Moral and Metaphysical Foundations for the Criminal Law of Collective Action

Criminal law is grounded in, and limited by, conceptions of moral responsibility. While there are substantial disputes whether one should understand criminal responsibility as simply that area of moral responsibility enforced by the state, virtually all criminal law theorists agree that the legit­imacy of state-imposed liability is threatened if people are routinely punished for acts for which they bear no proportionate moral responsibility.¹ The starting point for a philosophical inquiry into complicity law should be a sketch of its moral foundations.

But the really significant harms and wrongs of life—the looming incineration of our planet, the genocides, the systems of social hierarchy, etc.—are not the products of us acting alone. They reflect collective choices, intersecting plans, organized groups, and deliberately composed systems of incentives and reward. Likewise, the great achievements of freedom and equality: though led sometimes by individuals, they are products of social movements, alliances of parties, shared deliberation by voters, and implemented by agencies, armies, and bureaucracies.

In causal terms, history’s greatest heroes and villains are collectives not individuals. To be sure, Hitler, Stalin, and Tamerlane attract biographies, and their individual decisions and acts were crucial to history’s unrolling.

The morality of collective action—or, better put, the moral valence of collective acts—is of course a central topic of moral, social, and political life. And it has attracted a significant body of important philosophical inquiry. Such inquiry has typically taken three forms (although some­times combined): first, an analytical account of collective agency, which is meant to generate support for claims about praise and blame, or responsibility more generally, for the individuals involved in these acts; second, an account of the ways moral responsibility—including future- directed, action-guiding claims of duty—attach to groups (collectives) as such; and third, an account grounded in politics and normative sociology, about how the collectives construct a social ontology that legitimates persisting forms of inequality and subordination.

Put aside metaphysical speculations as to whether collections of agents can be said, non- reductively, to act. Indulge in the legal fictions of corporate acts and acts of state. It is indisputable that whenever a collective acts (or fails to act), individuals have acted (or failed to act). A crowd surges against a barrier—only because each person has stepped forward and pushed against those ahead. A country elects a president—only because millions have cast ballots, others have tallied those ballots, others have certified their results, and many more have accepted as legit­imate those results. How do we assess individual responsibility in such cases? Each individual has acted—but no individual’s act seems to have made a difference to the outcome. However, if many had not acted as they did, the harm would have been averted. This is the moral problem of parallel, marginal individual action, involving accumulations of similar individual acts with tiny or invisible effects. A proper analysis of the moral significance of these acts must look to, and beyond, questions of individual causal significance to broader concepts of “making a difference,” threshold-crossing, and what Derek Parfit famously called “moral mathematics.”²

Another difficult category of individual responsibility lies in the domain of non-parallel but individually significant acts, when the moral value of one act is affected by another’s distinct act. One such case might be the furnishing of a means to another’s bad end, when one person sells a crowbar to another planning to use it in a burglary; or a doctor provides medical aid to someone being tortured by an intelligence service, knowing that the victim will be tortured again if he survives.

The term “complicity” finds its technical, legal sense in such cases, and also its core moral sense, when the moral liability of one person depends on the acts of another. But the term may be more broadly used for the broader category of parallelism. As a first-order matter of morality, time presents two different types of questions that individuals might pose for themselves in such cases, or others might pose to them or about them: ex ante, what should individuals do—act or refrain, given the significance of their acts when summed or otherwise aggregated with those of others? Second, ex post, if harm has been done (or good, for that matter), what responses are appropriate, whether of blame and sanction, or praise and reward? Both questions hinge on the question of how the moral valence of one’s own act is affected by its inclusion in a broader class of similar acts.

Some cases present primarily causal/metaphysical puzzles rather than moral ones. When the intent of the peripheral parallel actor, or aider or abettor, is clear and maleficent, a char­acter judgment is easily assessed; and practices of shunning or reprobation have secure conse- quentialist anchors. Intent, as a matter of moral theory and social practice, drives a judgment of retrospective blame or prospective avoidance. Cases in which the individual act is surplus, or otherwise irrelevant to the scale of the harm, can matter for questions of the scale of moral responsibility or legal liability, whether to punishment or compensation. Think of one vote to elect a demagogue, when the margin of victory is far greater than one vote (and can be predicted to be so). Or a plant manager letting out a small amount of toxic discharge, aware that it is a common practice, and that the amount from one discharge has no determinable effect.

But descriptive metaphysics is less the issue than normative significance. We might agree that the cause of the house’s burning down was both the faulty circuit and the failure of the fire department to turn up in time to dowse the flames—even though the metaphysically par­simonious would not recognize the omission as a genuine cause. While fire investigators and tort lawyers might further disagree about which is to count as the cause, that disagreement lies in policy, not philosophy. The issues of overdetermination and inter-agent causation presented by complicity are similar. The central issue of causation lies in the attribution of relative degrees of significance, more than in the resolution of the puzzles of overdetermination. Philosophers can solve the metaphysical problem by articulating conceptions of causation—for instance, the “NESS” test,⁵ according to which each contributor to an overdetermined event can count nonetheless as a part cause of the event.

One way to address normative significance is to make use of the idea of being responsible for an attempt—that is, to treat an agent as liable so long as the agent’s act is of a type that generally causally contributes to the realization of a wrong (even if here it did not). If no typical excusing conditions are in place, we can treat the act as sufficiently anti-social in motivation, and risky in other iterations, that some form of sanction is warranted (or praise in the beneficent cases). It can also be the basis for a warranted judgment of character, and reveals the nature of the relations between the peripheral agent and those towards whom his ill-will or indifference is expressed.

The hardest puzzles are generated on a strictly Kantian account, because a peripheral agent could pass the relevant Kantian moral filters. Kantians require us to act in ways that accord with rules that all can defensibly follow. In principle, each match thrower can say, “I only throw matches when I know it won’t make a difference, because there are a sufficient number of other match-throwers so that my act is superfluous.” But even here, a proper understanding of the agent’s intentions—which necessarily include coordinating around others’ acts, if not identi­fying with them to accomplish a joint goal—makes it appropriate to assess the act in relation to that collective outcome. And so moral responsibility is warranted even in these metaphysically troublesome cases. We may summarize the case as follows: philosophical argument can support ordinary moral practices; the harms brought about collectively by several persons can be a basis for the warranted recrimination of a single agent, independent of any individually significant causal contribution. As a general matter, such recrimination is grounded in the intentional orientation of the agent—how she understands her actions in relation to what others do.

There are a number of different philosophical approaches, largely grounded in the phil­osophy of action, which provide further structure to this claim: that intentional or knowing contribution to an event produced through multiple individual acts is a sufficient basis for warranted response. Accounts by myself, Margaret Gilbert, Holly Lawford-Smith, Seumas Miller, and Julia Nefsky are among the contributions to this literature; many others are discussed, criticized, and defended elsewhere in this Handbook.⁶ Some of these philosophers, following the lead of Michael Bratman,⁷ build cooperation out of dense networks of individual intentions, coupled with reciprocal conditions (“I’ll do this, if you’ll do that”). Others, among them not­ably Margaret Gilbert, understand cooperation as built up not out of ordinary, first personal intentions (“I’ll do this, if you’ll do that”), but as reflecting a more radically social form of agency (“I’ll do this as part of our doing that”). My view falls in the latter camp: I understand collective agency as composed of individuals orienting their agency around a goal or activity they understand as irreducibly joint. In the focal case, this is a matter of individuals having “par­ticipatory intentions” to do their part in a collective act.

Much of this scholarship raises questions that go beyond the issues within accomplice liability, such as the difference between bare coordination and full cooperation; the ethical obligations that arise from (and may be constitutive of) cooperation; the ontology of social groups; and the reducibility of participatory intentions to ordinary individual intentions. But there is important carry-over, and potential cross-fertilization, between law and philosophy in this area. The first site of convergence concerns the basic structure of the accomplice’s intent, whose distinctive character is illuminated by the philosophical concepts ofjoint action and intentionality. But the real value may lie in the contributions of philosophy of action to two other standard questions within complicity law: the alienated accomplice and the knowing facilitator.

26.3