Actus reus revisited — causation and omissions

The actus reus of an offence comprises the prohibited external elements: conduct, circumstances, and consequences. This note revisits two foundational but analytically distinct problems: causation (when a defendant's act is legally responsible for a prohibited result) and omissions (when liability may attach to a failure to act).

Causation matters only for result crimes—murder, manslaughter, criminal damage, and others whose definitions require proof that the defendant caused a specified outcome. The inquiry divides into factual causation (but-for causation) and legal causation (ascribing the result to the defendant as a matter of normative judgment). Legal causation incorporates tests of substantial or operative cause, the rules on intervening acts (novus actus interveniens), and the so-called thin-skull rule.

Omissions liability raises a different question: when does the common law or statute impose a duty to act such that a failure to act—itself yielding a prohibited result—suffices for the actus reus? English law adheres to a no general duty to rescue; liability for omission turns on specific categories of duty. The orthodox account identifies duties arising from statute, contract, relationship, voluntary assumption of care, creation of danger, and public office. These categories are not closed, and their scope remains contested.

Both topics rest on normative foundations: causation tests police how far back the chain of responsibility runs; omissions doctrine delineates the boundary between acts and failures to act, and between moral blameworthiness and legal obligation. They sit at the intersection of legal doctrine, moral philosophy, and policy, and candidates must demonstrate command of black-letter rules, awareness of conceptual difficulty, and facility with the case law.

This note assumes familiarity with the basic structure of criminal liability but explores causation and omissions at the depth expected in FHS tutorials and examinations.

The modern law of causation and omissions is substantially judge-made, layered over centuries of incremental development. The but-for test has Victorian roots: in R v White [1910] 2 KB 124, the defendant put cyanide in his mother's drink intending to kill her, but she died of a heart attack before the poison could act. The Court of Criminal Appeal held that the defendant's act was not a factual cause of death because she would have died when and as she did even without the poisoning. White established that factual causation is a necessary but not sufficient condition.

The legal causation inquiry—requiring that the defendant's act be a substantial and operative cause—emerged through twentieth-century appellate decisions. R v Smith [1959] 2 QB 35 (Courts-Martial Appeal Court) held that where the defendant stabbed a fellow soldier and the victim died following poor medical treatment at a field hospital, the original wound remained an operative cause provided it was still an operating and substantial cause at the time of death. This set the template: medical negligence or delay seldom breaks the chain of causation.

The thin-skull rule—you take your victim as you find them—was settled in R v Blaue [1975] 1 WLR 1411 (CA). The victim, a Jehovah's Witness, refused a blood transfusion and died of her stab wounds. The Court of Appeal confirmed that religious or physical peculiarities do not break the chain.

Intervening acts were analysed in R v Pagett (1983) 76 Cr App R 279 (CA): the defendant used his pregnant girlfriend as a shield during a shootout with police; she was killed by police fire. Goff LJ held that the defendant's conduct remained an operative cause because the police response was a reasonable and foreseeable reaction. The voluntariness and foreseeability of the intervening act became the touchstones.

The law of omissions has older roots. Manslaughter by gross negligence has long permitted liability for failure to act where a duty exists: R v Instan [1893] 1 QB 450 (CCR) imposed liability on a niece who failed to care for her dependent aunt. The modern categories of duty were formalised by the mid-twentieth century, and R v Miller [1983] 2 AC 161 (HL) extended duty by holding that a person who inadvertently creates a dangerous situation owes a duty to take reasonable steps to avert the risk—a significant expansion.

Both causation and omissions doctrine reflect pragmatic judicial tailoring: the rules seek to avoid extending liability too widely (especially where the victim's own conduct or third-party interventions are involved) while capturing paradigmatic culpability. They continue to evolve, especially in the context of medical cases and vulnerable-victim scenarios.

Causation in fact: the but-for test

The but-for test asks: would the prohibited result have occurred but for the defendant's conduct? If the result would have occurred anyway, in the same manner and at the same time, the defendant's act is not a factual cause (R v White). This is a necessary condition but not sufficient to ground liability. The test works well for single-defendant, single-act cases but falters where there are multiple sufficient causes (see doctrinal debates below).

Causation in law: operative and substantial cause

Legal causation requires that the defendant's act be:

1. Operative: it must still be exerting an influence at the time of the prohibited consequence. An act that is entirely spent or superseded will not be an operative cause.
2. Substantial: it must be more than de minimis. The term 'substantial' means 'more than minimal', not necessarily 'the main' or 'sole' cause: there can be more than one operative and substantial cause (R v Benge (1865) 4 F & F 504).

In R v Cheshire [1991] 1 WLR 844 (CA), the defendant shot the victim, whose death resulted from complications arising from a tracheotomy performed by doctors. The Court of Appeal held that the gunshot wounds need only be an operating and substantial cause even if poor medical treatment contributed; only if the medical negligence were 'so independent of [the defendant's] acts, and in itself so potent in causing death', would the chain be broken. This formulation sets a high threshold for medical intervention to sever causation.

Intervening acts (novus actus interveniens)

An intervening act may break the chain of causation if it is:

• Free, voluntary, and informed: A fully voluntary act by the victim or a third party may relieve the defendant of responsibility, provided it was not reasonably foreseeable or a natural reaction to the defendant's conduct (R v Pagett).
• So independent and potent: Medical negligence, third-party acts, or the victim's own conduct must rise to this threshold. The test is conjunctive: independence and potency.

The courts distinguish:

The leading authorities on causation and omissions are best understood as a narrative progression, tracing the refinement of principle through appellate reasoning.

White (1910) established the necessity of factual causation. The defendant's conviction for murder was quashed because the cyanide did not cause death; the heart attack would have occurred regardless. The case stands for the proposition that intended harm does not ground liability for a result crime unless the actus reus is proved. It is the foundation of the but-for test and features in every causation problem.

Smith (1959), decided by the Courts-Martial Appeal Court, set the modern standard for legal causation in the face of medical negligence. The victim was stabbed in a barrack-room fight and dropped twice en route to the medical centre; treatment was substandard. Lord Parker CJ held that if the original wound was still 'an operating cause and a substantial cause' at death, it mattered not that other causes also contributed. The case is authority for the principle that subsequent negligence rarely breaks the chain.

Blaue (1975) settled the thin-skull rule for all characteristics. The victim refused a life-saving transfusion on religious grounds. Lawton LJ, delivering the judgment of the Court of Appeal, held that the defendant must take the victim as found, 'including her religious beliefs'. The refusal was not a novus actus; it was part of the victim's makeup. Blaue is central to problem questions involving victim peculiarities.

Pagett (1983) clarified intervening acts. The defendant fired at armed police while holding his girlfriend as a human shield; police returned fire, killing the girl. The Court of Appeal, per Goff LJ, applied the substantial-cause test and held that the police response was a reasonable act performed for self-preservation and the preservation of others—hence not a voluntary novus actus. The case developed the doctrine that a reasonable and foreseeable reaction does not break the chain, and it refined the novus actus framework into foreseeability and voluntariness.

Miller (1983), a House of Lords decision, is the leading authority on duties arising from creation of danger. The squatter fell asleep smoking, awoke to smouldering bedding, moved to another room, and did nothing. The House held he was under a duty to take reasonable steps to counteract the danger he had created. Lord Diplock's speech confirmed that a person who by his own act (even if inadvertent or non-culpable initially) creates a dangerous situation thereafter owes a duty. Miller fundamentally expanded omissions liability beyond static relationships and is frequently tested.

Cheshire (1991) refined Smith in the context of medical complications. The defendant shot the victim; the victim died from respiratory complications of a tracheotomy. Beldam LJ emphasized that the defendant's acts need not be the sole or even main cause, and that only treatment 'so independent … and in itself so potent' would break the chain. Medical negligence short of gross incompetence will not exculpate. Cheshire is ubiquitous in essays and problems on medical intervention.

Kennedy (No 2) (2007) addressed the limits of causation for drug-supply manslaughter. The defendant prepared a syringe of heroin and handed it to the victim, who injected himself and died. The House of Lords held (Lord Bingham delivering the leading speech) that the victim's self-injection was a free, voluntary, and informed act that broke the chain of causation. The supplier did not cause the death; the act of administration was the victim's alone. Kennedy is critical for omissions and causation in drug cases, and signals judicial reluctance to extend liability where the victim is an autonomous adult.

See practice questions below.

See further reading below.