likes nonconsensual sex, so that if the victim consented he would cease his assault. If both defendants are sufficiently close in culpability as to be lumped together in the most serious grade of culpability, then the criminal law may well be correct to draw its culpability distinctions differently for the noncausal property of consent than for the causal property of penetration.

It is also not quite true that the only reason for the criminal law to draw the causal/noncausal property distinction is in order to grade culpability systematically. If these kinds of properties differ in the universality with which they apply to criminal actions, that is a fact worth marking in systematizing criminal law. Although it is controversial—as we shall explore momentarily—all actions prohibited by Anglo-American criminal code have causal properties, while this is not true of noncausal properties. This is a fact worth marking, possible only if one distinguishes the two kinds of properties.

The second criticism of the causal/noncausal property classification is that it is incomplete. The argument is that certain actions are not divisible into their causal or their noncausal properties; rather such actions are said to have a nature that is neither. Such crimes are often termed "conduct" crimes, of which theft, rape, attempt, breaking and entering, and driving under the influence of alcohol are supposed to be examples.

This is a difficult criticism to get a handle on, since it seems so obviously false. Still, the criticism is a very popular one among criminal law theorists (Williams, 1983; Buxton; Fletcher, 1978), and it even infects the Model Penal Code when that Code (inconsistently) inserts "nature" of an action as an element in addition to "results" and "circumstances" (MPC § 2.02; see Moore, 1993). So the criticism must at least be taken seriously enough to be explained away as a mistake.

None of the supposed examples of "conduct crimes" turn out to require an analysis different than the orthodox analysis of action in criminal law. Take breaking and entering, for example. The actus reus of breaking and entering is breaking and entering a building not your own. That the building entered is owned by someone else is a noncausal property of the action required—a "circumstance element," in the language of the Model Penal Code (MPC § 1.13(9)). A breaking occurs when a willed bodily movement causes a window to be broken, and an entering occurs when a willed bodily movement takes place in the circumstance that an outward threshold of a building is crossed. There is no need for a nature to breaking and entering, since the actus reus of that offense is fully analyzable in the orthodox way.

The same analysis is adequate for the other supposed examples of "conduct crimes." The actus reus of rape is satisfied when a willed bodily movement causes penetration in the circumstance where there is a lack of victim consent. The actus reus of theft is satisfied when a willed bodily movement causes an item to move in the circumstances that the item in question is owned by another who has not consented to its taking. The actus reus of attempted murder is satisfied when a willed bodily movement causes a state of near success in killing to exist, and so on.

There are two apparent reasons explaining the persistence of this "conduct crimes" criticism despite its manifest falsity. One is due to the directness of the causal links between willed bodily movements and results in conduct crimes. Usually the causal chain between certain willed bodily movements and penetration in rape, for example, is very short. The shortness of the chain leads some to think that there is no causal link here at all. Yet a short causal chain is still a causal chain. In addition, once in a great while the chain is not so short, as when the defendant inserts the penis of another into the female (Commonwealth v. Dusenberry). Such cases make plain what was true all along: There is a causal property built into the actus reus of rape and other "conduct crimes," so that no sui generis "nature" of rape needs to be added into the analysis of that actus reus.

The second reason explaining the confusion about conduct crimes lies in certain linguistic facts. When we say, "The actus reus of theft is the moving of a chattel and such action of moving involves the causing of movement by that chattel," it may seem that the causal analysis is bogus. It sounds like saying, "The action of moving something involves moving that thing," which is trivial. Unnoticed is that the English language uses "moving" and "movement" in two quite different ways. "The moving of the chattel" refers to an action, whereas "the movement of the chattel" refers to a different event, the event of the chattel moving. The latter event could be caused by an action of moving, or it might not. It is thus a significant assertion to say that "the willed bodily movement caused the movement of the chattel" and even to say, "The moving of the chattel caused the movement of the chattel." The causing of movement of a chattel is thus a causal property of moving a chattel, and no sui generis idea of "nature" need be added to analyze this action.

The third criticism of the orthodox division is that the causal/noncausal property distinction is wholly indeterminate. The potential indeterminancy appears when we consider how much to include as part of what is caused by an actor's willed bodily movements. Consider the actus reus of killing an on-duty policeman. The standard analysis is that the causing of a death of a human being is a causal property any act must possess to be an instance of this prohibited act-type, and that the victim of such killing is a police officer, and on duty, are noncausal properties any act must possess to be an instance of the prohibited act-type. Yet, why isn't causing the death of something the causal property, and the circumstance that the killed thing is a person or noncausal property? Alternatively, why isn't the causing of death of an on-duty policeman a causal property, with no noncausal property? Without some control on how we individuate properties—a notoriously tricky business (Armstrong)—it would seem that the orthodox classification scheme can be manipulated at will.

There is no dearth of suggestions as to how to deal with this problem. Some have suggested a temporal criterion: if the fact exists at the time of the willed bodily movement, then it is a noncausal property of that act. Others have urged a conventional criterion: what is "customarily regarded" as part of what is caused forms part of a causal property (Buxton, p. 31). Still others have urged a moral criterion: ask whether the intent/knowledge distinction marks any significant difference in culpability vis-à-vis the property in question, and if it does not, call the property noncausal (Moore, 1993). One might even urge a metaphysical criterion: include just so much of the state of affairs prohibited in the causal property as corresponds to true causal laws (see Armstrong). None of these suggested responses, however, has proved adequate to the objection.

Despite this unanswered criticism, the second general truth about actus reus retains wide acceptance: in addition to a willed bodily movement, the actus reus of all offenses includes the consequences of that movement and the circumstance in which that movement took place. These consequences and circumstances constitute the causal and the noncausal properties, respectively, that any willed bodily movement must possess if it is to satisfy the actus reus requirement of the criminal law.

... View less

ACCOMPLICES

Accomplice liability rests on the premise that someone whom the law interchangeably calls an accomplice, accessory, aider and abettor, secondary party, or helper in the crime or crimes of his perpetrator, doer, or principal is derivatively liable for whatever crime or.... View more

ACCOMPLICES

Accomplice liability rests on the premise that someone whom the law interchangeably calls an accomplice, accessory, aider and abettor, secondary party, or helper in the crime or crimes of his perpetrator, doer, or principal is derivatively liable for whatever crime or crimes the principal commits. Punishment for accomplice liability is shared equally among principals and their helpers. Proof of the helper's liability is heavily mediated by the actions of the principal. If the principal commits a crime, the equal blame goes to the helper as well, provided that the crime that occurs is one the helper knew about and whose success the helper intended when he provided his assistance.

Accomplice liability's legitimacy rests on its demand that the helper's contribution be significant enough to justify his punishment, but not so significant, dominant, or manipulative as to wipe out altogether the responsibility of the principal. Someone who helps or tries to help someone else commit a crime exerts somewhere from no, to some, to too much constraint on his principal's autonomy. Too much influence exerted by the accomplice does not produce a case of accomplice liability; rather, it produces a case of principal liability for the overreaching helper in his agent's (or would-be principal's) "innocent" wrongdoing. No influence, or perhaps more accurately, no attempt to influence or support the principal, does not produce a case of accomplice liability because the helper has not done enough to make him sufficiently caught-up or "causally" related to the principal's crime. Neither is there a case of accomplice liability if the helper and the principal do not put themselves to the same task, either because the helper does not really care whether the principal succeeds in or even attempts a crime, or because the principal commits a crime or crimes that depart from, or are in excess of, the parties' common scheme. Cases falling in between those cases where the helper does either too much or too little are what one could call "pure" or "core" cases of accomplice liability in which the helper: (1) exerts some (but not too much) influence on the principal; (2) intends that the principal succeed in the jointly intended criminal act; and (3) the principal does in fact at least generally perform as the helper expects him to.

Principal liability: too much influence exerted by the helper

Cases of principal liability on the part of a would-be helper arise when the would-be helper acts in a way that allows us to say that it is as though the helper commits the crime himself. Certainly one can perform an action by getting others to do it. We say, for example, "Louis XIV built Versailles," even though the actual construction was not done by him. Indeed, we can think of cases where the principal is not a principal at all, but is simply, perhaps metaphorically, a tool, instrument, or means of someone else. Examples of such cases include cases where someone occupying what would otherwise be the position of the helper recruits a lunatic or a child to do the deed or tricks, forces, or even hypnotizes someone occupying what would otherwise be the position of the principal. These cases involve such coercion or manipulation of susceptible parties that the manipulated or coerced party's act is fishy enough for him to be called "not responsible" or for his act to be judged "not his own." Thus courts tend to reject the notion that providing a gun to a lunatic (the gunprovider being unaware of the lunatic's incapacity) to use to assault someone somehow makes the assault the gun-provider's and not the gunwielding lunatic's. For one person's act to be wholly someone else's, the person to whom we attribute the act must act in a way that shows he sees his act as such; one cannot, after all, use someone else inadvertently. Were, for example, a ring-leader to pay a safecracker to steal some jewels from a vault for a share of the profits, it is not as though the ringleader sees himself cracking the safe and stealing the jewels—he sees the safe-cracker doing it. The only evidence of his seeing himself doing it would be his placing such constraints on the safecracker's autonomy that it ceases to be the safecracker's intentional, purposeful, or deliberate act. Thus if the ringleader were to force the safecracker to crack the safe by putting a gun to his head or were he to trick the safecracker into believing that the safe and its contents really are the property of the ringleader, then the ringleader steals the jewels through the safecracker. In such a case, the ringleader would be the principal thief and not a helper at all, and the safecracker, who is seemingly the principal thief, would not be held responsible for his actions; instead, he would be viewed as an innocent means or instrument of the manipulative ringleader.

It is likewise an instance of principal as opposed to accomplice liability where A hands B a package into which A has secretly put a bomb for delivery to a victim A has in mind, or where A places B under duress by threatening B with a greater harm if B does not act on A's behalf than if B does. There A acts through B by seeing B not as a killer, but as an innocent dupe—a giant fuse, if you will. A sees himself killing the victim by manipulating or forcing B into doing A's dirty work for him. A harder case to classify is one in which a malicious felon places an innocent person or a police officer in circumstances where it is the innocent's right or the officer's duty to apply deadly force to repel the felon's threat of force, and the innocent or officer kills someone other than the malicious felon. In such cases the felon does not act through the innocent or officer because missing is the malicious felon's intention to use the killer. The felon's intention is likely that no such encounter materialize, except in so-called shield cases (where a third party is used by escaping suspects or those under siege as a shield against police gunfire), or in cases where one felon sends an innocent or confederate outside to a certain death in order to facilitate the malicious felon's escape. With such a bad intention and excessive risk at play, it is easy to see how in those cases we may conclude that the felon acts through the killer to deflect the justified use of deadly force away from the felon and toward another target.

When we are faced with questions of whether a would-be helper has manipulated the would-be principal to the point that the would-be principal's responsibility is wiped out altogether, the would-be helper/manipulator's conception of his own liability does not inhibit his conviction as principal. This is because the idea of "innocent agency" or "perpetration by means" is linked only to those cases where the principal intends to pursue an objective through the manipulative use of an agent. In other words, perpetrating harm through another is a narrower category of action than is causing another to do something harmful. Causing another to do something harmful, unlike perpetrating harm through another, is indifferent to whether the originating actor (whom we are considering treating as a manipulator) intends to reduce someone else to his influence or control. In other words, causing can be mechanical whereas using cannot. Accordingly, the ringleader who recruits an insane safe-cracker—not knowing of the safecracker's affliction—may in some important sense cause the ensuing theft, but does not commit the theft through the insane thief. Missing there is the ringleader's intent to use, manipulate, or otherwise act through the safecracker. When, however, the harmful act is orchestrated by a user or manipulator who is counting on the agent's susceptibility, incapacity, or lack of responsibility, the idea of innocent agency or perpetration-bymeans describes cases where the manipulated agent is a lunatic, a child, someone duped as to material facts, or anyone who cannot choose what is good and right due to coercion or any other constraint on the innocent agent that is known to the dominant party. In such cases it makes no difference whether the harm is committed by lying, stealing, frightening, shooting, stabbing, or nonconsensual intercourse (as in an infamous British case in which a husband misled an intoxicated man into thinking the husband's wife wanted intercourse with the intoxicated stranger) (R. v. Cogan & Leak, 1976 Q.B. (Eng. C.A.)). Each of these cases instantiates principal, not accomplice, liability on the part of the dominant party.

Core cases of accomplice liability

What must the helper do to be an accomplice to the principal's crime? In what way is it as though helpers who do not coerce or manipulate their principals commit their principal's crimes? The helper has merely helped. But helping, say, burglary, is not committing burglary. Nor is helping burglary trying to commit burglary, any more than argue is equivalent to "try to convince" or warn is equivalent to "try to alarm." Anglo-American law nevertheless treats a helper as a principal so long as the helper intentionally contributes to the principal's crime. In other words, if the helper pitches in the requisite contribution to the principal's crime, then the helper is punished identically to the principal since Anglo-American law long ago "abrogated" or statutorily eliminated the historically recognized distinction between the amount of punishment one deserves for, say, giving a burglar a crow bar and actually performing the breaking and entering oneself (Standefer v. United States, 447 U.S. 10 (1980) pp. 15–20; Smith). Whether a defendant is (1) a perpetrator (or "principal in the first degree") who actually performs the criminal act; (2) a perpetrator (or "principal in the second degree") who provides some assistance at the scene of the crime; or (3) a helper or an accomplice (or "accessory before the fact") whose aid is given in advance (as in the planning stages) or contemporaneous with but away from the scene makes no difference for purposes of punishment. Indeed, only a helper (or "accessory after the fact") whose assistance comes after the crime has taken place—typically in the form of concealment of the crime—receives a more lenient punishment than those criminals who make their contribution before or during the crime.

While this position of "equivalency" of punishment between those who commit crimes and those who help them is uncontroversial in Anglo-American law, there is some controversy over what sorts of help or contribution count as acts of accomplice liability. We know already that too much influence—too great a contribution—is not a case of accomplice liability at all, but rather a case of principal liability on the part of the would-be helper. So too is it possible that the would-be helper has not done enough for his contribution to count as an instance of accomplice liability. Although there are various formulations of how much the helper must add to the criminal venture, it is clear that encouragement of any sort, whether it be in the form of soliciting or asking the principal or another helper to commit or participate in a crime, cheering on the principal (or another helper), or merely promising the principal (or another helper) to help if necessary is enough to keep the helper on the hook for the principal's actions.

When accomplice liability is based on encouragement, the doctrine operates identically to the doctrine of conspiracy. The doctrine of conspiracy holds that parties who agree to commit a crime are jointly liable not only for the agreement, but also for the carrying out of the conspiratorial objectives (Pinkerton v. United States, 238 U.S. 640 (1946)). An agreement can be formal or explicit as well as implicit or arrived at through "nods and winks." Merely imitating another's behavior, however, falls short of an agreement, even if the imitation is mutual. Thus the Supreme Court has held that a conspiracy to fix prices cannot be proved in the absence of some communicated intention to pursue a specified pricing strategy. Simply pursuing such a strategy—even aware that it is strongly in other competitors' interests to do the same—is "consciously parallel" behavior, but does not make out a case of conspiracy (Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939)). Many cases of accomplice liability, however, are also cases of conspiracy. Those that are not are cases that originate not with encouragement between parties, but when material aid such as a car, a crowbar, a gun, or poison are supplied, or when the helper serves as a getaway driver or lookout. In these instances—where there is no communication of the helper's intent to help—there is no conspiracy between the principal and the helper to commit a crime. Accordingly, the helper's shared responsibility for aiding the principal's crimes in these instances is based on the law of accomplice liability, as opposed to the laws both of accomplice liability and conspiracy.

When it is aid and not encouragement that the helper adds, it must be "actual" aid that "mattered" or "made a difference" to the principal's actions (Kadish, pp. 358–359). Only when the helper's actions could not have been successful in any case is there no accomplice liability. Thus the question "how much contribution?" comes down simply to whether the helper meant to contribute to the principal's crime, and any action that betrays such an intention will be deemed sufficient to constitute an act of accomplice liability. Thus it has been held that lending a man a smock to keep a battery victim's blood from staining the batterer's suit made enough difference to the batterer to justify our treating the smock-lender as a batterer (judgment of 10 May 1883, 8 RGSt 267; cited in Fletcher, pp. 677–678). So too might an angry judge's interception of a telegram have mattered in a murder, since if the victim had received the telegram, he could have anticipated the gunman behind him on reading the crucial wire: "Four men on horseback with guns following. Look out" (State ex rel. Attorney General v. Tally, 102 Ala. 25, 69, 15 So. 722, 734 (1894)). Even a door opened for a burglar could conceivably make a difference to burglary through the window.

The above are only exceptional examples; but even basic cases of accomplice liability, such as where a helper lends his principal a crowbar for a burglary or drives him to the sites of the crime, are not cases where the helper actually causes the crime, even if the principal has no crow bar of his own or cannot drive a car. It follows that an otherwise superfluous helper whose opening a bank door hastens a robbery by seconds is on the hook as accomplice to the robbery, as flimsy as his contribution may be. Thus the real issue here is not so much whether the helper has "caused" or even "made a difference" to the principal's crime, but rather, whether the helper has put himself to helping, or has tried to help.

What must the helper know about the principal's intentions? It is not enough that the helper encourage or aid the principal in the principal's crime or crimes. In order to be an accomplice in the principal's crime(s), the helper must in addition know what it is that is being helped, and must intend that the acts of encouragement or aid facilitate the principal's criminal venture. In other words, there must be some level of attunement between the parties before blame for the principal's actions can be shared equally by the helper. So, if the helper lends a crow bar to his neighbor unaware that the neighbor intends to use it for a burglary, then the helper is not an accomplice to the burglary, even though he has contributed material aid. Because the aid was provided in ignorance of what use it would be put to, the aid cannot, by itself, make out a case of accomplice liability against the helper. This is precisely the problem the law faces with providers of goods (e.g., retail hunting stores) and services (e.g., lessors of hotel rooms) who naively help along a criminal venture. A helper's contribution simply cannot count as accomplice liability unless the helper knows what the principal is up to.

Even when the helper does know what the principal is up to, there are two considerable problems to our finding the helper to be an accomplice in what the principal ultimately does: (1) the helper may be indifferent to how or even whether the helper's contribution operates on the principal; and (2) the principal may depart "upward" from the common scheme by committing excessive crimes not imagined by the helper, or depart "downward" from the common scheme by raising a complete or partial excuse from liability that may (or may not) be open to the helper.

The helper's level of commitment to the principal's criminal venture

Often the principal and helper have divergent levels of commitment to the contemplated crime. For example, the lessor of a hotel room that the lessor knows the lessee will use for gambling, drug distribution, or prostitution may well know to what unlawful use the premises will be put. But the lessor still may not, in the words of the celebrated Judge Learned Hand, have a sufficiently "purposive attitude towards" the venture so as to be "associated" with it in a way that demonstrates "that he wishes to bring [it] about" (United States v. Peoni, 100 F.2d 401 (2d Cir. 1938). Instead, the lessor may even prefer to lease the room to a law-abiding lessee (that way the lessor can avoid trouble) but is willing to rent to anyone who can pay the going rate. In such a case, the lessor's contribution to the crime counts as aid, but unless there is reason for us to believe that the lessor is somehow in on the scheme, the lessor's aid falls short of the "purposive attitude" toward the principal's crimes that the law of accomplice liability requires. No doubt a case of accomplice liability is made out if the lessor takes a commission from the lessee's venture or charges the lessee extra to insure against the risk the lessor incurs by leasing to a criminal who may for obvious reasons be bad for the lessor's business.

This means that a helper cannot be accidentally liable for or caught up in the principal's crimes, even when the crime that the principal has committed is one for which accident is not necessarily an excuse. For example, a (principal) driver can be convicted of drunk driving even if it is perfectly reasonable for him to believe he has had too little to drink to have become drunk. To be accomplice to the driver's drunk driving, however, the helper must do more than merely fail to take cost-justified precautions against doing or saying anything that may make the principal's drunk driving more likely. Instead, the helper must mean to facilitate drunk driving; he must, in Judge Hand's terms, demonstrate a purposive attitude toward bringing about the crime in question. So a bartender or social host may in fact be assisting drunk driving by keeping an inaccurate tally of how much their guest or customer has had to drink, they may know that such a result is likely, and nonetheless not have within the letter of the law helped drunk driving—not if such an outcome is not the bartender or social host's intention.

The principal's departures from the common scheme

When the principal departs "upward": the helper's liability for the principal's excess. Assuming that there is at least a crime toward which the helper has the requisite intention or "purposive attitude," then what happens if the principal commits other crimes in addition to or instead of the crime or crimes that the helper means to help? In evaluating the helper's responsibility for the principal's excessive criminality, the law is understandably unsympathetic to the claims of the too-finicky helper, who complains at trial that the principal deviated, however slightly, from the common scheme. Therefore if the helper asks the principal to take the victim's gold watch by snatching it from the victim's wrist, but the principal obtains it by threatening to expose the victim's criminal record to his golf club, a court would not let the helper off the hook for being an accomplice to the principal's blackmail. The criminal objective of stealing the watch remains the same—the principal's deviation only in means fails to demonstrate a lack of attunement between the parties. If, however, the helper lends a crowbar to the principal, believing that the principal intends to enter a house in order to steal a television, the helper will be off the hook for playing a role in, say, arson, if that is what the principal unexpectedly does instead upon entering the house. So long as there is attunement as to "essential matters" or so long as the crime in question is of the same "type" or within the "contemplation" of the range of crimes anticipated by the helper, the principal's departures will not save the helper from liability for what the principal ultimately does (Northern Ireland v. Maxwell, (1978) 3 All E.R. 1140; Regina v. Bainbridge, (1960) 1 Q.B. 129). Determining just what it was that was contemplated demands thorough knowledge of the enterprise, a matter that is made easier when there is a conspiracy: the more formal the better.

When the principal exceeds the helper's expectations, still a minority of courts have expanded the liability of the helper for the principal's excesses (People v. Luparello, 231 Cal. Rptr. 832 (Ct. App. 1987)). It is safe to say that decisions which hold helpers on the hook for their roles in, for example, intentional murder by a principal when the helper asked the principal to commit assault or at most kidnapping, reveal the most extended or outermost limits of a helper's liability. Yet the overwhelming majority of courts regularly do make this stretch when principals commit intentional or accidental killings—even unforeseeably or contrary to a carefully thoughtout plan—during the course of certain "inherently dangerous felonies," such as burglary, robbery, arson, rape, kidnapping, and prison escape (Model Penal Code § 210.2(1)(b)). As a result, a getaway driver who means to aid in robbery is liable for the "murder" of a store clerk who dies of a heart attack when confronted by the armed principal robber. Outside of this area of homicide, known as "felony murder," helpers are typically held liable only for their principals' actions that are within, or at least adjacent to, their common scheme.

When the principal departs "downward": The helper's relation to the principal's excuses. So far we have been focusing entirely on helpers' excuses that have to do with what the helper knows, intends, and does. But often the principal will have an excuse that will allow him to avoid altogether, or at a minimum reduce his responsibility for, what he has done. In such cases, courts have developed strategies for establishing the connection, if any, between the principal's full or partial excuses and the helper's liability.

Take for example a version of Shakespeare's Othello, in which Iago calmly and maliciously drives Othello into a blind rage and incites him to kill his wife Desdemona by making Othello believe (falsely) that Desdemona had been unfaithful to him. Let's assume that Othello's rage would make him eligible for the partial excuse of "provocation" or "extreme emotional disturbance," which precludes a murder conviction and instead makes his crime more accurately described as the less-grave offense of manslaughter. The excuse is only partial because Othello still deserves some punishment; it is not as though Iago acted (completely) through him. Othello was in a rage, but still knew what he was doing—knew that he was retaliating against his wife. While Othello is still partially responsible for what he has done, his rage reduces "down" to manslaughter an otherwise murderous act. Consequently Othello's punishment will be five or so years rather than the life imprisonment he would have been eligible for had he been thinking more clearly at the time of the killing.

But what is Iago's relation to Othello's partial excuse? There are four options open to us here, each of which is expressed in at least some Anglo-American court opinions: (1) allow Othello's rage to benefit Iago on the ground that a helper's liability cannot exceed whatever crime actually takes place. This view is appealing to the extent that it enforces the notion that a helper's liability is derivative of the principal's—the helper cannot help a crime that does not take place, whatever the reason may be that it fails. This view is unappealing, however, to the extent that it lets the helper borrow (perhaps unfairly) defenses such as intoxication, mistake, insanity, and duress, which may be utterly personal to the principal; (2) allow Othello's rage to benefit Iago only if Othello was "justified" in part in killing Desdemona as opposed to merely "excused." This has the appeal of letting the helper exploit a principal's actions that, at least in part, "interfere with the rights of no one." Unfortunately, to give the helper this benefit not only requires taking a position on which defenses are in fact justified as opposed to excused, but also threatens the unwelcome result of allowing a helper to claim he was justified in doing something when the very facts that make the act justified are unknown to him (as in battering someone whom unbeknownst to the batterer the principal had the right to repel in self-defense); (3) deny Iago the benefit of Othello's rage by "grafting" Othello's action onto Iago's intention in order to make out a case of murder on Iago's part and manslaughter on Othello's. While this way each party would have to raise his own defenses, the problems are considerable in that this would permit a defendant to be convicted as accomplice to a murder when no murder took place; and (4) conclude that Iago has attempted murder while Othello has committed manslaughter. This position (Model Penal Code § 5.01 (3)) recognizes that the helper's help has in an important sense failed or misfired (thus the reduced liability) and features the additional benefit of reflecting that accomplice liability is more about what the helper puts himself to than what he actually accomplishes. Whichever of these four strategies a jurisdiction adopts for dealing with problems of a helper's liability for the actions of a principal who may have a full or partial defense will determine whether the helper is punished even though the principal is not, whether the helper can borrow a principal's excuses (or his justifications), or a position somewhere in between.

What crimes can be helped?

Although the abrogation of the distinction between principals and helpers has equalized their punishments, the abrogation will never be able to eliminate altogether the distinction between helping and doing when it comes to identifying whether certain actors are liable at all for whatever it is that has happened. For example, suppose an antiprostitution law that criminalizes "selling sex." Obviously the prostitute is the seller, but what has the "John" or buyer of sex done? Sold sex? Helped the prostitute sell sex? Nothing criminal at all?

This problematical aspect of the law of accomplice liability comes up frequently in the context of two-party cases requiring the participation of two persons as opposed to the run-of-the-mill offense that requires a perpetrator and a victim. Dangerous games such as Russian roulette or drag racing are examples of such two-or-more-party offenses where the law of accomplices has an uncertain role to play. It is not all that unusual for courts to say that lucky survivors of dangerous games have somehow killed the unlucky players who have died from shooting themselves in the head or by driving their cars off the road into ditches or oncoming traffic.

For example, in People v. Abbott & Moon, 84 A.D.2d 11 (N.Y. App. Div. 1981), Moon was drag racing with Abbott, who killed Patricia Hammond and her two passengers, who had entered the intersection through which Moon was racing at 80–85 and Abbott at over 90 miles per hour at the time of the wreck. Although Moon was driving worse than unsafely, he was lucky enough to avoid ramming into anyone. While Abbott's liability for the three deaths was obvious, Moon's conviction of criminally negligent homicide and reckless driving also was upheld on appeal. The court explained that

[w]hile Moon did not personally control Abbott's vehicle which struck the three victims, it could reasonably be found that he "intentionally" aided Abbott in the unlawful use of the vehicle by participating in a high-speed race, weaving in and out of traffic, and thus shared Abbott's culpability. . . . Moon associated himself with the high-speed race on a busy highway and took part in it for nearly two minutes over a distance in excess of one mile. Actually his conduct made the race possible. He accepted Abbott's challenge and shared in the venture. Without Moon's aid Abbott could not have engaged in the high-speed race which culminated in the tragedy. (p. 15)

For this reading of complicity the New York appellate court cited criminal-law expert Wayne LaFave, who has noted that such a view "has much to recommend it" (LaFave and Scott, p. 673).

Although calling Moon an accomplice in the fatalities that arose out of his excessive risk-taking has an elemental appeal (he was, after all, a wrongdoer), it is analytically impossible. Consider again the passage quoted above in which the court observed: "Actually his [Moon's] conduct made the race possible." Indeed it did, and this is precisely why each racer is analytically precluded from helping the race. Help can be withheld, or it wouldn't be helping at all. In other words, because the relation of helping (unlike doing or perpetrating) to the ultimate harm is synthetic or empirical, not analytic or true by definition, the actions of helping and doing are distinct and should be so treated. Thus if the crime analytically, elementally, or definitionally requires two or more parties, then the required parties cannot, merely by participating, possibly "help" an activity to which they are by definition essential. Certainly a buyer does not help a seller in an exchange transaction by paying for goods any more than an unmarried person helps a bigamist by marrying him or her, a betrothed couple help each other get married by marrying, or someone helps someone else kiss by simply kissing them.

Here we are not talking simply about cases of "joint principality," under which two parties divide the elements of an offense; for example, two parties rob when one commits the assault (the frightening of the person) and the other the larceny (the taking of the property). Since both the force or threat of force and the taking of the property are analytically, elementally, or definitionally necessary to any robbery, neither party is helping robbery; both are committing it. Oppositely, where the help of one party is necessary only as an empirical or synthetic matter—that is, where the helper does not fulfill a statutory definition of crime or one of its elements, but his actions happen to be necessary for the crime on these facts, then he is helping and not doing, no matter how he may characterize his own actions. For example, that a getaway driver may be necessary for a successful robbery must be observed to be known; getaway drivers are not analytically or definitionally necessary to robbery, which occurs whether or not the perpetrators have a car. Consequently, getaway drivers are helpers, not principals or joint principals, regardless of how they may characterize their actions.

Despite considerable confusion in court opinions—see Commonwealth v. Atencio, 189 N.E.2d 627, 630 (Sup. Jud. Ct. of Mass. 1963)—and academic commentary—see Fletcher, pp. 654–655—multiparty game cases, like exchange transactions, do not instantiate helping by one whose participation is analytically a necessary condition of the crime itself. This is not to say that drag racing and Russian roulette foreclose altogether the doctrine of accomplice liability. Spectators cheering on a drag race could be liable for helping the homicide. Well-known (and still controversial) decisions like Wilcox v. Jeffrey, 1 All E.R. 464 (King's Bench Division 1951), (where a magazine writer, for the purpose of writing about the performance, "helped" Coleman Hawkins play jazz illegally in the United Kingdom) have so held. Cheering spectators are helping drag racing (as Natalie Wood so enthusiastically did in Nicholas Ray's Rebel Without a Cause) and thus are liable as accomplices in the unlucky racer's demise. But a lucky drag racer who avoids disaster—who neither bumps, cuts off, nor swerves into another racer, driver, or pedestrian—"helps" nothing.

Although American law insists on treating helpers and doers identically, the cheering spectators should have an excuse, albeit a partial one: they were merely helping. Not only is the lucky survivor helping nothing, but neither is he jointly principal in the killing, given that the crime with which such defendants are customarily charged—manslaughter—has two elements: (1) excessive risk-taking and (2) causing death. Manslaughter is not, analytically, a two-or-more-party offense; nor is it divided into one (one steers, the other accelerates?) as obscene phone calling could be were one person to dial and the other to speak obscenely. Moon was charged with manslaughter, not with drag racing. To use the necessary participation as a means of describing the role as that of helping the unlucky player's actions papers over the grammatical, even moral, distinction between helping and doing.

... View less