Coley & Ors v R

We have heard these three cases in succession because they have some features in common. Each raises a (different) question connected with the interplay between the law relating to voluntary intoxication and the law relating to insanity or (non-insane) automatism. Each calls, however, for consideration of its very particular facts. Neither individually nor collectively do they provide an occasion for any wide-ranging general statement of the law of insanity, still less of loss of capacity generally. We know that this area of the law is under active consideration by the Law Commission, which work will, we think, be of value. Although there have historically been very few cases which raise insanity, that has been because the statutory provisions governing the disposal orders which must be made if there is a verdict of insanity have historically inhibited attempts to rely on it. More recent changes in those disposal provisions may well lead to an increase in numbers. Any review must, critically, address both the law of loss of capacity and the means of disposal in such cases, so as to pay proper regard both to the interests of the individual defendant and to the public risk which he represents.

The defendant, Scott Coley, was convicted of attempted murder. He was 17 at the material time. He had no previous conviction of any significance, nor any history of violence to anyone or anything. He lived at home with his parents and worked with his father. He was described by neighbours as a quiet and pleasant young man. He was a regular and quite heavy user of strong cannabis. He liked knives, of which he had a collection in his bedroom, and which he sharpened from time to time.

He and his family were on excellent terms with the woman who lived next door. She was a single lady living with two teenage daughters and she had a partner who visited frequently. The partner was a cousin of the father of the defendant. The two households shared some interests, including the online playing of video games, of which a frequently used one was called Battlefield: Bad Company 2. It involved assuming the role of a combatant and scoring points according to (among other things) the number of opponents killed, particularly by close-quarters engagement and with knives.

On a Sunday in the summer of 2011, the defendant’s parents went to bed early, as was their habit owing to the need for an early start next day. He watched an action film called “Triple X”. According to his own account, which was not questioned, he went to bed at about 2330. At just after midnight he arrived on the upstairs landing of the house next door and pushed open the door of the main bedroom, inside which the lady neighbour and her partner were asleep. He was dressed in dark clothing with a balaclava over his head and he was carrying a large and sharp hunting-type knife (referred to as a “Rambo” knife). The lady neighbour woke, saw him, got up from the bed and screamed very loudly. Her partner leaped out of bed and confronted the defendant. The defendant stabbed him repeatedly with his knife, chiefly to the chest, arm, back and head. The partner nevertheless wrestled with him and in effect threw him downstairs. The defendant ran out of the house and into the garden of another neighbour who lived on the other side of his own house, where he went into some bushes at the bottom of the garden.

The neighbour’s partner had been very seriously injured and nearly died. His life was saved by the prompt action of a neighbour skilled in first aid and by the paramedics who were quickly on the scene.

When the alarm was raised, the defendant emerged from the next-door garden, carrying the knife and his balaclava. He was heard to say that he wanted his mother. He was also heard to say that he was not going to go down for this but would rather kill himself. And he was heard to say that he thought he had done something really wrong. After some coaxing, his father managed to persuade him to hand over the knife and balaclava, and the defendant went into his own kitchen. There he removed his top and waited quietly.

The dark clothes in which the defendant was dressed were different from what he had been wearing that day. On his own account to the police, which was not questioned, he had gone to bed in his boxer shorts. It follows that he had got up, selected dark clothing and an identity-concealing balaclava, armed himself with the knife, entered his neighbour’s house, advanced to her bedroom door and pushed it open. He had been able to enter the house because there was a spare set of keys to it kept in his kitchen; it was an example of the close friendship between the two households that his family looked after the next door house if the occupants were away. When he went into the kitchen after the event, he put the keys on the top of the refrigerator, not in their usual place, and tucked under some clothing which happened to be there.

When he was interviewed next day by the police, the defendant made no bones about having been responsible for the terrible injuries to his neighbour’s partner. He asserted that he had ‘blacked out’ and woken up standing outside the bedroom door and pushing it open. He had heard the scream, but had then ‘blacked out’ again and he said he had no further memory until he was in the bushes in the garden of the house on the other side of his own. Thus he said that he had no memory at all of the attack on the neighbour’s partner. His account to the police was the same as that to the court of trial, except that in the former interview he gave some unsatisfactory, and arguably untruthful, answers to the effect that he had not known of any keys and did not know how he had got in. His evidence was that he had no idea why he had done what he did.

The defendant had consumed a good deal of cannabis that day. His last cannabis cigarette had been at around 2200 that evening. He was a regular heavy user and his use worried his parents. He told the court that he had been using it since he was 13, and had on previous occasions found that it made him ‘paranoid’ and had caused him to black out. Three psychiatrists gave evidence. All were clear that the defendant suffered from no underlying or abiding mental illness or disorder, nor from a personality disorder. They each said, however, that it was a real possibility, if not most likely, that the defendant had been in the course of a “brief psychotic episode” induced or triggered by the cannabis, when he had done what he did. If so, they all thought that it was an unusually brief episode, but they were unable to offer any other psychiatric explanation for what he did. They also said that it was possible that he had been acting out the role of a character in the video game. One of them said that this may have taken (and in his view probably did take) the form of a delusion that he was such a character.

The defendant asked the judge to leave to the jury the issues of (a) insanity and (b) automatism. After argument, the judge declined to do so. The issue before the jury was, accordingly, whether the defendant had formed the intention to kill. After a long retirement, it concluded that he had. In this court, Mr Fisher QC renews his submission that the defendant was entitled to have the issues of insanity and automatism left for decision by the jury.

The argument for the availability of insanity runs as follows:

We should make it clear that in this appeal the question for us is not whether there really was a psychotic episode, as the doctors hypothesised, or not. That would be a question which could only be determined by a jury, if the evidence gave rise to possible insanity in law. The issue for us, as it was for the trial judge, is whether if the doctors were right, it would or might have been a case of insanity.

The judge ruled that insanity was not available on the evidence. He held that this was a case of voluntary intoxication, rather than insanity. The possible abnormality of mind could not be classified as insanity in law because it arose from an external and not an internal cause and was self induced. It was a temporary malfunctioning of the mind caused by the application to the body of illegal drugs, and that did not constitute insanity within the M’Naghten rules. He relied upon the principled limitation imposed by the law on the extent to which voluntary intoxication can be a defence to criminal charges. He also relied on the decision of this court in R v Quick [1973] QB 910.

Mr Fisher contends that this was not a case of intoxication but rather had passed to a recognised condition of mental illness, namely a psychotic episode, no matter how transient. He contends that insofar as Quick holds otherwise, it is wrong and inconsistent with the law’s recognition of the difference between intoxication or drunkenness simpliciter, on the one hand, and a disease of the mind induced by drunkenness on the other.

We agree that the law has long recognised the distinction which Mr Fisher identifies between intoxication and a disease of the mind induced by intoxicants. That distinction can be found in many cases. Mr Fisher rightly identifies one of the earliest in the direction of Stephen J to the jury in R v Davis (1881) 14 Cox CC 563, which was approved by Lord Birkenhead LC giving the sole speech in the famous intoxication case of Beard (1920) 14 Cr App R 160 at 194:

“But drunkenness is one thing and the diseases to which drunkenness leads are different things, and if a man by drunkenness brings on a state of disease which causes such a degree of madness, even for a time, as would have relieved him from responsibility if it had been caused in any other way, then he would not be criminally responsible”.

We also agree that, as Lord Birkenhead also made clear, insanity which is temporary is as much insanity as that which is long-lasting or permanent. Davis was a case of a defendant suffering (temporarily) from delirium tremens. That, self evidently, is not intoxication. It is, if anything, the opposite. It is a condition brought about by the protest of the brain and nerve receptors against the removal of intoxicants to which the body has become accustomed.

We do not doubt that the possible state of mind in which this defendant stabbed the man next door can properly be called a mental abnormality (or, in the nineteenth century language of M’Naghten, a defect of reason) which is recognised medically by psychiatrists. As a matter of fact, some care may need to be exercised in discovering exactly how the expressions “psychosis” or “psychotic” are used, which was not much explored in the evidence in this case, save to explain that the doctors were speaking of a state in which the mind becomes detached, to a greater or lesser extent, from reality. Generally, as we understand it, these expressions are more often encountered as descriptions of symptoms than as constituting a mental illness in their own right. The underlying cause of such symptoms may vary. Well understood ones certainly include schizophrenia and bi-polar disorder, which are no doubt mental disorders or illnesses. Another well known possible cause of psychotic symptoms is drug abuse, which is not a mental disorder. But with that caveat, we agree that to speak of a psychotic episode is no doubt to speak of a temporary abnormality of the brain or mind and thus of a defect of reason for the purposes of the M’Naghten rules.

However, the key thing to understand is that whether there is or is not a “disease of the mind” for the purpose of the M’Naghten rules is, and has to be, a question of law and not of medical usage. It makes excellent sense for medical people to classify a great variety of conditions as recognised medically. It enables statistics to be gathered, resources allocated, diagnoses understood with reasonable consistency and treatment to be advised: see the discussion in the slightly different context of diminished responsibility in R v Dowds [2012] EWCA Crim 281. But the law has to cope with the synthesising of the law of insanity with the law of voluntary intoxication. The first calls for a special verdict of acquittal and very particular means of disposal. The latter is generally no defence at all, but may be relevant to whether the defendant formed a specific intention, if the offence in question is one which requires such: DPP v Majewski [1977] AC 443. In most, but not all, intoxication cases, the intoxication will be possibly relevant to a serious offence allegedly committed but will afford no defence to a lesser offence constituted by the same facts: for example causing grievous bodily harm with intent (s 18) and causing grievous bodily harm without such intent (s20), or of course murder and manslaughter. In the development of the common law, intoxication was historically regarded chiefly as an aggravation of offending, rather than as an excuse for it. For all the reasons explained in Majewski, the law refuses as a matter of policy to afford a general defence to an offender on the basis of his own voluntary intoxication. The pressing social reasons for maintaining this general policy of the law are certainly no less present in modern conditions of substance abuse than they were in the past.

The precise line between the law of voluntary intoxication and the law of insanity may, we do not doubt, be difficult to identify in some borderline cases. But the present case falls comfortably on the side of the line covered by voluntary intoxication. It matters not that the condition of the defendant as observed in the aftermath of his attack on the neighbour was not that of conventional intoxication, in the sense that he was not, for example, staggering or unable to speak clearly. If the doctors were right about his state of mind, his mind was to an extent detached from reality by the direct and acute effects on it of the ingestion of cannabis. Every intoxicated person has his mind affected, and to an extent disordered, by the direct and acute effects of the ingestion of intoxicants; all intoxication operates through the brain. Not infrequently it would be perfectly legitimate to say of a very drunken man that his mind had become detached from reality by the intoxication; that is obviously true, for example, of the drunken man who suffers delusions as a result of the drink, but the proposition is not limited to that case. In order to engage the law of insanity, it is not enough that there is an effect on the mind, or, in the language of the M’Naghtenrules, a ‘defect of reason’. There must also be what the law classifies as a disease of the mind. Direct acute effects on the mind of intoxicants, voluntarily taken, are not so classified. That is the distinction drawn by Stephen J in Davis and maintained ever since. In the slightly different legal context of diminished responsibility a similar distinction is recognised: see R v Wood [2008] EWCA Crim 1305. Mr Fisher’s superficially simple argument amounts to treating every ‘defect of reason’ as a ‘disease of the mind’, but that is not the law.

Quick was not a case of intoxication. The defendant was a nurse at a mental hospital accused of assaulting a patient. He was a diabetic. He claimed that he had done what he did involuntarily when in a hypoglycaemic state, that is to say when his blood sugar fell to the point that he lost any control of his actions. The fall in blood sugar was, if it had occurred, attributable to the insulin which he had taken for his condition, coupled with some alcohol and without enough food to keep the level correct. He claimed that the defence of automatism was available to him, that is to say complete loss of voluntary control of his actions. The trial judge ruled that if his case was truthful it was insanity in law rather than automatism, and withdrew automatism. The defendant was not prepared, in the then state of the law as to disposal after a special verdict of not guilty by reason of insanity, to advance the defence of insanity. Instead, he appealed on the grounds that the different defence of automatism which, if it succeeded, would result in outright acquittal, ought not to have been taken away by the judge. Thus he contended, in effect, that he was not insane if his case was true. This court agreed. Lawton LJ said this:

“Our task has been to decide what the law means now by the words ‘disease of the mind’. In our judgment the fundamental concept is of a malfunctioning of the mind caused by disease. A malfunctioning of the mind of transitory effect caused by the application to the body of some external factor such as violence, drugs, including anaesthetics, alcohol and hypnotic influences cannot fairly be said to be due to disease. Such malfunctioning, unlike that caused by a defect of reason from disease of the mind, will not always relieve an accused from criminal responsibility. A self-induced incapacity will not excuse – see R v Lipman [1970] 1 QB 152, nor will one which could have been reasonably foreseen as a result of either doing, or omitting to do something, as, for example, taking alcohol against medical advice after using certain prescribed drugs or failing to have regular meals while taking insulin. From time to time difficult borderline cases are likely to arise. When they do, the test suggested by the New Zealand Court of Appeal in Reg v Cottle [1958] NZLR 999, 1011, is likely to give the correct result, viz: can this mental condition be fairly regarded as amounting to or producing a defence of reason from a disease of the mind ?”

It will be seen that there are two stages to this decision. Lawton LJ held that since the cause of the defendant’s condition was the application of an external factor, this was not insanity. Secondly, he held that whilst it might be automatism, that condition did not amount to a defence if it was voluntarily self-induced or caused by doing, or omitting to do, something which the defendant ought to have foreseen might lead to the condition. Later in another hypoglycaemia case, R v Bailey (1983) 77 Cr App R 76, also a case of claimed automatism, this court applied that approach and observed (at 80) that mere failure to take food after an insulin injection may not necessarily involve the criticism that the consequences should have been foreseen. The appeal, however, failed because automatism was inconceivable on the facts. However, in saying what it did, the court drew a sharp distinction from voluntary intoxication. That is because voluntary intoxication with alcohol or illegal drugs is an a fortiori case. No further enquiry is needed into whether the consequences ought to have been foreseen. All of this is thus entirely consistent with the voluntary intoxication rule to which we have referred. Drugs or alcohol are an external factor. When voluntarily taken their acute effects are not treated by the law as a disease of the mind for the purposes of the M’Naghten rules. Such a case is governed by the law of voluntary intoxication.

It is well known that the distinction drawn in Quick between external factors inducing a condition of the mind and internal factors which can properly be described as a disease can give rise to apparently strange results at the margin. An often quoted example is illustrated by R v Hennessy (1989) Cr App R 10, where the diabetic claimed not low blood sugar but high blood sugar (not hypoglycaemia but hyperglycaemia) and the court concluded that the latter did arise internally and thus could give rise to what the law classifies as a disease of the mind. We observe however, three things. First, Hennessy approved and applied Quick, which represents the law binding on us. More importantly, second, the arguably unsatisfactory distinction between the two conditions which a diabetic may experience has nothing to do with the case of voluntary intoxication, which is a plain one because of the clear law as expounded in Majewski and elsewhere. Third, any future rationalisation of the law of insanity must take into account the rules for disposal, which concerned the court in Hennessy as they do in practice in many cases in the Crown Court.

We conclude that the judge was right to rule that the defence of insanity did not arise in Coley’s case.

Mr Fisher also submitted that the judge ought to have left automatism to the jury. Automatism, if it occurs, results in a complete acquittal on the grounds that the act was not that of the defendant at all. It has been variously described The essence of it is that the movements or actions of the defendant at the material time were wholly involuntary. The better expression is complete destruction of voluntary control: Watmore v Jenkins [1962] 2 QB 572 and Attorney-General’s Reference (No 2 of 1992) [1994] QB 91. Examples which have been given in the past include the driver attacked by a swarm of bees or the man under hypnosis. ‘Involuntary’ is not the same as ‘irrational’; indeed it needs sharply to be distinguished from it.

In the present case the doctors were asked several times whether the defendant was acting “consciously” when he did what he did. We understand the difficulties of selecting appropriate adverbs, but this one carries some risk of difficulty. He was plainly not unconscious, in the sense of comatose. But automatism does not require that, and if it did it would be even more exceptional than it undoubtedly is. On the other hand, his mind may well, if the doctors were right, have been affected by delusions or hallucinations and in that sense his detachment from reality might be described by some as an absence of conscious action. Such condition, however, clearly falls short of involuntary, as distinct from irrational, action. In the present case, however, the use of the word does not now obscure what the medical evidence amounted to. Here the doctors were at some pains to explain that the defendant would, despite their hypothesis of psychotic episode, have been capable of complex organised behaviour. It is plain that a person acting under a delusion may act in such a way, and clearly this defendant did. He must have made the decision to dress specifically for his intrusion next door, and to arm himself with his knife. He made the decision to find the keys and let himself in. That was not, as it seems to us, capable of being described as involuntary action, whether or not it was driven by delusion or other psychotic detachment from reality. That was, moreover, the plain tenor of the medical evidence. Dr Maganti said that if he was in a psychotic state, as he hypothesised, he was “conscious” and “in control of his body”. Dr Kennedy said that such a person would be “in voluntary control of his limbs and conscious in the sense that they are aware of what they do physically.” Dr Nimagadda came closest to asserting the possibility of lack of voluntary control when he expressed the view at one point that the defendant had assumed the role of the video game character and “it was not him acting…It was the person in character.” But a later answer clarified this. The doctor said “He is conscious in a way but it is conscious in the belief that he is a character. He does not have an awareness of what he is doing.” That is a description of irrational behaviour, with a deluded or disordered mind, but it is not a description of wholly involuntary action.

Quite apart from the fact that the evidence was of voluntary, if irrational action, the defence of automatism is not available to a defendant who has induced an acute state of involuntary behaviour by his own fault. This court so held in Quick, and approved the decision in Bailey and Hennessy. Whilst concluding in Quick that the question of automatism ought to have been left to the jury, the court held that it would be necessary for the jury to scrutinise the extent to which the defendant had brought upon himself whatever condition he may have been in. That was in the context of taking some alcohol, or insufficient food, when also taking insulin, which is an area where there is plainly scope for debate as to the responsibility of the defendant. But the voluntary consumption of intoxicants leading to an acute condition is the prime example of self-induced behaviour.

Furthermore, the judge in the present case held that since the jury must have before it the question whether the defendant had formed the intent to kill, or had not done so because of his state of mind, the possibility of automatism ran with that question and a separate direction upon automatism was simply an unnecessary complication. It would not have been so, of course, if there had been before the jury an alternative charge which was not of specific intent, such as causing grievous bodily harm contrary to section 20 Offences Against the Person Act 1861. But there was not. We are satisfied that automatism simply did not run in this case for the reasons already explained: (a) the defendant was clearly not acting wholly involuntarily and (b) he had induced his condition by voluntary intoxication. On the further part of the judge’s reasoning, like him we find it difficult to envisage a case of an involuntary intent. On the facts of this case we agree that the verdict of guilty involves a clear finding that despite his state of mind, the defendant intended to kill and thus that he acted voluntarily. We do not, however, think it safe to say that in every case in which automatism is indeed a possible and legitimate conclusion, it should be removed from the jury if they have a decision to make about specific intent. That may particularly be so if the jury is invited to infer intent from the action, which may be a very short-lived action; if the action might indeed have been involuntary, such inference would not be safe and the jury ought in such a case to confront the issue of involuntary automatism before it goes on to intent.

We should add that for the Crown Mr Price QC advanced as a supplementary argument a similar contention in the event that insanity was a viable defence which should have been left to the jury. The argument runs as follows. If the defendant did not know the nature and quality of his act, he cannot have intended to kill. Thus the decision that he did intend to kill resolves the issue of insanity against him in any event. We do not agree. We think it is possible to conceive of cases of an insane intention to kill. It is certainly possible to imagine cases, as Mr Price properly recognised, of an insane belief that one’s act is not wrong. But in any event, if there is a genuine possibility of a verdict of not guilty by reason of insanity, it is our view that this should not be withdrawn from the jury simply on the basis that its answer on the issue of specific intent may be closely related to its answer on insanity. Quite apart from anything else, to withdraw from the jury an issue of insanity which it has heard debated is capable of colouring its decision on intent.

For these reasons Coley’s appeal against conviction must be dismissed.

The judge passed a determinate sentence of 18 years’ custody. His reasons for doing so demonstrate that he had approached the task with considerable care and in a sequential and logical manner. We agree with him that this was a grave example of attempted murder. It was in no sense a spontaneous attack, but was planned, albeit not before that night. It was characterised by wearing a disguise, taking a fearsome knife, and invading the home of the neighbour in the middle of the night when everyone was asleep. The injuries caused were very serious, and the judge could see that the injured man was painfully continuing to live the dreadful experience. Against that were to be set the judge’s finding that the defendant did not represent a continuing danger, that he was of otherwise entirely good character and that any violence or aggression at all was wholly out of character. It was, the judge concluded, an isolated, motiveless and inexplicable crime, apparently triggered by the use of drugs and influenced by the very violent video game.

The judge was rightly guided by the SGC guidelines relating to attempted murder. He referred to the fact that they pre-dated the addition to Schedule 21 of the Criminal Justice Act 2003 of paragraph 5A dealing with knife murders. As the guidelines demonstrate, the sentence for attempted murder must bear sensible relation to the gradations of the full offence as established in statute by that schedule (although those sentences are of course a great deal longer) and will normally then vary according to the gravity of injury inflicted. There can be no complaint about the judge’s conclusion that for an adult this offence would have fallen into the general range between 17 and 25 years, probably towards the upper part of it because of the taking of the knife, which for the full offence, in the case of an adult, now attracts its own starting point under paragraph 5A. The judge’s reasoning appears to have taken the adult range, lifted it for the knife factor, and then reduced it for age. However, if the very broad relationship to the much longer sentences for the completed offence of murder is, correctly, to be a factor of significance, it needs to be noted that Schedule 21 provides for a starting point for a defendant under 18 of a 12 year minimum term, and that that is not altered by paragraph 5A. As this court has made clear on many occasions, the right sentence for an offence does not suddenly leap from one point to another on the defendant’s eighteenth birthday. The adjustment for age is necessarily flexible. We are however persuaded that a term of 18 years did not in the end make sufficient adjustment for the youth and otherwise impeccable character of the defendant, or for what would have been the statutory starting point at his age for the completed offence. A sentence approaching twenty years custody is a very long one indeed for a young man of 17-18 and will find him a much different person by the time it is over. We think that the appropriate sentence for this young man for the appalling offence he committed is one of 15 years youth custody, and we substitute that sentence. To that limited extent, the appeal against sentence is allowed.

McGhee was convicted of wounding with intent. He had sought to advance automatism at trial but the judge had ruled that it was not available. His appeal challenges that ruling. Insanity was not in question. Voluntary intoxication was, but of course the offence was one of specific intent so that he was able to ask the jury to say that his intoxication meant that he had not formed the necessary intent.

The defendant’s medical background, although not his violence on the night in question, can only inspire sympathy. He has the misfortune to suffer from a grave form of tinnitus. Tinnitus is a medical condition caused by damage to the auditory pathway which results in persistent and permanent internal ringing in the ears. There is no cure for this disease, the debilitating effect of which is frequently compounded by the inability to sleep. Therapy is confined to palliatives, designed to alleviate the symptoms and in particular to assist with sleep. Since 2006, the defendant has suffered from profound hearing loss, and a particularly gross form of tinnitus, medically described as “severe-catastrophic”. To assist with his sleep, in 2009 he was prescribed the benzodiazepine tranquilliser temazepam for use at night. Because of his tendency to take more than prescribed, his girlfriend supervised his dose. However, the defendant frequently found that he still could not sleep, and, although temazepam is accompanied by a warning not to drink, he resorted to using alcohol as well. In his distress, he also sometimes physically hit his head on the wall with a view to rendering himself unconscious.

On the evening of 29 October 2010, he took his prescribed nightly dose of temazepam, supervised by his girlfriend. Although it is not clear precisely when, he also drank a good deal.

Just before 4am the following morning, he went to a 24-hour off-licence in Finsbury Park, which was manned by the manager (Mr Shah) and his assistant (Mr Jabarkhel). The events inside that shop were captured on closed circuit television: the footage was available in the Crown Court, and we too have viewed it. It is clear from that that the defendant was intoxicated and, indeed, he has always accepted that he was.

He entered the shop, and purchased soft drinks and a bottle of spirits. He selected the soft drinks from a shelf, and put them on the counter. The alcohol was behind the counter, and he selected both the type (whisky) and the brand, which he specifically identified to the shop assistant who served him. Initially, the defendant was polite, and made amiable small talk with Mr Jabarkhel, whilst Mr Shah served other customers, although Mr Jabarkehel did not really want to engage with him because of his obvious intoxication. The footage shows the defendant slouching over the counter, and talking for many minutes. He took a chocolate bar, which he ate and for which he did not pay - but Mr Jabarkhel did not seek to challenge him about that, probably because of his condition.

After some minutes, the defendant started to make abusive comments to Mr Jabarkhel, particularly about his ethnic background and religion. To avoid confrontation, the shop assistant went to smoke a cigarette outside. The defendant also left the shop with his purchases; and, when Mr Jabarkhel tried to go back in, he was prevented by the defendant who stood in the doorway and elbowed him. The defendant invited him to fight, and started to shadow box towards Mr Jabarkhel’s face. No charges arose from that early part of the events.

However, another customer, Towan Williams, arrived at the shop and saw the defendant swearing and pushing Mr Jabarkhel. When Mr Williams tried to enter the shop, the defendant became aggressive and pushed him against the wall. Mr Williams tried to back away, and punched the defendant in self-defence. An altercation then took place, with the defendant and Mr Williams wrestling each other on the floor, out of the camera’s eye. However, the defendant accepted that, at one stage whilst on the floor, he pressed his fingers into Mr Williams’ eye, causing him pain and blurred vision for two days, which required medical treatment.

The incident with Mr Williams over, the defendant then walked off; but returned to the shop about 10-15 minutes later armed with a large kitchen knife and with a t-shirt attached to his head like a bizarre head-dress. The CCTV footage shows him entering the shop, and immediately trying to attack both Mr Jabarkhel and Mr Shah, who are seen trying to escape. Mr Shah jumped over the counter, and ran to the exit and away, pursued briefly by the defendant. Mr Jabarkhel was not so fortunate: the defendant caught him and stabbed him in the arm with the knife, causing a wound through his clothes, before running from the shop leaving the knife on the counter. The defendant is then seen entering and leaving the shop once or twice more, before leaving the scene altogether.

The defendant could not remember any of those events. They resulted in him being charged with assault occasioning actual bodily harm to Mr Williams, and wounding Mr Jabarkhel with intent to cause grievous bodily harm.

At the initial hearing, on 12 April 2011 the defendant pleaded guilty to the assault on Mr Williams, and indicated that, in respect of Mr Jabarkhel, he would be willing to plead guilty to causing grievous bodily harm contrary to section 20. However, he said that he was unwilling to plead to the section 18 offence, because, at the relevant time, he was in a state of automatism such that he was unable to form the required specific intent to cause grievous bodily harm. The Crown indicated that those pleas would be unacceptable. The court giving him permission to do so, the defendant then formally withdrew his guilty plea to the assault on Mr Williams, and the matter was set down for trial before Mr Recorder Morris-Cole on 21 October 2011.

The Recorder made two rulings that day. First, he rejected the defence submission that the defendant had no choice but to take alcohol to alleviate the symptoms of his medical condition: he ruled that his intoxication was voluntary. He also ruled that the defence had not satisfied the evidential burden upon them necessary to leave automatism to the jury. He found that, in all of the circumstances, there was no evidence that, at the time he stabbed Mr Jabarkhel, the defendant was suffering from automatism in the sense of a “total destruction of voluntary control”.

Following those rulings, the trial proceeded on the basis of a three count indictment. The Defendant accepted that he was guilty of both the section 47 offence against Mr Williams and the section 20 offence against Mr Jabarkhel. With regard to the section 18 offence, the Recorder directed the jury that, although the defence of automatism was not open to the Defendant and that an intoxicated intent was still an intent, nevertheless they could only convict the Defendant if they were sure that, at the time of the stabbing, he intended to cause Mr Jabarkhel grievous bodily harm. The jury found the Defendant guilty of all three offences.

Mr Winship’s careful submissions on behalf of McGhee, for which we are grateful, are these.

We need not repeat what is said above at [22] in the case of Coley. Automatism involves wholly involuntary action as distinct from irrational action. That is how, correctly, the question was approached in McGhee’s case by counsel on both sides and by the recorder. There were two experts called. Dr Falkowski, a consultant psychiatrist, gave evidence that in his view the defendant was clearly aware of his actions and in control of them, even if his judgment was impaired. If that was right, no question of automatism arose. The possibility that automatism ought to have been left to the jury arose from the evidence of Professor Birch, who is not a medical doctor but a consultant pharmacologist and a senior professor of biomedical science specialising in psychopharmacology and with a particular expertise in the action of drugs on the brain.

Professor Birch gave evidence on a number of topics. One part of his evidence explained that the temazepam and alcohol might produce what he described as “paradoxical effects”. By that he meant that although temazepam is designed to inhibit anxiety and thus to keep the taker calm, because it acts as an inhibitor of the processes of the brain, it may also inhibit the control which would otherwise restrain the taker from anti-social action and thus end up having the opposite (“paradoxical”) effect. In his words, the drug, especially if potentiated by alcohol, can reduce social self-restraint and lead to disinhibited behaviour, such as aggression or over-reaction to trivial provocation. He used the analogy of the drug peeling away the layers of an onion which consisted of the taker’s normal social self-restraint and exposing the core of aggressive reaction. He gave evidence that serious violence can occur in this way as a result of the combination of alcohol and benzodiazepines.

Professor Birch was additionally asked whether it was possible that, in the later part of this incident (ie in the attack on Mr Jabarkhel), or for that matter earlier, the defendant might have been acting “without any form of conscious control”. His answer was:

“I do not think it was organised conscious control. It may be a small amount of conscious control. I suspect probably not.”

A little later came this exchange:

“Q: You said that conscious control may not be what we are seeing here ?

A: No that is right….

Yes, I mean, bits of it look as though it could be a sort of automatic kind of behaviour.”

Mr Winship focuses on those answers as providing an evidential base requiring the question of automatism to be left to the jury.

Once, however, these answers are put in the context of Professor Birch’s evidence it is completely clear that he was speaking of disinhibited behaviour rather than of behaviour over which the defendant had no voluntary control whatever. The first answer cited above was followed immediately by this amplification:

“I think he was probably reacting spontaneously to something which…I was talking about, the core of the onion, the aggressive behaviour which came out as the result of being disinhibited.

I do not feel that conscious control is really the way of describing it. I mean, conscious control means that you have actually thought about it and you have decided that this is the course of action to take.”

The second passage cited above was followed immediately by this:

“Q: So parts of it look as though they could be automatic….Which parts of what you see look as though they could be ?

A: Well, his response to whatever was the provocation in the fight, and perhaps coming back again.”

There were further questions and answers to like effect. Thus the evidence made it clear that what Professor Birch was speaking about were the paradoxically disinhibiting effects of temazepam plus alcohol and the manner in which that can lead to a person behaving in an aggressive fashion from which he would normally be restrained in the absence of the substances taken. Disinhibition is exactly not automatism. Moreover, one has only to look at the camera footage and the defendant’s clearly voluntary behaviour in the shop and outside it over quite an extended period for it to be apparent that automatism was simply not in question. Intention was, on the count laid under section 18, but not automatism. The recorder was quite right.

Additionally, as we have explained above at [24], it is very likely that even if the defendant had been in a state of automatism, the defence would have failed on the grounds that he induced it through his voluntary fault. He had voluntarily drunk himself into what was agreed on all sides to be a state of current intoxication. To the extent that his intoxication was worsened by the combination of temazepam with alcohol, he was well aware of the dangers of taking them together. That he was in distress at the very unpleasant tinnitus would not alter that. This, however, did not arise.

In those circumstances, this appeal against conviction must be dismissed.

The charge in this case was the aggravated form of arson, contrary to section 1(2) of the Criminal Damage Act 1971. Specifically, the defendant was charged with starting a fire, being reckless as to whether the life of another would be endangered. He had started a fire in his own house. The persons whose lives were potentially in danger were his neighbours in the next door semi-detached house.

The defendant was born on the 5^th July 1968, and so was 42 years old at the material time. He owned and occupied a semi-detached house in Newport, South Wales. He was employed by the local council, a steady job which he had held for about 25 years without complaint about him.

From about 2005, and probably for reasons related to the illness and death of his mother, the defendant suffered from clinical depression. From about the same time he began to abuse alcohol. He appears to have developed a pattern of drinking heavily during a period of leave, and then abruptly ceasing to drink prior to returning to work. The sudden cessation of his excessive drinking had in the past resulted in episodes of alcohol psychosis or alcohol-induced hallucinosis. In October 2005, December 2005, and again in December 2010, he had to be admitted to hospital following such episodes; on the last of those occasions he was detained under section 2 of the Mental Health Act. A report from a consultant forensic psychiatrist described him as ordinarily shy and amenable, but displaying uncharacteristic outbursts of aggression and paranoia when affected by such conditions.

It was common ground in the Crown Court that the defendant was suffering such an episode on the Friday 29^th April 2011. As before, it had been brought about by heavy binge drinking throughout most of the days of the preceding week’s Easter leave, and then by stopping suddenly on Saturday 23^rd or Sunday 24^th with a view to going back to work on Monday 25^th. He did return to work, but felt ill on Tuesday 26^th and could not continue. Over the next two days, his family had become concerned about his mental health and had sought medical assistance. He was complaining of hearing voices and his brother described him as ‘talking into space’. On the morning of Friday 29^th April 2011, he was seen to be apparently cutting the grass in his garden using an electric lawnmower which was not plugged in. At some stage, the Defendant removed from his house, and stored in his garden shed, a box containing his passport and important documents relating to his mortgage, house insurance and pension. At about 1130 he was seen taking a petrol can into the house: he was later to say that he did so because voices in his head were telling him to burn the house down. Thereafter, in what was admittedly a deliberate act, he poured petrol from the can onto the carpet in at least two places in his living room. Vapour from the petrol was ignited, and a fire began. The Defendant later told his brother: “I turned the gas cooker and fire on, threw the petrol, went to light the fire, it blew me against the wall”. It seems he then panicked and went upstairs.

By good fortune, neighbours saw the fire and intervened before much damage had been caused. They alerted the family who occupied the adjoining semi-detached house. The Defendant was seen at an upstairs window, and a ladder was used to bring him down. He was treated at the local hospital for the effects of smoke inhalation, and was then transferred to a psychiatric hospital. No one else was injured, and the only damage was to the Defendant’s own home.

It was accepted by the Defendant that he had set the fire. It was further accepted that he knew the nature and quality of his act, and that what he did was wrong. But for the fact that the only damage was to his own property, he would have pleaded guilty to simple arson. However, he contended that he had given no thought at all as to any risk which his actions might pose to others, and on that basis denied that he was guilty of arson being reckless as to whether the life of another was endangered.

The facts were largely agreed. The judge ruled that the circumstances of the case brought it within the principle in DPP v Majewski [1977] AC 443. He held that the mental disorder from which the Defendant was suffering when he started the fire was a direct result of his voluntary consumption of alcohol. Since the offence was one involving recklessness, he would direct the jury that in considering whether the Defendant was reckless as to the risk to the neighbours, they “would have to look at this defendant as if he had not been drinking”. Following that ruling, the Defendant pleaded guilty. He accepted and accepts that but for the condition which he was in, he would have foreseen the risk to the neighbours.

Insanity does not arise in this case. The defendant was affected by a mental disorder, but he knew what he was doing and he knew that it was wrong. What he challenged was the assertion that he was reckless as to the endangering of the lives of the neighbours. He said that because of his mental condition, a risk to them had simply not occurred to him.

Since R v G [2004] 1 AC 1034; [2003] UKHL 50 a man is reckless as to a possible result where he is himself actually aware of the risk that it may occur: see Lord Bingham at [41]. That was not a case of voluntary intoxication; it was a case of 11- and 12-year old defendants who had not applied their minds to the risk. The former rule, set out by the House of Lords in DPP v Caldwell [1982] AC 341, was overruled in R v G. Caldwellhad been a case of voluntary intoxication. In R v G the House of Lords did not have to decide how its new (“subjective”) test of recklessness fell to be applied to a case of voluntary intoxication. Lord Bingham, however, adverted at [36] to the fact that it was unfortunate that the meaning of recklessness had arisen in Caldwellin the context of intoxication, and added that

“…one instinctively recoils from the notion that a defendant can escape the criminal consequences of his injurious conduct by drinking himself into a state where he is blind to the risk he is causing to others. In R v Caldwell it seems to have been assumed…that the risk would have been obvious to the defendant had he been sober.”

Lord Steyn, at [60], associated himself with those remarks. Lord Rodger, in the only other substantial speech, did not attempt to answer the question of voluntary intoxication, but drew attention to the view of Lord Lane CJ (obiter) in the earlier case of R v Stephenson [1979] QB 695 that intoxication represented an exception to the generally subjective test of recklessness. There remains some room for doubt as to whether the aggravated offence of reckless arson, charged in the present case, is governed by this principle or should be regarded as an offence of specific intent – see for example the (dissenting) opinion of Lord Edmund-Davies in Caldwell at 361D and the passing obiter reference in R v Heard [2008] QB 43 at [31], where, however, the offence under consideration (sexual assault) was one requiring not recklessness but (basic) intent. We see some force in the argument that voluntary intoxication ought not to be a defence to an offence involving recklessness, even subjective recklessness; it may fall for decision in a later case whether Lord Lane’s view in Stephenson correctly represents the law now that Caldwell recklessness has passed away.

It is, however, not necessary to resolve that issue in the present case. The crux of the judge’s decision was that this was a case of voluntary intoxication. We agree that it was a condition which was caused by past voluntary intoxication, but the defendant was not in any sense intoxicated at the time of the offence. He had stopped drinking some five or six days earlier. As in the old case of Davis (see [15] above), he was not suffering from a direct or acute reaction to the voluntary taking of intoxicants. His condition at the time of the offence was one of mental disorder with psychotic symptoms which included the hearing of voices and hallucinations. One of the doctors described the condition as follows:

“Alcoholic psychosis describes a cluster of different psychotic conditions relating to alcohol misuse. It includes delirium tremens and alcoholic hallucinosis…

The occurrence of agitation paranoia, auditory and visual hallucinations and confusion following sudden withdrawal of alcohol would point towards a diagnosis of delirium tremens. In certain circumstances there might be seizure associated with it. I note that Mr Harris has a history of seizures in the past….The complete resolution of symptoms is usually within a few days of stopping alcohol, with or without medication.”

The defendant’s condition was also complicated by clinical depression. He was in fact continuing to take both anti-depressant and anti-psychotic medication.

The argument for the Crown in this case is that the mental illness from which the defendant was suffering was brought on by his past voluntary drinking. Therefore, it is contended, it should be treated in the same way as if he were still drunk. We agree that there is scope for the argument that an illness caused by his own fault ought as a matter of policy to be treated in the same way as is drunkenness at the time of the offence. This would, however, represent a significant extension of DPP v Majewski and of the similar principle expounded in Quick, which likewise concerned a case where what was asserted was an acute condition (there of automatism) induced arguably by the defendant’s fault. A great many mental illnesses have their roots in culpable past misconduct of the sufferer: those attributable to many years of past drug or alcohol abuse are perhaps the most obvious, but there could be many other examples, such as perhaps a culpable failure to follow a recommended medical regime, or maybe the consequences of traumatic brain injury caused by one’s own drunken driving. Whether the Majewski approach ought to be extended to such cases may be a topic which might be addressed in the forthcoming work of the Law Commission on loss of capacity, and it should, no doubt, be the subject of proper public debate. But in the present state of the law, Majewski applies to offences committed by persons who are then voluntarily intoxicated but not to those who are suffering mental illness. This defendant was, it is clear, suffering from a condition of mental illness when he set fire to his own house. That it was not long-lasting does not mean that it was not a true illness. In our view he was entitled to have tried the question of whether, in the condition in which he was, he was actually aware of the risk which he created for his neighbours.

It follows that we grant Harris leave to appeal against conviction, allow the appeal and quash his conviction. We were helpfully addressed on the question of whether it is in the public interest to order a re-trial. The defendant had been receiving treatment voluntarily for some time before his trial and was making real progress. He had reduced his drinking to modest proportions and was taking the anti-depressant and anti-psychotic medicine prescribed. The judge made, entirely appropriately, a community order with a condition of mental health treatment designed to ensure that the same regime continued. The offence is now a little short of two years ago and he has behaved without complaint ever since. Since the facts are undisputed and will remain on record in the event of any further misbehaviour, we do not consider that the public interest calls for a re-trial.

We return to the point of departure at [1] above. These cases do not provide an occasion for a general review of the law of insanity, automatism or intoxication, but are illustrations of its application. A general review by the Law Commission into these related areas, such as is currently in train, will be welcome. It is we think essential that any such review addresses also the vital question of disposal following the differing verdicts which may ensue.