Intoxication and mistake

This final substantive week draws together the architecture built across the year: actus reus, mens rea, homicide, offences against the person, property offences, sexual offences and the general defences. Intoxication and mistake are not merely discrete topics. They are pressure-points in the law’s account of culpability. A defendant who is mistaken may lack the mental element required by the offence, or may act under circumstances which, if true, would provide a defence. A defendant who is intoxicated may say either that the prosecution cannot prove mens rea, or that the facts appeared to him in a way which would otherwise excuse or justify his conduct. The law’s response is sharply differentiated. Honest mistake usually matters because criminal liability depends on what D intended, knew, believed or foresaw. Voluntary intoxication, by contrast, is often treated as culpable in itself, particularly where the offence is one of basic intent or the alleged mistake concerns defensive force. The result is a body of rules which is coherent only if understood as a compromise between subjectivist principle and public-protection policy. The central doctrinal contrast is this: a sober honest mistake may negate mens rea or support a defence even if unreasonable, whereas a mistake caused by voluntary intoxication is commonly excluded for basic intent offences and for self-defence. Yet intoxication is not a status of outlawry. If the offence requires a particular ulterior purpose or specific intention, voluntary intoxication may prevent the prosecution proving that element. Involuntary intoxication may be relevant more broadly, but only if it truly prevents proof of the necessary mens rea; if D forms the intent despite being drugged, liability remains possible. For Durham Criminal Law, the topic is best revised as a set of nested questions. First, what is the relevant element: offence mens rea, dishonesty, consent, belief, or defensive circumstance? Secondly, what kind of mistake is alleged: factual, legal, moral, induced by intoxication, or independent of intoxication? Thirdly, what kind of intoxication is present: voluntary, involuntary, therapeutic, Dutch courage, or unexpected disinhibition? Fourthly, what is the classification of the offence: specific intent, basic intent, or a statutory offence using reasonableness language? The best exam answers do not recite Majewski mechanically. They identify the exact point at which mistake or intoxication intersects with the actus reus and mens rea of the offence studied in earlier weeks.

The common law has long been suspicious of intoxication. The older judicial instinct was moralised: drunkenness was not to excuse violence, disorder or sexual aggression. But the modern law could not simply ignore the logic of mens rea. If an offence requires proof that D intended a particular result, or knew a circumstance, it is hard to maintain a conviction where intoxication genuinely prevented that state of mind from existing. The history of the law is therefore a history of managed inconsistency. The law permits intoxication to negate some forms of mens rea, while refusing to allow self-induced incapacity to operate as a general excuse. The classic pre-modern position appears in the treatment of Dutch courage. In Attorney-General for Northern Ireland v Gallagher [1963] AC 349, D formed an intention to kill, drank to steady his nerve, and then killed. The House of Lords rejected the suggestion that intoxication at the time of the act erased responsibility: the prior formation of the murderous design supplied culpability. The decision reflects a broader idea that intoxication cannot be used as a device for insulating intentional wrongdoing. The decisive modern structure was established in DPP v Majewski [1977] AC 443. The House of Lords drew the familiar distinction between offences of specific intent and offences of basic intent. Voluntary intoxication may be evidence that D lacked specific intent; it is no answer to a basic intent offence. The rationale was partly evidential and partly normative. If a person voluntarily takes alcohol or dangerous drugs, he is treated as reckless as to the risk that he may commit offences requiring no more than recklessness or general intent. That recklessness in becoming intoxicated is then allowed to stand in for the mens rea of the basic intent offence. The classification has been criticised because it is not conceptually pure. ‘Specific intent’ does not simply mean intention, and ‘basic intent’ does not simply mean recklessness. Murder is specific intent; assault occasioning actual bodily harm is basic intent; criminal damage is generally basic intent; burglary with intent may contain a specific ulterior element. The categories function as policy labels rather than as a taxonomy of mental states. Mistake developed along a different path. The criminal law gradually abandoned any general requirement that mistakes must be reasonable. For many offences, an honest mistake of fact, however foolish, prevents proof of the required mens rea. In self-defence, cases such as R v Williams (Gladstone) (1984) 78 Cr App R 276 and Beckford v R [1988] AC 130 affirmed that D is judged on the facts as he honestly believed them to be. Parliament later preserved that broad subjective approach in Criminal Justice and Immigration Act 2008, s 76, but expressly excluded mistaken beliefs attributable to voluntarily induced intoxication. In sexual offences, Parliament moved differently. The Sexual Offences Act 2003 makes D’s belief in consent relevant only if it is reasonable. That is not a common-law intoxication exception; it is a statutory definition of the offence. Historically, therefore, the law’s movement has been uneven: mistake is increasingly analysed through the structure of mens rea, while intoxication remains governed by a moralised and policy-driven compromise.

The first principle is that mistake is not, in itself, a defence. It is a way of denying an element of liability or of supplying the factual basis of a recognised defence. If D takes V’s umbrella believing it to be his own, the mistake matters because it may prevent dishonesty or the intention permanently to deprive in theft. If D strikes V believing, honestly, that V is about to attack him, the mistake matters because self-defence assesses the necessity of force by reference to the circumstances as D believed them to be, subject to the statutory intoxication limitation. If D penetrates V believing V consents, the mistake matters only if the offence’s statutory requirements allow that belief to count. Under the Sexual Offences Act 2003, D’s belief in consent must be reasonable; an honest but unreasonable belief will not suffice. The second principle is that the relevance of mistake depends on the precise mens rea of the offence. This is why Durham students should cross-reference earlier weeks. For murder, voluntary intoxication may be evidence that D lacked intent to kill or cause grievous bodily harm. For unlawful act manslaughter, however, the base offence may be one of basic intent, and voluntary intoxication will usually not prevent liability if the elements of that offence are otherwise established. For non-fatal offences, intoxication may be relevant to s 18 of the Offences against the Person Act 1861 because that offence requires a specific intent to cause grievous bodily harm or resist arrest, but not ordinarily to s 20 or assault occasioning actual bodily harm. For theft, because dishonesty and intention permanently to deprive are central, intoxication may in principle prevent proof of those mental elements, but the evidential burden in practice is demanding: juries may infer intent despite drunkenness. For criminal damage, voluntary intoxication is generally no answer to simple criminal damage because recklessness is sufficient; however, an honest belief in lawful excuse under Criminal Damage Act 1971, s 5 may be relevant, subject to the separate problem of intoxicated mistake.

The case-law is best organised around five propositions. First, voluntary intoxication may prevent proof of a specific intent but is no defence to crimes of basic intent. That is the Majewski proposition. Its practical significance is enormous: it explains why drunken defendants may escape murder but be convicted of manslaughter, or why intoxication may matter for s 18 but not s 20. The decision should not be treated as a neat piece of mens rea theory. It is a rule of attribution grounded in culpability for becoming intoxicated and in the perceived need to protect the public from drink- and drug-fuelled violence. Secondly, intoxication cannot be deliberately engineered as a defence. Gallagher remains the paradigm: D who forms the plan and drinks for courage cannot rely on the subsequent drunken state to deny liability. The criminal law looks at the transaction as a whole, not at an artificially isolated instant. Thirdly, involuntary intoxication is relevant only if it prevents proof of mens rea. Kingston is the leading authority. D was drugged by others, but the House of Lords held that if the jury were sure he formed the relevant indecent intent, liability followed. The case is morally troubling because intoxication lowered D’s inhibitions, but doctrinally it insists that disinhibition is not absence of intention. Fourthly, the boundary between voluntary and involuntary intoxication is sensitive to the nature of the substance and D’s appreciation of risk. Hardie is important because the Court of Appeal refused to treat Valium, taken for calming purposes, as equivalent to alcohol or dangerous drugs. The question was whether taking it was reckless. Heard later confined the category by stressing that voluntary consumption of alcohol before sexual assault could not negative a basic intent offence. Fifthly, drunken mistake as to defensive necessity is excluded. O’Grady refused self-defence where D, intoxicated, mistakenly thought the victim was attacking him. Hatton applied the same principle after the statutory and common-law refinement of self-defence. Williams and Beckford remain vital for the general rule that honest mistake may suffice; O’Grady and s 76(5) mark the intoxication exception. Lipman occupies a special place. D took LSD, hallucinated that V was a snake, and killed her. He lacked the specific intent for murder but was convicted of manslaughter. The case remains the clearest practical demonstration of the specific/basic distinction in homicide. For exams, the landmark cases should not be listed chronologically. They should be deployed functionally: Majewski for the classification rule; Lipman for murder to manslaughter; Gallagher for Dutch courage; Hardie for therapeutic involuntariness; Kingston for involuntary intoxication with actual intent; O’Grady and Hatton for intoxicated self-defence; Heard for basic intent in sexual assault.

See practice questions section below.

See further reading section below.