Actus reus — conduct, causation, omissions

Actus reus is the external element of an offence: the conduct, circumstances, consequences, or omission which the law identifies as part of the prohibited wrong. In the first week of Cambridge Part IA Criminal Law, it is tempting to translate actus reus as the guilty act and then move quickly to mens rea. That is too crude. Many offences do not require an act in the narrow sense; some require a state of affairs, some require a prohibited result, and some are committed by failing to act when the law imposes a duty. Equally, many offences require not merely that something happened, but that it happened because of the defendant’s conduct in a legally relevant sense.

Three ideas structure this topic. First, criminal law ordinarily punishes conduct rather than mere thoughts, character, or status. This reflects both liberal restraint and evidential necessity: the criminal process should not condemn inward states unless they are manifested in outward conduct. Secondly, where an offence is result-based, the prosecution must prove causation. It must normally show that the defendant was a factual cause of the result and that legal responsibility for the result should be attributed to him. Thirdly, omissions are exceptional. English criminal law generally imposes no duty to rescue strangers; liability for failing to act arises only where a recognised duty exists, such as a statutory duty, a contractual duty, a duty arising from a relationship, an assumption of responsibility, or the creation of a dangerous situation.

The topic is foundational because it recurs throughout the paper. In homicide, causation explains when an assault becomes manslaughter or murder. In non-fatal offences, it determines when injuries are attributed to the accused. In criminal damage, the actus reus includes destruction or damage to property belonging to another. In theft, the conduct element is appropriation, but it occurs within a wider set of circumstances and mental elements. In offences of strict liability, the actus reus may carry much of the normative work because mens rea is reduced or absent.

For Cambridge purposes, this week should be used to build a method rather than to memorise isolated rules. In a supervision or Tripos answer, begin by identifying the offence structure: conduct, circumstances, consequence, and any omission. Then ask whether the defendant’s bodily movement was voluntary; whether any consequence was factually caused by the defendant; whether any intervening event breaks legal causation; and, if liability is based on non-action, whether a recognised legal duty exists. The best answers do not recite the but-for test and stop. They explain why causation and omission doctrines are not purely scientific inquiries but normative devices for allocating criminal responsibility.

The modern language of actus reus and mens rea is a relatively tidy vocabulary imposed on a less tidy historical development. Early common law criminal liability grew from forms of wrongdoing rather than from a general analytical structure. The old felony forms were concerned with concrete harms: killing, burning, wounding, stealing, breaking and entering. Over time, lawyers began to distinguish the external element of the wrong from the fault element accompanying it. The Latin terminology suggests conceptual elegance, but the underlying law remains highly offence-specific.

Historically, criminal law was suspicious of liability for omissions. The common law developed in a social and political culture which treated criminal punishment as an especially coercive form of state power. To impose a general duty to rescue would have required courts to say when one person becomes responsible for another’s welfare. That is a legislative and moral choice of considerable breadth. English law therefore preferred specific duties. Parents must care for children; those who assume care may incur responsibility; those who create a danger must take reasonable steps to avert it; drivers must stop and report accidents. The result is a patchwork rather than a code.

Causation also developed pragmatically. The criminal courts did not adopt a unified metaphysical theory of causation. They asked whether the defendant’s contribution was sufficiently connected with the result to justify liability. The but-for test offered a simple starting point, especially in cases such as White, but it could not decide harder cases. If a victim receives poor medical treatment, acts in panic, refuses treatment on religious grounds, or injects drugs supplied by the defendant, the issue is not merely whether the defendant was a historical condition of the death. It is whether the later event should be treated as interrupting the chain of legal responsibility.

The twentieth century saw increasingly explicit judicial formulation of principles. Smith and Cheshire addressed medical treatment after violence. Pagett dealt with responsive acts by third parties. Miller articulated a duty arising from creating a dangerous situation. Stone and Dobinson recognised assumption of responsibility in an informal domestic setting. Kennedy reaffirmed the significance of the victim’s free, deliberate and informed act in causation. These cases are not just factual illustrations; they reveal the law’s attempt to mark boundaries between personal autonomy, public protection, and fair attribution of responsibility.

The historical lesson is that actus reus is not a purely mechanical threshold. It encodes assumptions about the proper reach of criminal law. The insistence on voluntariness protects against punishment for reflex, accident, or automatism. The reluctance to impose liability for omissions protects individual liberty and avoids uncertain duties of altruism. The causation doctrines prevent liability from turning into unlimited responsibility for every consequence to which the defendant made some historical contribution. In Part IA, this background matters because examiners reward answers which see doctrine as structured by principle, not merely by case-name accumulation.

1. The actus reus is offence-specific. There is no free-standing crime of causing harm or behaving badly. Every offence must be analysed according to its own definition. The actus reus may consist of conduct, circumstances, consequences, or a combination of these. In criminal damage, the result is destruction or damage; the property must belong to another; and the defendant’s conduct must cause that damage. In theft, the conduct is appropriation and the circumstances include property belonging to another. In murder, the result is death and causation is central. A good answer first disaggregates the offence rather than asking in the abstract whether there was an actus reus.

1. Conduct must ordinarily be voluntary. Criminal liability is generally premised on a voluntary act or omission. A bodily movement produced by a reflex, spasm, or complete loss of control is not ordinarily the defendant’s act for criminal law purposes. Voluntariness is conceptually prior to mens rea: if there is no legally attributable conduct, the question of fault may not arise. Do not confuse this with the defence of automatism, which has its own doctrinal rules. At the actus reus stage, the simple point is that criminal law is concerned with agency, not mere bodily occurrence.

1. Result crimes require causation. Where the offence requires a consequence, the prosecution must prove that the defendant caused it. Causation is usually divided into factual and legal causation. Factual causation is commonly expressed through the but-for test: but for the defendant’s conduct, would the prohibited result have occurred when and as it did? If the answer is yes, factual causation fails. White is the classic illustration: the defendant poisoned his mother, but she died of an unrelated heart attack before the poison operated. He was not guilty of murder because the poison did not cause death.

The leading cases fall into three groups: factual causation, legal causation, and omissions.

White is the indispensable starting point for factual causation. The defendant attempted to poison his mother, but she died of natural causes before the poison took effect. The case shows that wicked intent and dangerous conduct cannot substitute for proof that the prohibited result was caused. For murder, the prosecution must prove that the defendant caused death; otherwise liability may lie for an attempt, but not the completed offence. White is therefore a discipline against result-based liability by moral intuition.

Smith and Cheshire concern medical treatment following injury. Smith held that the original wound remained an operating and substantial cause despite poor medical treatment. Cheshire went further in insulating defendants from arguments based on negligent treatment. The later medical act will only rarely break the chain. The underlying policy is clear: one who inflicts a serious injury cannot ordinarily avoid responsibility because doctors then treat the victim imperfectly. Yet the doctrine must preserve a limit. If treatment is wholly independent and overwhelmingly responsible for death, the defendant’s contribution may cease to be legally significant.

Pagett concerns responsive intervention. The defendant used a pregnant woman as a human shield and fired at armed police; police returned fire and killed her. The police response did not break causation because it was a reasonable act done in self-defence and in response to the danger created by the defendant. The case is best understood as a responsibility principle. When the defendant creates a situation in which others predictably respond to danger, those responses are normally attributed back to him unless they are free, deliberate, informed, and unreasonable in a way that makes them independent.

Kennedy marks a limit. The defendant prepared a syringe of heroin; the victim self-injected and died. The House of Lords held that the victim’s voluntary and informed self-administration broke the chain. The case is central because it prevents causation from becoming a device for circumventing the autonomy of the immediate actor. Where an adult of sound mind freely chooses to inject himself, the supplier does not cause the administration in the relevant legal sense, though he may be guilty of drug offences.

Stone and Dobinson and Gibbins and Proctor establish omission duties based on responsibility for vulnerable persons. In Stone and Dobinson, the defendants took in Stone’s sister, who became helpless, and failed to obtain medical assistance. Their limited capacities did not prevent liability: they had assumed responsibility. In Gibbins and Proctor, a father and his partner failed to feed a child. The duty arose from parental responsibility and household control. These cases show that omission liability is not confined to formal legal relationships; it may arise from the reality of dependency and undertaking.

Miller is the modern authority on creation of danger. The defendant accidentally started a fire while sleeping, woke, realised what had happened, and moved to another room without taking steps to extinguish it or summon help. The House of Lords treated the subsequent failure as capable of satisfying the actus reus. Miller is doctrinally elegant because it avoids pretending that the initial accidental act was culpable. Liability arises when the defendant becomes aware of the danger he has created and fails to take reasonable steps. It is a bridge between conduct and omission, and a frequent source of examination problems.

Bland is different but conceptually important. It concerned withdrawal of life-sustaining treatment from a patient in a persistent vegetative state. The House of Lords characterised withdrawal as an omission rather than a positive act and held that continuation of treatment was not required where it was not in the patient’s best interests. The case demonstrates that the act-omission distinction is morally and legally contested, especially in medical contexts. It is not merely a matter of physical description; it depends on how the law characterises professional duties.

See practice questions section below.

See further reading section below.