Why actions matter to the criminal law

Compare the following two cases:

Poison: Patient is terminally ill and is in a lot of pain that cannot be relieved through palliative measures. In order to bring his suffering to an end, Doctor injects Patient with some poison and Patient dies.

Dehydration: Patient is terminally ill and is in a lot of pain that cannot be relieved through palliative measures. Doctor orders that Patient should not be given any more food and water, and Patient dies of dehydration.

In Poison, Doctor commits murder. In Dehydration, Doctor does not commit any crime at all. But the only difference between these two cases is that Doctor has performed an act in Poison whereas all he is guilty of in Dehydration is an omission. Doctor could only be held liable for his omission in Dehydration if he had a duty to continue feeding Patient and giving him water; but he would have had no such duty if there was no point in continuing to treat Patient and giving Patient food and water is regarded as part of Patient’s ‘treatment’ (as it was in Airedale NHS Trust v Bland). In contrast, Doctor will always be under a duty not to take positive steps to bring Patient’s life to an end.

But we can’t just rest content with saying that Doctor is held liable in Poison while he is held blameless in Deyhydration because he breached a duty in Poison while he did not in Dehydration. That just raises the question of why there is a duty not to act in Poison when there is no duty to act in Dehydration. Why is what Doctor did in Poison so much worse than what he failed to do in Dehydration that the law will say that there was a duty not to act in Poison while there was no duty to act in Dehydration? After all, if we are just looking at the consequences of what Doctor did, Doctor’s actions in Poison look a lot better than Doctor’s omissions in Dehydration. In Poison, Doctor brought Patient’s suffering to a quick and painless end. In Dehydration, Doctor condemned Patient to a very difficult, agonising and prolonged death (though doing so may still have been in Patient’s ‘best interests’ given the alternative of allowing Patient to continue to suffer weeks or even months of extreme pain before the final end).

Some would argue, on this basis, that (to quote Lord Mustill in Bland) the law is ‘both morally and intellectually misshapen’ in distinguishing between Poison and Dehydration and that we should not find Doctor guilty of murder in Poison. Others (see, for example, Finnis, ‘Bland: crossing the Rubicon?’ (1993) 109 LQR 329) would agree with the premise of this argument but disagree with the conclusion: they would argue that we can only restore order to this area of the law by overruling Bland and holding that Doctor is just as much guilty of murder in Dehydration as he is in Poison: giving a patient food and water should not be regarded as being part of a patient’s ‘medical treatment’ and even if there is no further point in giving the patient medical treatment, there is still a duty to give the patient food and water so long as the patient is able to ingest it. So a breach of that duty that brings about death, when accompanied with an intent to kill, should be regarded as murder.

From a religious point of view, it is easy to see why we might distinguish between Poison and Dehydration. In Poison, Doctor is intervening to alter the natural course of events that would otherwise determine when Patient dies – which is not allowed if you think that that natural course of events represents God’s will for when Patient should die. In contrast, in Dehydration, Patient dies of natural causes – if you don’t get water for long enough, you die. And while we could argue about whether Doctor violated God’s will in withholding food and water from Patient – with the result that, from a religious perspective, his failure to act in Dehydration is just as bad as his act in Poison – the fact that Poison represents a much more obvious case of ‘playing God’ than Dehydration may be enough to justify our drawing a distinction between the two cases.

However, in a secular society, religious arguments for distinguishing between Poison and Dehydration are out of bounds. Can we find a non-religious basis for distinguishing between these two cases? The normal secular arguments for why the law should draw distinctions between acts and omissions – that if the law is too ready to find duties to act this will (1) undermine civil liberties; (2) create a great deal of uncertainty in the law; (3) deprive people of opportunities to earn credit from others for doing the right thing when they weren’t compelled to do so; (4) ‘punish’ people for putting themselves in positions to help others in need by subjecting them to legal duties that people who have made more selfish choices will never incur – don’t seem to apply here to distinguish Poison from Dehydration. If the law said there were a duty to act in Dehydration, doing so wouldn’t seem to have any of the undesirable effects that would normally accompany the law’s finding that one stranger owed another stranger a duty to save him or her from harm. However, it may be that we can make a ‘civil liberties’ argument in favour of the law’s finding that Doctor had a duty not to act in Poison when it says that he did not have a duty to act in Dehydration.

The argument goes as follows: if the law finds that it was permissible for Doctor to kill Patient in Poison, then it will not be long before the law finds that it was compulsory for Doctor to kill Patient in Poison. This is because the law gives doctors very little leeway in treating their patients – either the doctor had a duty to administer a particular treatment, or he had a duty not to administer that treatment: the law rarely (never?) says to a doctor ‘It’s entirely up to you whether or not to you administer that treatment.’ So if the law said it was permissible for Doctor to kill Patient in Poison then it’s hard to see how it could not then make the further leap of saying that it was compulsory for Doctor to kill Patient in Poison. But saying this would involve a very serious interference with Doctor’s civil liberties: many doctors do not want (for whatever reason) to kill their patients, and would seriously resent the law’s compelling them to kill their patients. And if doctoring were seen as coming with a duty to kill in certain circumstances, then the number of people wanting to become doctors would radically diminish. The same problem does not arise in relation to the law’s saying that Doctor has no duty to feed or water Patient in Dehydration. If the law follows that up by saying that it is compulsory for Doctor not to give Patient any food or water in Dehydration, most doctors would be able to be reconciled with that. The fact that the law might get in the way of a doctor acting on her normal instinct to give a patient food or water – even if nothing more can be done for the patient – would not be experienced by the doctor as a violation of her civil liberties any more than the fact that the size or location of her hospital places a limit on how many patients she can treat, or her need to respect her patient’s wishes places a limit on what she can do for that patient. The state of the law would just be one more fact which places a limit on how much she can do for the patients she would otherwise wish to treat, and which she just has to learn to live with.

The defence of insanity

The cases of Anders Breivik (who shot 69 people on Utoya Island in Norway) and James Holmes (who shot 12 people in a cinema in Aurora, Colorado) raise again the issue of when a criminal defendant should be able to take advantage of a defence of insanity. As laid down in the M’Naghten Rules (‘MR’, for short), the defence of insanity will be available to a defendant if, at the time he acted, he was ‘labouring under such a defect of reason, from a disease of the mind, as not to know the nature and quality of the act he was doing; or if he did know it, that he did not know what he was doing was wrong.’

One criticism that it is often made of the defence of insanity is that it is medically outdated – that our notions of when someone can be said to have been medically insane in acting as he did have developed since 1843, when the MR were laid down, and the law has failed to keep pace with those developments. However, this criticism is unjustified: in my view, the MR are not intended at all to reflect clinical or medical ideas as to when someone should be said to be insane.

Instead, I think the MR are directed at identifying cases where: (1) it was impossible for the defendant to alter his conduct to make it conform to the law (which accounts for the requirement that the defendant must not have known what he was doing or must not have known that what he was doing was wrong), and (2) the defendant was unable to alter his conduct to make it conform to the law because he suffered from an internal condition which means that the defendant is likely to do what he did again, and therefore constitutes a public danger that needs to be dealt with (which accounts for the requirement that the defendant’s incapacity arises out of a ‘disease of the mind’).

On this view, then, the MR are not at all concerned to reflect modern medical notions of when someone can be said to be insane: they arise out of a dual concern to uphold the requirements of the rule of law – which would dictate that someone not be held liable for breaking the law if he was incapable of being guided by its dictates – and to ensure that people who get the treatment they need to deal with conditions that they have that, if unaddressed, will prevent them from complying with the law in the future.

So it really does not matter whether Breivik or Holmes were medically insane when they acted as they did (which they probably were). They were both clearly capable of conforming their conduct to the law, and as such can be properly held accountable for their failures to exercise that capacity. If – having been imprisoned for their crimes – they also require medical treatment to deal with their psychiatric conditions (as Peter Sutcliffe, the Yorkshire Ripper, did, eventually being incarcerated in Broadmoor Hospital under the Mental Health Act 1983) then they should receive it; but their psychiatric states cannot excuse them from being punished for their failures to obey the law.

All of this is not to say that the law on when someone will be found not guilty by reason of insanity is perfect. Four criticisms of the law can be made.

(1) Stigma. There is a concern that defendants who fall within the MR might be put off pleading that they are ‘not guilty by reason of insanity’ because of the stigma attached to being labelled as ‘insane’. This concern gets particularly strong in cases like Sullivan (epilepsy) and Hennessy (diabetes) where what triggered the defendant’s incapacity to comply with the law was an internal condition that no one would ordinarily categorise as ‘insanity’. If the MR are really not concerned with medical notions of insanity, there is a case for saying that we should not use the term ‘insanity’ to refer to the defence to which the MR apply. However, it is difficult to think of what alternative name we could give this alternative defence. ‘Not guilty by reason of internal incapacity’ feels a bit too clinical.

(2) Civil liberties. Another reason why defendants might be unwilling to plead ‘not guilty by reason of insanity’ is that if the plea is successful, then the court is then empowered under s 3 of the Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 to make an order requiring the defendant to be restrained in hospital for an ‘unlimited period of time’. Now – the court does not have to make such an order (though it does in the case where the sentence for the offence which the defendant was charged with is fixed by law (effectively, murder cases)) – it could make a more limited order, or even an absolute discharge. But the possibility that a court might make such an order in their case would make a defendant – particularly one who is charged with a relatively petty offence such as shoplifting – hesitate long and hard before making a plea of ‘not guilty by reason of insanity’ for fear of being trapped in a One Flew Over the Cuckoo’s Nest situation of being trapped in an institution where you are not insane, but are treated as though you are, and are eventually driven crazy by the experience. One way of getting round this problem would be to amend the 1991 Act to require a defendant who has been committed to hospital having pleaded ‘not guilty by reason of insanity’ to appear before a judge at regular intervals so that the judge (preferably the one who presided over the defendant’s trial) can evaluate their treatment and condition and ensure that they are not being treated inappropriately.

(3) Falling between two stools? The defence of ‘not guilty by reason of insanity’ exists alongside the ‘defence’ of non-insane automatism – which is not really a defence, but a denial that the defendant committed the actus reus of the crime because an external cause (for which the defendant was not at fault) meant that the defendant was unaware of what he was doing (the defendant ‘blacked out’) or was aware of what he was doing, but was unable to control his bodily movements (the defendant was acting reflexively). However, there may be one situation which falls between these two defences, because it does not come under either of them. This is the situation where due to an external cause (for which the defendant was not at fault) the defendant thinks that there is nothing wrong – either legally or morally – with acting in a particular way, and he goes on to act in that way, where acting in that way amounts to the actus reus of an offence. For example: D is hit on the head by a stone. When he comes to, he thinks that he is the emperor of the world, and that all women are required to have sex with him, if he so wishes. He proceeds to have sex with V against her will. D won’t be able to rely on a defence of non-insane automatism because he was not acting as an automaton in having sex with V – he knew what he was doing, and was in control of his body. But will he be able to take advantage of a defence of ‘not guilty by reason of insanity’? The difficulty is that his delusions were triggered by an external cause – that of being hit on the head by a rock. Perhaps it could be argued that the external cause triggered a ‘disease of the mind’ and in that way allow him to take advantage of the defence of insanity. But this seems hard to reconcile with cases like Bailey, where the defendant acted under the influence of an insulin injection, and was held not to be able to take advantage of the defence of insanity. (Though it looks like the defendant knew what he was doing in that case – though he might not have been aware of the effects of his actions.)

(4) Strict liability. In DPP v Harper, the Court of Appeal ruled that the defence of insanity did not apply to strict liability offences – in that case, driving with excess alcohol in one’s bloodstream. This seems wrong. If the MR are directed at someone’s being unable to conform their conduct to the law, then there seems no reason why they should not apply as much to strict liability offences as they do to offences which require some degree of mens rea for their commission. A defendant who seeks to take advantage of the MR is not denying that he had the mens rea for the offence with which he is charged: he is making a much more basic point – that he should not be held responsible for his actions because they were either not his actions or they were performed in ignorance of the fact that they might be contrary to the law (and he was not at fault for that ignorance – so the maxim ignorantia legis non excusat does not apply). (Having said that, there are a couple of cases which suggest that a strict liability offence will capture someone who was completely incapable of avoiding committing that offence, and on that basis they refused to find that the offence in question was a strict liability offence. They are Sherras v De Rutzen (refusing to find that the offence of selling alcohol to an on-duty policeman was a strict liability offence as it would be impossible for someone selling alcohol to know whether a policeman was on-duty or off) and Lim Chin Aik v R (refusing to find that the offence of entering Singapore when having been prohibited from doing so was a strict liability offence as the notices prohibiting someone from entering Singapore were only published in Singapore, and so it would be impossible for someone outside Singapore to know whether they were prohibited from entering Singapore). These cases may be confusing absolute liability – where it is no defence to say ‘It was impossible for me to avoid the prohibited conduct’ – and strict liability, where in principle it should be a defence to say that it was impossible to avoid doing the thing which is prohibited.)

Why intentions matter to the criminal law

The intentions with which you acted matter to the law in a couple of different respects:

(1) The fact that D had a particular intention in acting as he did can make his conduct unlawful, when if he did not have that intention his actions would have been perfectly lawful. In such a case, we can say that D’s intention was wrong-making.

(2) The fact that D had a particular intention in acting as he did can make his conduct more seriously wrongful than it would have been had he not had that intention. In such a case, we can say that D’s intention was aggravating.

Wrong-making intentions

Consider the following two cases. In both cases, Patient is terminally and is in crippling pain.

Allowed: Dr Good gives Patient a dose of morphine that is sufficient to relieve his pain, realising at the same time that if Patient receives a dose of morphine of that quantity, then Patient’s death is likely to be significantly accelerated. There is no other means of relieving Patient’s pain. Patient subsequently dies from the dose of morphine that he received.

Not Allowed: Dr Bad gives Patient a dose of morphine of exactly the same quantity as that administered by Dr Good. But when Dr Bad administers the dose, he intends that the dose that Patient receives will kill him, thereby ending his torment. Patient subsequently dies from the dose of morphine that he received.

(It should be noted that there is actually no medical evidence that doses of morphine administered in sufficient quantities to relieve pain can actually accelerate people’s deaths; but that does not affect the discussion below.)

The only difference between these two situations is that in Allowed, Dr Good does not intend (in the sense of acting with a particular aim or purpose) to kill Patient, whereas in Not Allowed, Dr Bad intends to kill Patient.

In every other respect, the situations are identical. Both doctors act with the ultimate intention of relieving Patient’s pain (note that you can have more than one intention in a given situation because you can have more than one aim or purpose in acting in a particular way) – it’s just the means by which they intend to relieve Patient’s pain differ. And in terms of the doctor’s physical actions, they are identical: both administer the same dose of morphine, and both bring about Patient’s death at the same time.

So the only difference between the two situations is what is running through the doctors’ heads at the time they inject the morphine into Patient. But that difference is enough to make Dr Bad a murderer, while Dr Good is not guilty of any offence and is regarded as having acted perfectly legitimately, indeed laudably. Why does the difference in the two doctors’ intentions matter so much to the criminal law? We can consider a number of different explanations:

(1) Usurpation. Probably the most historically accurate explanation of why Dr Bad’s intentions make what he did unlawful is that in deciding to kill Patient, he is ‘playing God’ – usurping a decision (as to when Patient will die) that it was for God to make. However, in a secular society, such an explanation is regarded as illegitimate. However, is it possible to argue that Dr Bad’s decision to take Patient’s life was wrongful because it involved Dr Bad in usurping a decision that was for Patient to make? The answer must be ‘no’ because Dr Bad would still be regarded as murderer even if he was simply carrying out Patient’s instructions in injecting him with morphine. We have to look elsewhere for explanations as to why Dr Bad’s intentions make him a wrongdoer.

(2) Patients’ interests. Maybe we could argue that it is in the interests of patients for the law to say that their doctors are not allowed to act with the intention of killing them. The idea is that patients would do worse under any other alternative rule, such as ‘A doctor is allowed to act with the aim of killing one of his or her patients if conditions x, y and z are satisfied.’ Four arguments might be made for saying that patients would be worse under an alternative rule. (i) A slippery slope argument that once we drop the bar on doctors’ intentionally killing their patients by adopting an alternative rule, various institutional and political pressures will result in that alternative rule being gradually reformed and relaxed, and patients will do worse under those successor rules than they would have done had the original bar on intentional killing been retained. (ii) A pessimistic argument that while patients might do better under the alternative rule, if the alternative rule is applied correctly, human error means that the alternative rule will not be applied correctly, and patients will consequently do worse under the alternative rule than they would have done under the original rule which barred intentional killing under any circumstances. (iii) A psychological argument which observes that human beings do not cope well under circumstances of cognitive dissonance – where they are subjected to two conflicting emotions or reactions. In such a situation, human being resolve the dissonance by changing the way they look at the situation that is triggering their emotion or reaction so as to ‘mute’ one of the dissonant emotions or reaction. A doctor who intentionally kills a patient under an alternative rule might well experience a period of cognitive dissonance where his training makes him think ‘It will be a bad thing if this patient dies’ and the law (which gives effect to the alternative rule) makes him think ‘It is okay to bring this patient’s life to an end’. The doctor might well resolve this dissonance by beginning to downgrade the importance of saving his patients’ lives. In other words, he will become desensitised – with sinister implications for his other patients. So the doctor’s patients will get worse treatment from the doctor under the alternative rule than they would have done had the doctor been forbidden from ever intentionally killing a patient. (iv) A combined argument, which says – on the basis of points (ii) and (iii), above – that under an alternative rule, patients will become scared to consult their doctors and will be especially scared to go into hospitals, and will as a result do worse than they would have done had the bar on intentional killing of patients been retained.

I think that all of the arguments (i) – (iv), above, have something going for them, and help to justify the continued existence of a rule barring doctors from intentionally killing their patients, while at the same time doctors are allowed to do things that they know will result in their patients’ dying, so long as doing those things is a necessary and proportionate means of treating their patients’ conditions.

Aggravating intentions

Consider the following two situations. In both cases, Parents have a two-month old Baby.

Useless: Parents neglect to feed Baby because they are spending all their money on drugs; as a result, Baby dies.

Evil: Parents neglect to feed Baby because they resent having to spend any money on her, and hope that she dies; Baby does eventually die.

Parents will be guilty of gross negligence manslaughter in Useless; while they will be guilty of murder in Evil. This is because Parents acted (or, more accurately, failed to act in breach of their duty to feed Baby) with the aim or purpose of killing Baby in Evil, while they had no such aim or purpose in Useless. (I disregard the possibility that Parents might be ‘held’ to have had an intent to kill in Useless under Woollin, on the basis that they foresaw that death was virtually certain to occur. If this troubles anyone, just assume that Parents were so high all the time in Useless that they thought Baby wouldn’t die even if he wasn’t given anything to eat.) Why does the fact that Parents had an intent to kill in Evil make their offence much more serious than the offence committed by the Parents in Useless, who had no such intent? A couple of different explanations can be given:

(1) Character. The Parents in Evil have worse characters than the Parents in Useless. The Parents in Useless did not care enough about the value of their Baby’s life. The Parents in Evil did not care at all about the value of their Baby’s life. This difference in character may lead us to want to stigmatise the Parents in Evil and the Parents in Useless in different ways – by giving them different names (murderers in Evil; manslaughterers in Useless) – so  that people will know what sort of characters they are dealing with when they come across them.

(2) Danger. Because the Parents in Evil have worse characters than the Parents in Useless, the Parents in Evil are more dangerous than the Parents in Useless. The Parents in Evil have proven themselves willing to kill other people, when the Parents in Useless have not provided any evidence that they are willing to do the same. Given this difference in dangerousness, we might want to lock up the Parents in Evil for longer than the Parents in Useless, both to prevent the Parents in Evil doing any more harm to anyone else, and to provide greater opportunities for their rehabilitation.

Why consequences matter to the criminal law

The consequences of your actions matter to the criminal law in a couple of different respects:

(1) The law distinguishes between successful attempts and unsuccessful attempts, so that defendants who are guilty of a successful attempt are treated worse under the criminal law than defendants who are guilty of an unsuccessful attempt – when the difference between a successful and an unsuccessful attempt can just be a matter of luck.

(2) Crimes of constructive liability make a defendant who is already guilty of committing one criminal offence liable for a much more serious offence if his initial crime has certain consequences; and this is so even if the defendant did not intend or foresee that his initial crime would have those consequences. Examples of constructive liability in the criminal law are: (a) murder (where the defendant merely intended to cause gbh to his victim, but death resulted as well); (b) constructive manslaughter; (c) causing death while driving uninsured/disqualified/unlicensed; (d) maliciously inflicting gbh (where all that was foreseen was that the victim would suffer some kind physical harm); (e) assault occasioning actual bodily harm; (f) accomplice liability under the law on joint enterprise.

This post attempts to explain: (1) why the law might distinguish between successful and unsuccessful attempts; and (2) why the law might recognise crimes of constructive liability.

Successful and unsuccessful attempts

Contrast the following two cases:

Hit: Assassin trains his gun on Target, 500 yards away. Assassin pulls the trigger and Target is shot dead.

Miss: Assassin trains his gun on Target, 500 yards away. Assassin pulls the trigger but a sudden gust of wind means that the bullet just misses Target.

In Hit, Assassin will be guilty of murder and get a mandatory life sentence; in Miss, Assassin will only be guilty of attempted murder and won’t get a mandatory life sentence. Why the difference?

A lot of people think that the law should treat Assassin the same in both Hit and Miss as his intentions were exactly the same in both cases, and it’s just a matter of luck that Assassin killed his victim in Hit but not in Miss. I think what this view overlooks is that in both Hit and Miss, Assassin does two things wrong, not one. In both cases, Assassin tries to kill Target – and that’s wrong. But the other thing wrong that Assassin does is that he knowingly exposes Target to the risk of suffering bad luck, by being killed. When Assassin pulls the trigger, he makes it a matter of luck whether or not Target is killed, and he was aware that he was doing that.

So we need to punish Assassin both for (1) trying to kill Target and for (2) knowingly exposing Target to the risk of suffering bad luck. But how do we do this? My suggestion is that we punish Assassin for (1) through a fixed penalty (say, 15 years), and we punish Assassin for (2) through a conditional penalty which Assassin will incur if Target does end up suffering bad luck and is killed. This conditional penalty is a fitting penalty for the wrong committed by Assassin in knowingly exposing Target to the risk of suffering bad luck: if Target ends up suffering bad luck, then Assassin will also suffer bad luck by incurring the conditional penalty.

The result is that we end up punishing Assassin more in Hit than in Miss. In Hit, Assassin incurs both the fixed penalty and also the conditional penalty, as the bad luck that he knowingly exposed Target to the risk of suffering has materialised. But in Miss, Assassin only incurs the fixed penalty: Assassin shares in Target’s good luck – as Target is not killed, Assassin escapes the conditional penalty.

Constructive liability

Contrast the following two cases:

Scary: Man and Wife are driving home from a party, where Man caught Wife kissing another man. Man (who is driving) says to Wife, ‘When we get home, I’m going to beat you to an inch of your life.’ Wife is so scared by this threat that she jumps from the moving car and scars her knees.

Funny: Man and Wife are driving home from a party, where Man caught Wife kissing another man. Man (who is driving) says to Wife, ‘When we get home, I’m going to beat you to an inch of your life.’ Wife laughs and says ‘I’d like to see you try.’

In Scary, Man is guilty of assault occasioning actual bodily harm; while in Funny he is only guilty of an assault. And this is so even though Wife’s reaction to Man’s words may have just been a matter of luck. Why the difference?

We can’t use the ideas discussed in the previous section to explain the difference between Scary and Funny as Man may not have knowingly exposed Wife to the risk of suffering actual bodily harm in either case. (In Scary, he may have been completely astonished that his Wife would react to his threat by jumping from a moving car.) So we need a different idea to explain why the law treats these two cases differently.

We can draw on the idea of a moral threshold to explain why the law might treat Man more harshly in Scary than in Funny.  The idea is that Man crosses a moral (actually, legal – but ‘moral’ sounds better) threshold when he threatens Wife with a beating: he does something wrong. We don’t want him to do this. Now – how do we stop him from crossing that threshold? My suggestion is that by putting Man on notice that if his threat causes his Wife to suffer actual bodily harm, he will be subject to a more serious penalty than he would be if his Wife did not suffer any such harm, the law does two things to help encourage Man to stay the right side of the line of not threatening his Wife:

(1) The law makes Man uncertain as to what punishment he will receive if he crosses the moral threshold and threatens his Wife. He won’t know whether he will get the standard penalty for committing an assault, or a more serious penalty in the event that his assault results in his Wife suffers actual bodily harm. This uncertainty can be expected to deter Man from crossing the moral threshold of threatening his Wife as it makes it much harder for him to determine whether threatening his Wife would be ‘worth it’. (Note that there is no rule of law objection to placing Man in a position of uncertainty as to what punishment he will receive if he assaults his Wife, as Man could easily avoid the uncertainty by simply not assaulting his Wife.)

(2) If Man tries to calculate what the probability is that he will end up receiving a penalty for assault occasioning actual bodily harm, he will be forced to think about what the chances are that his assault will result in his Wife suffering actual bodily harm. In thinking about this, he may be brought to realise what effects his assaulting his Wife will have on her, and will come to the conclusion that given those effects, he should not assault his Wife and will voluntarily decide not to cross the threshold of assaulting his Wife.

So constructive liability may encourage people not to cross the moral thresholds to which constructive liability attaches by: (1) deterring them from crossing those thresholds by making them uncertain what liability they will incur if they cross that threshold; and (2) persuading them not to cross those thresholds by making them reflect on what the effects of crossing those thresholds will be.

The law on intoxication

The basic point

There is no such thing as a defence of intoxication under the criminal law. If you ever find yourself writing ‘D may be able to take advantage of a defence of intoxication’ you are doing something very, very wrong. You need to stop and think – why am I saying this?

Intoxication is only relevant to whether or not a defendant has committed a criminal offence if it means: (1) he did not commit the actus reus of the offence; or (2) he did not have the mens rea for the offence; or (3) he mistakenly believed in facts which, if true, would mean he had a defence to being found guilty of committing that offence. If none of (1), (2) or (3) apply, then the defendant is guilty – the fact that he acted in an intoxicated state will be completely irrelevant.

Kingston illustrates the point. Defendant’s drink was spiked, his inhibitions were consequently loosened, and he ended up sexually assaulting a boy. He knew what he was doing at the time and was in control of his body. He was found guilty: he had the actus reus of the offence (unconsented to touching of a sexual nature), he has the mens rea (intention to touch, or recklessness as to touching) and was not acting under a belief which, if true, would have given him a defence. The fact that the defendant would not have acted as he did had his drink not been spiked was irrelevant. He was still responsible for his actions. If I offered you a million pounds to kill your best friend, and you took the money and killed your best friend, you could not dream of trying to get off the charge by arguing, ‘Had I not been offered the money, I would never have killed my best friend.’ And yet people still try to argue that Kingston should have been acquitted. Whatever it was that caused you to loosen your inhibitions – money, drugs, drink, lust – if you are still responsible for your actions, you can still be held accountable for them under the criminal law.

Let’s now look at how the law handles cases when a defendant can argue that either (1), (2) or (3), above, apply.

No actus reus

In this sort of case, the effect of D’s intoxication either means that D had no idea what he was doing, or D was not in control of his body. In either case, D might be able to argue: I didn’t have the actus reus because I didn’t cause whatever harm it is that my body did while I was blacked out or had no control over my body’s physical movements. If this argument works, then D is entitled to be acquitted on the basis of non-insane automatism (not insane automatism, because the automatism had an external cause – the intoxication). But there is one exception. If D was responsible for his own intoxication then we can say (at least where the actus reus of the offence requires that it be shown that D caused a particular harm) that D did have the actus reus. By causing himself to get drunk, he was the ultimate cause of the harm that his body did while D was blacked out or not in control of his body’s physical movements. So we can contrast two cases:

Wild Thing 1. D gets so drunk in a pub that he ends up having convulsions. As he is flailing around, one of his fists hits a fellow drinker, V. In this case D has the actus reus of assault occasioning actual bodily harm (assuming that the hit did abh to V). D  ultimately caused his fist to touch V by getting so drunk that he went into convulsions, and this caused (we are supposing) V to suffer actual bodily harm.

Wild Thing 2. Same as Wild Thing 1 except D was drinking what he thought was orange juice, but it had been spiked by T. In this case, D does not have the actus reus for assault occasioning actual bodily harm because he did not cause his fist to hit V. T did that; so if anyone is going to be guilty of assault occasioning actual bodily harm, it should be T. But D is entitled to be acquitted here on the basis of non-insane automatism.

No mens rea

In this sort of case, the effect of D’s intoxication means that D can deny that he had the mens rea for the offence with which he has been charged. For example, if, say, he was charged with murder, D would be arguing – my intoxicated state means that I did not have an intent to kill or an intent to cause gbh. If, on the other hand, he was charged with malicious wounding, D would be arguing – I did not appreciate that there was an unreaonable risk that V would suffer some physical harm from my actions. Here are a couple of situations where D would be able to make such an argument:

Polar Bear. D gets blind drunk in a pub. As he walks outside the pub, alone, he sees what he thinks is a polar bear coming towards him. (In fact, it is V, dressed in a white parka.) Thinking that the polar bear’s head would look great on his wall, D grabs hold of what he thinks is the polar bear and cuts its head off with a handy knife. On sobering up, D is told that he decapitated V and is being charged with murder. D has the actus reus for murder: he caused V’s death by sawing V’s head off. But D will argue that he did not have the mens rea for murder when he was sawing V’s head off: he intended to kill a polar bear, not a human life in being.

Note that this argument would not work if D was getting blind drunk in order to get up the courage to go home and kill his wife. In such a case, we could dial back the moment when D caused V’s death to the moment when he was getting drunk in the pub, and say that that act of getting drunk caused V’s death because V would not have died had D not gotten blind drunk, and there was no obvious break in the chain of causation after the moment he got drunk to break the chain between the getting drunk and V’s dying. And at the moment D got drunk, he did have an intent to kill – he was getting drunk with the aim or purpose of killing his wife. So in that situation, D would not be able to deny that he had the mens rea for murder, contemporaneous with the actus reus, and would consequently be convicted of murder. This is known as the ‘Dutch courage rule’ (first articulated by Lord Denning in Attorney-General for Northern Ireland v Gallagher).

But let’s assume the ‘Dutch courage rule’ doesn’t apply in Polar Bear: D was getting drunk just for the hell of it. In such a case, D would argue that he should be acquitted because he did not have the mens rea for murder.

Darts. D gets blind drunk in a pub. He gets up, intending to have a game of darts. But he is so drunk, he thinks that the darts in his hand are actually little wriggling mice. He is so disgusted, he throws one of the ‘mice’ at a nearby drinker, V, shouting, ‘Look at this! That’s horrible!’ When D sobers up, he is told that V was actually struck in the eye by the dart that he threw at V, and that D is going to be charged with malicious wounding, contrary to s 20 of the Offences Against the Person Act 1861. D has the actus reus: what he did in throwing the ‘mouse’ at V caused V to be wounded. But D will argue that he did not have the mens rea: when he threw the ‘mouse’ at V, he did not foresee that V would suffer ‘some kind of physical harm’ as a result of his actions.

In cases where D seeks to rely on his own intoxicated state in order to deny that he had the mens rea of the offence with which he has been charged, a very simple rule (called the Majewski rule) applies to determine whether D can do this. In the case where D was to blame for getting drunk, the courts ask – Had D not been drunk, would he have had the mens rea for the offence with which he has been charged? If the answer is ‘yes’ then D will not be allowed to deny that he had the mens rea for the offence with which he has been charged. If the answer is ‘no’, then D will be allowed to persist in his plea that he did not have the mens rea for the offence with which he has been charged, and will be acquitted on the basis of no mens rea.

Let’s see how the Majewski rule would apply in our two situations:

Polar Bear. Here D’s drunken state means that he did not have the mens rea for murder, but he was also to blame for getting drunk, so we might prevent him denying that he had the mens rea for murder depending on the application of the Majewski rule. So we ask – Had D not been drunk, would he have had an intent to kill or an intent to cause gbh? The answer, pretty obviously, is ‘no’ so D is allowed to persist in his plea that he did not have the mens rea for murder and will be acquitted on the basis of no mens rea.

Darts. Here D’s drunken state means that he did not have the mens rea for malicious wounding, but he was also to blame for getting drunk, so we might prevent him denying that he had the mens rea for malicious wounding depending on the application of the Majewski rule. So we ask – Had D not been drunk, would he have foreseen that there was a risk of V suffering some kind of physical harm when he acted as he did? Well, had D not been drunk, he would have realised he was holding a dart, not a mouse, so he plainly would have foreseen that there was a risk of V suffering some kind of physical harm when he threw the dart at V. So the answer to our question is ‘yes’ and as a result D will not be allowed to deny that he had the mens rea for malicious wounding. As a result, D will be convicted of malicious wounding – he committed the actus reus, and is not allowed to deny that he had the mens rea, and will have no relevant defence.

The textbooks try to sum up the above by saying that a defendant who is charged with a criminal offence will – if he was to blame for getting drunk – be allowed to rely on his drunken state to deny that he had the mens rea for the offence if the offence was one of ‘specific intent’ (= mens rea requires an intent to produce a particular outcome), but not if the offence was one of ‘basic intent’ (= mens rea can be established merely by showing that the defendant foresaw that his actions would have a particular outcome). But this is dangerously vague. It would be better just to ask, in a case where the defendant seeks to rely on his drunken state to establish that he did not have the mens rea for the offence with which he has been charged, and the defendant was to blame for getting drunk, the Majewski question: Would the defendant have had the mens rea had he not been drunk? Answering that question will allow you to determine whether or not the defendant will be allowed to deny that he had the mens rea without having to worry about whether the offence that the defendant has been charged with is one of ‘specific intent’ or ‘basic intent’.

Mistaken belief in defence

In this sort of case, D’s intoxicated state means that he believed facts were true, that if they were true would have given him a defence, such as self-defence, or duress. Examples:

Banana Man. D is high on LSD. In his drunken state, he believes V is running towards him with a knife, and is about to stab him. D takes out a gun and shoots V dead. In fact, V was not holding a knife but a banana, which he was urging D to eat.

Down at Heel. D is in the pub with V. V is telling D about his financial problems, and says to D, ‘You’ll take care of me, won’t you? You’d really regret it if you didn’t.’ D is pretty drunk and misinterprets V as threatening that he will kill D unless D immediately gives him money. D pulls out a gun and holds up the pub, netting £2,000, which he instantly hands over to an astonished V.

In these cases, the courts ask a very simple question – Was D to blame for his intoxicated state? If he was, then he will not be allowed to rely on his mistaken belief to argue that he has a defence. But if he was not to blame for being intoxicated, and if the facts that D’s intoxicated state led him to believe were true would – if true – have given him a defence, then D will be allowed to raise that defence to the offence with which he has been charged.

So if, in Banana Man, someone had slipped D a tab of LSD without his realising it, and shooting V would have been a reasonable thing to do had V actually been about to knife D, then D will be acquitted of murder on the ground of self-defence. (Note that in Banana Man, the fact that D had the actus reus and mens rea for murder is easily established.) But if in Banana Man, D had taken the LSD, he will not be allowed to rely on his mistaken belief to argue that he should be acquitted on the ground of self-defence. (Note that this is an exception to the normal approach, established in Gladstone Williams, that a defendant will be entitled to raise a defence of self-defence so long as he honestly believed in facts that, if true, would have given him a defence of self-defence.)

The same is true in Down at Heel. If D was to blame for getting drunk, he will not be allowed to rely on a defence of duress when he is charged with theft. But if someone had been spiking his orange juice, he will be allowed to rely on his mistaken belief that V was threatening to kill him to set up a defence of duress. (This is assuming that on the facts that he believed to be true he would have had a defence of duress.)

The one case where a defendant who was to blame for getting drunk will still be allowed to rely on a mistaken belief in facts which, if true, would have given her a defence was established in Jaggard v Dickinson. Under s 5(2) of the Criminal Damage Act 1971, it is a defence to being convicted of causing criminal damage to property belonging to another if you believed the owner of the property would have consented to the damage. The defendant, in a drunken state that she was to blame for getting into, attempted to break into her friend’s house believing that her friend would be happy for her to do this. In fact, she had got the wrong house and was breaking down someone else’s door. She was acquitted of criminal damage to property on the basis that she honestly believed that the owner of the door would have been happy for her to break down the door. This was so even though the only reason she thought this was because she was drunk, and she was to blame for getting drunk.

Criticisms of the law

The law on intoxication is sometimes criticised on the basis that –

(1) It is unclear.

(2) It does not provide a defence to someone like the defendant in Kingston.

I don’t think either of these criticisms are justified. The law is pretty clear (as you can see from my summary above), and there is no way the law should be acquitting the defendant in Kingston. But this does not mean that the law in this area is beyond criticism. Two criticisms can be made:

(3) In Darts we end up convicting the defendant of a very serious offence – malicious wounding – simply on the basis that he was to blame for getting drunk. This does not seem right. If the defendant, when he got drunk, did not foresee the risk of other people coming to some physical harm as a result, he is being found guilty of an offence for which such foresight is required when he had no such foresight. He is being found guilty because he is prevented, under the Majewski rule, from denying that he had such foresight. Convicting someone on the basis of a fiction seems very unsatisfactory. It looks like the defendant in Darts is being convicted on a constructive basis – he did something bad when he got drunk, and then the offence he ends up being convicted of depends on the level of harm that resulted from his getting drunk without any requirement that he needs to have foreseen that harm.

(4) The workings of the above rules seems to produce some arbitrary results. Consider:

Tramp 1. Two tramps, Vladimir and Estragon, get very drunk and take a lot of drugs and fall asleep near each other. Vladimir wakes up, still intoxicated, and thinks that the sleeping Estragon beside him is a writhing mass of snakes that are about to attack him. Vladimir panics, pours the remaining alcohol near him onto Estragon and sets it on fire. Estragon dies. Vladimir is charged with murder but is acquitted on the basis that he lacked the mens rea for murder: he did not have an intent to kill a life in being. See R v Lipman.

Tramp 2. The same facts as Tramp 1, but in this case when Vladimir wakes up, he wrongly thinks that Estragon is attacking him with a knife. Vladimir strangles Estragon. Vladimir is charged with murder and is convicted on the basis that he had the actus reus for murder (he caused Estragon’s death), and the mens rea for murder (he intended to kill Estragon), and he can’t rely on his mistaken belief that he thought that Estragon was attacking him because that mistaken belief was the result of his intoxicated state, and he was to blame for being intoxicated. See R v O’Grady; R v Hatton.

But what, morally, is the difference between Tramp 1 and Tramp 2 that would justify us in acquitting Tramp 1 of murder, and finding Tramp 2 guilty? The treatment of Tramp 2 looks even more arbitrary given the fact that under Gladstone Williams a non-intoxicated but still stupid belief that one is about to be attacked might entitle a defendant to be acquitted of murder.