The Willed Trance: Volition, Voluntariness and Hypnotised Defendants

Traditionally, jurists have distinguished between voluntary/involuntary behaviour via the theory of volition. Though perceived as the conventional approach, this paper argues that the volitional understanding of voluntariness is an inadequate instrument for assessing complex behaviours which seemingly portray a striking level of intelligence and purposiveness on the part of the accused. In particular, the phenomenon known as hypnotically-induced behaviour, which forms the focus of this paper, is one such troublesome case. To this end, the version of the volitional theory most staunchly advocated by Professor Michael Moore is singled out for scrutiny, due to his strong sentiments supporting the application of his philosophy to these aforementioned behaviours. In contrast to Moore, this paper suggests that the position most recently proposed by the Law Commission of England and Wales within their discussion paper on the defences of insanity and automatism is to be preferred. Specifically, the Commission recommend substituting the theory of volition for that of ‘control’ as a means for assessing the voluntariness of any given behaviour. This paper submits that a theory of control has two major advantages over the traditional theory of volitionalism. First, the possession/absence of control more accurately reflects the contemporary system of criminal law in England and Wales. Second, a theory of control is more conceptually defensible as an explanation for why behaviours performed under hypnosis are typically perceived as involuntary.


Introduction
This paper offers a critical examination of the fundamental principle that criminal responsibility requires voluntariness of action. Specifically, it centres on the traditional meaning of a voluntary act as 'willed bodily movement', 1 and attempts to apply this theory to instances of complex, Purposive conduct, such as that committed under hypnosis. Given that the commission of a criminal act is 'rarely accompanied by a complete lack of consciousness and bodily control', hypnotism represents an exemplar of how difficult judgments need to be made regarding the degree of loss of consciousness and attention at the specific time. 2 Contrary to the position of Michael Moore 3 -the leading proponent of the traditional theory of voluntary action-it is contended that volitionalism is unable to adequately defend the proposition that defendants behaving under hypnotic influences act involuntarily. This verdict is reached by drawing upon the work of scholars who have directly challenged the application of Moore's volitional theory to a number of problematic cases. 4 These include somnambulism, behaviour under or due to hypnosis, and that which is more generally characterised as conduct performed when the actor is dissociated.
As a result, this paper examines the proposals set forth by the Law Commission of England and Wales. 5 The Commission suggest that the concept of volition should not be the decisive factor in determining whether the conduct of an accused was performed voluntarily. 6 Rather, in framing their reformed defence of involuntary conduct (otherwise known as automatism) the Commission emphasise the notion of 'control' and its relationship to voluntariness. 7 The theories of 'control' and 'volition' are subsequently distinguished, before the former most position is supported as a more accurate representation of current criminal law doctrine. If hypnotism operates as we commonly believe it does, a lack of control, as opposed to a lack of volition, is also a more accurate description of the defect present in these individuals. The paper concludes, however, by briefly acknowledging the numerous practical difficulties inherent in allowing a defendant to successfully plead automatism when he/she has allegedly been hypnotically induced to commit the crime in question. As such, while a hypnotism defence may be theoretically plausible, it is highly unlikely that society will encounter a defendant receiving a full acquittal through its practical usage any time soon.

Automatism and the Voluntary Act Requirement
Eminent legal scholars often state that the requirement of actus reus, otherwise known as a voluntary act, is the minimum condition necessary for the imposition of criminal culpability. 8 Together with mens rea, the actus reus is derived from the familiar slogan entitled actus non facit reum nisi mens sit rea, which means 'an act does not make a person guilty unless their mind is also guilty'. 9 This conveys that criminal culpability requires blameworthiness of both mind and behaviour. In other words, the voluntary act must generally be accompanied by either intention, knowledge, recklessness, etc.
Given this initial requirement that the offence for which the accused is charged was engaged in voluntarily, it is unsurprising that an involuntary act is thus perceived as one that is undeserving of blame, punishment or societal condemnation. 10 Indeed, to inflict punishment for involuntary actions would fail to fulfil the philosophical tenets of both utilitarians and retributivists-the two leading philosophies of punishment in criminal law. While it is accepted that, in theory, utilitarians 11 may punish an accused in order to achieve a general deterrent effect regardless of individual voluntariness, 12 it is uncontroversial to say that there is little personal deterrence value to punishing one who is not acting voluntarily. 13 Furthermore, for retributivists 14 it is simply unjust to punish someone who has not freely chosen to violate legal norms. 15 As Wilson purports, '[t]he state does not punish involuntary conduct because doing so is inappropriate, just as it is inappropriate to hold the runaway train responsible for colliding with the school bus, an infant for taking sweets from a sweet shop, or a dog for biting the postman'. 16 Due to these concerns, the law of England and Wales has developed the common law defence of automatism, which negates the actus reus of a crime where the agent behaved involuntarily. 17 Strictly speaking, automatism negates an element of the prima facie case, 18 and is therefore not a 'defence' per se. 19 Rather, it is presented as conceptually distinct from the ordinary legal defences. 20 With automatism, the accused is not even the author of the criminal wrong. 21 Instead, he is a person to whom the criminal wrong happened. The basis of the claim to avoid censure is not, as with standard excuses, 'Don't blame me. I'm only human', but rather 'Don't blame me. It was not me who did it'. 22 Consequently, non-insane automatism is typically defined as requiring the following three criteria: First, the accused behaved in a completely involuntary manner. 23 Second, the involuntariness must have resulted from an external factor (such as concussion or hypnotic influences) 24 ; or, potentially, an internal factor, (such as a leg spasm). Third, there should be no prior fault on the defendant's part. 25 The classic example of this lattermost requirement would be motoring offences or sleepwalking acts triggered by voluntary alcohol intake preceding the episode of involuntariness. 26  Before proceeding, it should be stressed that while theorists often refer to a voluntary 'act' requirement, it is undoubtedly false that acts are the only legitimate objects of criminalisation. 28 Many jurisdictions, including England and Wales, recognise various offences in which criminal liability need not be based on the performance of an act, such as omissions and possession offences. Some of these crimes are notoriously controversial, but there is surely no good reason to suppose that persons can only be responsible for physical acts. 29 Insofar as criminal liability should track moral responsibility, restricting the imposition of liability solely to acts is unwarranted, for we are certainly morally responsible for a wider array of events than just physical acts. 30 Therefore, it is preferable to adopt Botterell's approach and refer instead to the existence of a voluntary conduct principle, which is broad enough to cover a wide range of behaviours. 31 Nonetheless, because this paper focuses largely on the voluntariness of particular actions, the terms 'act', 'action' and 'conduct' are used interchangeably.

The Traditional Meaning of Voluntariness
It is easy to explain why involuntary behaviour should not attract punishment. It is far more difficult to explain just what exactly is missing in these sorts of cases. Despite the lack of consensus regarding the precise definition of a 'voluntary act', 32 Michael Moore within his book entitled Act and Crime 33 offers a re-articulation of the traditional view which perceives the concept of volition (or the 'will') as the key to voluntariness. 34 Specifically, voluntary action is a movement of the body which follows from the will or 'a volitionally caused bodily movement'. 35 'Volitions' are functional mental states which execute a person's desires, beliefs and more general intentions by causing the basic bodily movements that satisfy these more general states. 36 Arguably, this also covers so-called instinctive reactions, such as the 'sudden movement of a tennis player retrieving a difficult shot; not accompanied by conscious planning, but certainly not involuntary'. 37 By way of example, Wilson helpfully summarises the volitional theory as follows: In the typical account, the desire to act in a particular way is thought to translate itself via the will into the bodily movements that comprise the action. So, a child might conceive a desire to throw a cup, and then send mental orders to the muscles in his shoulder, arm, and fingers to contract and expand all in their correct sequence so that the projectile is properly launched. 38 Therefore, it is because of this effort of will that action can be designated as voluntary and which justifies censuring the person. 39 These foregoing statements reflect a traditional view in the philosophical literature that treats 'mind' and 'body' as two distinct entities, with the mind operating like a kind of puppet master that resides within and 'pulls the strings' of the body. 40 In other words, action, like any other event on this view, must be caused by something. For Moore, action is voluntary when it is caused by an exertion of will. 41 This theory further helps to explain why the law distinguishes between 'X's arm went up' and 'X raised his arm'. 42 Without more, the former does not qualify as voluntary since it may simply have been caused by external force. The latter, however, with its implicit existence of volition (the actor's 'undertaking' or 'trying') appears to suffice. 43 This theory is equally applicable to other evident cases of involuntariness such as spasms, and the knee-jerk reaction caused by a doctor tapping the patellar tendon with a reflex instrument. In each case, the will does not precede the movements in question. 44 Thus, the behaviour exhibited is not caused by the actor's decision-making processes. In sum, these types of behaviour are more accurately described as events that happen to an actor, as opposed to actions he brings about. 45 Unfortunately, this theory begins to face difficulties when applied to more complex phenomena which do not neatly fit on either side of the voluntary/involuntary continuum; most notably, dissociative states. 46 The psychological notion of dissociation poses vexing problems for understanding human action and its relation to culpability. 47 Dissociation is defined in the Diagnostic and Statistical Manual of Mental Disorders as 'a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment'. 48 Similarly, Steinberg opines that 'dissociation is a fragmentation of consciousness'. 49 These are very broad and rather vague definitions, but examples include somnambulism, 50 fugues, 51 and (for some theorists, like Moore) behaviour performed under hypnosis. 52 To be clear, unconsciousness is not essential for involuntary action to exist. 53 Rather, unconsciousness merely facilitates a finding of involuntariness. Each of the examples asserted above can arguably invoke a severe disturbance or alteration of self-consciousness, resulting in a sense of detachment from the usual perception of reality. 54 If these states make it unduly difficult for the dissociated agent to comply with legal norms, then fairness requires that they are not attributed with responsibility. 55 However, uncertainties arise because these conditions share little resemblance to reflexes or spasms; the dissociated agent engages in conduct demonstrating both accurate understanding of the environment and goaldirectedness, suggesting that the bodily movements involved are intentional, voluntary actions. 56 Although this paper focuses on the difficulties posed by hypnosis, many of the arguments presented are equally applicable to other states of dissociation, including those asserted above.

Hypnosis and the Law
Hypnotism has baffled a variety of disciplines for hundreds of years, leading to numerous suppositions about its existence and effect. 57 Although the profusion never settled into a single cogent theory, the 21st century has embraced hypnotism with open arms. 58 Today, the professions of medicine, 59 dentistry, 60 psychology 61 and psychiatry 62 all utilise hypnosis. In sharp contrast, law is undoubtedly the most reluctant profession to embrace hypnotic phenomena. 63 This unwillingness likely stems from the uncertainty surrounding this condition. 64 Despite modern technological, scientific and medical advances, the nature of hypnotism and its effects remain shrouded in mystery. Indeed, developments in the area of criminal defence law are almost inexistent. In England and Wales, the sole authority to address hypnosis as a potential criminal defence is a statement of obiter dictum in Quick, 65 which suggests that behaviour performed under 'hypnotic influences' 66 should be regarded as constituting automatism.
A second, and perhaps more well-known example is the American Model Penal Code which explicitly lists 'conduct during hypnosis or due to hypnotic suggestion' as an example of involuntary behaviour. 67 Virtually identical provisions have also been enacted into the criminal codes of Montana, 68 New York 69 and Kentucky. 70 Thus, in Anglo-American theory at least, hypnosis could suffice for a defence of automatism. In practice, however, there are no recorded cases where hypnosis has formed the basis of a successful defence to a criminal charge. 71 As Garvey succinctly states, 'hypnosis today is more about stage entertainment and smoking cessation. One searches in vain for real crimes committed under hypnosis'. 72

The State versus Non-State Debate
While no theorist has wholly agreed with another, modern hypnotism theories can be roughly divided into two main camps: the 'state' camp and the 'non-state' camp. 73 The former maintains that hypnosis renders the subject's action involuntary; the individual suffers an impaired state of consciousness in which they are mere pawns at the mercy of the omnipotent hypnotist. 74 Conversely, the 'non-state' camp contend that hypnotised subjects act voluntarily; they merely enter a state of relaxation and suggestiveness. 75 The view of Michael Moore falls under the 'state' camp position. For Moore, behaviours performed under hypnotic suggestion wholly lack the component of volition, which he deems as essential to voluntary action. Because 'volition' and 'voluntary' are synonyms in his usage, 76 hypnotised defendants do not act voluntarily, and should thus satisfy the automatism defence.
It is submitted that Moore's approach is heavily influenced by the traditional 'trance' theory of hypnosis. The most common scientific definition of trance theory includes three primary elements: (1) the subject is unaware of or insensitive to his environment, (2) the subject has unusual experiences, such as hallucinations or false beliefs, and (3) his actions are performed unconsciously and involuntarily. 77 Evidently, both Moore and the trance model perceive the hypnotised subject as being incapable of performing voluntary/volitional activity. As such, the remainder of this paper examines hypnotism through the lens of the volitional thesis in order to decipher whether hypnotic behaviour should constitute automatistic conduct. 78

Hard Cases for the Volitional Thesis: Hypnosis and Dissociative States
Professors Bernard Williams and Stephen Morse examine Moore's postulations on this point by contending that the volitional theory is incapable of defending the proposition that dissociated defendants act involuntarily. 79 The first argument advanced by Williams directly addresses the behavioural facts surrounding cases like hypnosis and somnambulism. He submits that these behaviours are so responsive to the environment and so seemingly intelligent that they tempt one to conclude they must be voluntary actions persons perform. 80 Thus, when referring to a scenario of somnambulistic behaviour, Williams states: There is no doubt that Lady Macbeth has picked up the light, found the door, undone its bolt, and carefully come down the stairs. Moreover, it is not a matter of a mechanically determined routine which merely looks as though it were responsive to perceptual cues; some somnambulists will walk around pieces of furniture that are not in their normal place. So why should we say that these movements only look like actions? 81 The preceding statements are equally applicable to behaviour performed under hypnosis. It is submitted that there are two ways to interpret this submission. 82 First, Williams could be suggesting that certain behavioural patterns are, themselves, sufficient to classify as actions. 83 In other words, if behaviour looks like voluntary action, then that itself should be the decisive factor in determining whether it is voluntary action, and thus, whether criminal liability should attach. This sounds identical to the philosophy of Gilbert Ryle 84 who belonged to the school of psychological thought known as 'Behaviourism'.  Green 1959). 78. Note that this paper will not discuss whether hypnotically-induced behaviour actually is/is not voluntary. Rather, this paper will assume that hypnosis is a real phenomenon, and that it is possible for a defendant to behave involuntarily as a result of its effects. The specific question guiding this thesis is why hypnotic behaviour is involuntary; ie is it because hypnotised defendants lack volition or, alternatively, control? See below. and assessed. 86 Consequently, behaviourists outright reject the view that internal mental states (like volitions) can influence behaviour. 87 Against the internal cause view of, say, mental disease, Ryle urged that we would not be able to tell on such a view whether 'the inner lives of persons who are classed as idiots or lunatics are as rational as those of anyone else. Perhaps only their overt behaviour is disappointing'. 88 In response to this argument, Moore correctly purports that behaviourism has fallen into disfavour ever since the scientific interest to understand the various complexities of the human mind began in the 1970s. 89 Indeed, a large body of scientific literature categorically proves that internal mental processes can, in fact, affect our behaviour. 90 Thus, in the present era, we now have good reason to think that human action may well have an internal, hidden nature so that surface indicators (like behaviour) may sometimes be misleading. 91 It is submitted that Moore's theory of action helpfully attempts to explain this hidden nature in terms of volitions. To summarise, the contention that somnambulistic and like behaviours look very much like voluntary actions does not thereby entail that they are voluntary actions-nor has Williams successfully falsified Moore's volitional theory. 92 A more convincing way to construe Williams's approach is to contend that such intelligent and environmentally-responsive behaviours could only be possible if they are willed. 93 Thus, he would be directly using Moore's own theory in order to argue that hypnotically induced behaviours are, in fact, voluntarily committed. This is certainly not an illogical approach. Unlike reflexes or spasms, the bodily movements of the dissociated agent are not random; they express the agent's underlying desires, beliefs and more general intentions. 94 For instance, the hypnotised subject performs the specific wants and desires instilled in them by the hypnotist, and the sleepwalker can walk downstairs, engage in the intricate task of driving, or even kill; depending upon their individual desire/belief set. Indeed, it is surely implausible to suggest that the harms done are completely random: they must express the agent's underlying desires and beliefs. In each case, some kind of mental state is effectively executing these desires and beliefs by causing the bodily movements that satisfy them. So why is this mental state not a volition? 95 While these preceding statements are applicable to hypnosis, current neuroscientific understanding would seem to doubt the suggestion that sleepwalkers truly act upon repressed desires. In fact, it is now usually postulated that their complex actions arise not from mental states, but rather, from the 'central pattern generators' located in the spinal cord. 96 In regard to hypnosis, however, Moore has three responses to the contention that hypnotically-influenced actions are the product of volitions, none of which are free from scrutiny.

Moore's First Response-'Dissociated Movements Are Caused by the Wrong Kinds of Volitions'
In the first of Moore's rebuttals, he attempts to explain why these behaviours look so much like voluntary actions, even though he believes they are not. Specifically, he asserts that hypnotised persons lack true, 'full-blooded' volitions, but instead have sub personal, 'volition-like states that execute some of our 86 background states of desire, belief and general intention'. 97 In other words, he suggests that hypnotised subjects and their ilk do not voluntarily act because their complex patterns of behaviour are caused by sub personal, less-than-full volitions-the kind no human being has access to, and should not be held responsible for. 98 Essentially, all that Moore is saying here is that the difference between ordinary action and dissociated action is that in the former case the behaviour is caused by a 'true' or 'proper' volition, whereas in the latter case the behaviour is caused merely by a 'sub personal' or 'fake' volition. For Moore, only the former should be classed as voluntary action. In turn, this provides an explanation for why dissociated movements look so much like everyday actions; not only do they share some of the same beliefs and desires as when we are ordinarily awake, but these desires are then executed by a counterfeit volition which very much resembles a true volition. 99 They therefore look a lot more like action than reflexes or spasms, because they are a lot more like action. 100 Nonetheless, dissociated movements do not qualify as voluntary actions so long as they are not caused by true volitions.
Respectfully, however, it is submitted that this argument is a metaphysical extravagance. 101 While philosophers, computer scientists, and psychologists can debate the matter, no one is close to proving the existence of any functional structures in the brain, like volitions, let alone their sub personal relations, ie Moore's 'volition-like' states. 102 What is required here is new evidence showing the exact brain structures that are needed to perform the executory functions of volitions. 103 It is true that studies on the supplementary motor area of the brain have been conducted in the past; however, this area still remains shrouded in mystery. 104 Moreover, it seems somewhat ill advised to expect members of a jury to discover such entities, for which no proof exists, and which have no equivalent in the everyday psychology that juries and lay-people rely upon to understand behaviour. 105

Moore's Second Response-'Dissociated Persons Do Not Volitionally Act Because They Are Unconscious'
Moore's second rebuttal centres on the notion of 'consciousness'. Specifically, he asserts that the dissociated do not voluntarily act because: Consciousness seems essential as part of our self-boundaries. So, if we (our conscious selves) are asleep, hypnotised, or otherwise not active, then we don't will anything. 106 The term 'consciousness' is open to misunderstanding, for it has a myriad of different meanings. 107 This paper will adopt the definition used most often in common parlance, namely, being awake and aware of one's self and environment. 108 Moore's argument, therefore, is that the lack of consciousness involved in dissociative states prevents the commission of willed/voluntary activity. Few would disagree with Moore that a sleepwalker-a classically dissociated agent-represents the prime example of an unconscious person. 109 However, the same cannot be said for hypnosis. A large body of recent research suggests that, 'hypnosis is not the same as being unconscious or asleep', 110 and that persons subjected to hypnotic influences remain 'wide awake, alert and discerning'. 111 Indeed, studies have shown that most participants report possessing a heightened sense of awareness, concentration and focus, which combine to increase the subject's receptivity to suggestion. 112 After reviewing the substantial amount of evidence in this field, including EEG 113 tests which are performed to measure the activity of a subject's brain waves, psychologist Graham Wagstaff concludes, 'there now seems to be a broad consensus among researchers that hypnotic participants do not lose consciousness'. 114 It could thus be seen as a mistake on Moore's behalf to conflate hypnosis with somnambulism by placing them both under the umbrella heading of 'dissociative states'. 115 Nevertheless, because Moore's approach hinges on the belief that hypnosis does constitute a state of impaired consciousness, the following discussion will, for argument's sake, assume that Moore is correct. Yet, even if we accept his submission, this does not necessarily entail that these agents are incapable of committing volitional acts. In other words, a lack of consciousness does not necessarily equate to a lack of volition. 116 This is where Professor Stephen Morse takes issue with the adopted approach. As Morse opines: I believe that Michael Moore does not adequately defend why the ability to be self-conscious is required for volitional action, nor does he indicate how much consciousness is necessary. It is possible that unconscious agents are not responsible and will thus be exempt from punishment, but are nonetheless acting persons. 117 It is submitted that Morse raises a particularly salient point here. Specifically, by questioning the degree-vagueness of the volitional theory, he highlights the fact that Moore never attempts to explain how much consciousness an agent actually needs to possess before he can be said to perform a voluntary/ volitional action; or, consequently, how much consciousness he needs to lack before his action becomes non-volitional/involuntary. In contemporary literature, it is uncontroversial to say that 'consciousness' is not an all-or-nothing concept. Rather, it occurs in varying levels and degrees; people can be more or less conscious than others. Consequently, it is also true that individuals can be more or less dissociated than others; or more or less hypnotised. Moore's theory fails to pinpoint the exact stage at which consciousness becomes involuntariness.
To further buttress his contention that unconscious defendants may volitionally act, Morse provides the example of R v Cogdon, 118 an unreported Australian case in which a somnambulist axe-bludgeoned her daughter to death. 119 Supposing that a properly constituted Mrs Cogdon would and should feel guilty for the death of her daughter, Morse infers that she would be guilty of something. He further infers that the something she would be guilty of is something she did. 120 While Morse may be correct that Mrs Cogdon has something to feel guilty about, this is arguably not a supposed action that she did while asleep. She should feel guilty for being the kind of person that would possess an underlying desire for her daughter to be dead. 121 As Moore, in response to this argument purports, '[w]e each have responsibility for our character, and Mrs Cogdon has some bad aspects to hers for which she should feel guilty about. Yet, responsibility for character should not be confused with the responsibility that the law cares about, which is being responsible for our actions'. 122 Feeling guilty for her character defects is thus quite compatible with her not having performed a voluntary action when her daughter was killed. 123 Alternatively, she may feel guilty because her body was causally responsible for her daughter's death. Nonetheless, being causally responsible for an event is not the same as being criminally responsible. 124 A vast number of people experience feelings of guilt for things that happen even when they are not intimately connected to the occurrence. 125 For instance, if a parent's child or family dog causes harm to another person, it is surely natural for the parent to experience feelings of guilt and responsibility, despite the lack of active supervision on their part. 126 To summarise, it is therefore submitted that Morse's argument fails here. The fact that dissociated agents may feel emotional guilt for the resulting harm caused does not automatically entail that they committed a voluntary act.

Moore's Final Response-'Dissociated Persons Cannot Volitionally Act Because They Lack Access to Good Reasons'
Moore's final response is closely tied to the previous argument. He suggests that the lack of consciousness involved in dissociative states, such as hypnotic trance, prevents the resolution of conflicting desires and intentions, and this resolving function is one of the crucial features of volitions. Moore elucidates thusly: To serve such a resolving function, volitions must be responsive to all (or at least a fair sample) of what one desires, believes and intends. And this is what being asleep, being unconscious, or being hypnotised prevents. These states seem to break the unity of consciousness that allows volitions to be formed that are responsive to all of one's desires, beliefs, and general intentions, and not just responsive to a small subset. 127 The terminology of this quote is complex, but all that Moore is saying here is this: an impaired state of consciousness goes hand in hand with an impaired decision-making process. This is because dissociated agents lack access to the reasons which would ordinarily be available to help them deliberate about what to do when conflict is present; they have been 'sealed off' 128 so to speak. The agent lacks access to the reasons not to cause harm, and thus cannot effectively restrain his behaviour. 129 This is Moore's foremost explanation for why the behaviour of the dissociated agent does not count as volitional action. Unfortunately, this explanation seems to prove too much, for it would negate action in 121 a substantial array of cases in which our best considered judgment is that action surely occurred. 130 For instance, numerous crimes against the person are crimes of passion, committed in heightened emotional states such as fear and rage. 131 Consider the enraged, jealous agent who discovers spousal infidelity and kills immediately in the heat of passion, or the grieving agent who kills immediately in response to a lesser provocation while suffering extreme emotional disturbance. 132 In each case, it is fair to say that the agent's emotions may 'seal off' his restraining reasons which would help to prevent him from behaving unlawfully. As Morse opines, 'the provoked agent may be just as unable to access his restraining reasons for action as the hypnotised person or the sleepwalker'. 133 Of course, Moore only intends to negate volitional action in the latter two cases, yet it is arguable that his account would negate volitional action in all of the aforementioned cases. Under our current system of criminal law, claims of provocation would, at best, constitute only a partial excuse of loss of control, which reduces a charge of murder to manslaughter. 134 As such, if there is an explanation for why the law should sometimes negate volitional action in cases of dissociation, it cannot be because these agents have their countervailing reasons 'sealed off'.

Dissociative States as Excuses?
Despite the foregoing arguments and counterpoints, Morse prefers to err on the side of caution by submitting that dissociated agents do volitionally act, but will likely be excused for their criminal conduct. Presumably, the alternative he favours is the insanity defence. According to Morse, 'the defendant who acted in a dissociated state is more like a legally insane actor than like an actor who harms as a result of a reflex movement'. 135 In England and Wales, the insanity defence requires that the accused suffered from a defect of reasoning caused specifically by an internal factor, namely, a condition which constitutes a disease of the mind. 136 Morse offers three reasons in support of treating cases of dissociation as instances of legal insanity. First, a successful plea of either defence results in exemption from responsibility, and thus, no blatant injustice will be done from either approach. Second, claims of dissociation are difficult to establish and easy to fake, 137 so it may be beneficial to place the burden of proof on the defence. 138 Finally, Morse argues that many of the arguments presented by Michael Moore are 'as consistent with excuses as they are with act negation'. 139 For instance, when Moore states that the ability to restrain oneself by being able to access and be guided by good reasons is a crucial component of volitional conduct, this is virtually indistinguishable from the concept of 'irrationality' which Morse has consistently claimed is at the heart of any legitimate insanity defence. 140 In regard to Morse's first point, while it is correct that both defences result in exemption from responsibility, it should be stressed that the outcomes of each defence post-exemption differ dramatically. If an automatism defence succeeds, the agent receives an unqualified acquittal. 141 In contrast, a successful plea of insanity may result in the agent's indefinite committal to a mental institution. 142 As such, the proposition that 'no blatant injustice' will be done, is rather dubious. Within the context of hypnosis, it is highly unlikely that such defendants require compulsory detention or treatment due to the extraordinary nature of the precipitating event. Indeed, compulsory detention of a hypnotised subject is most likely impossible by virtue of the amendment made by the Domestic Violence, Crimes and Victims Act 2004, which restricts hospital orders to conditions treatable in a special hospital. 143 The second of Morse's points, namely, that the burden of proof should be placed on the defendant due to the ease in which dissociation can be feigned is, however, particularly pertinent to hypnosis. It is very difficult to determine when a person is hypnotised and when he is not, as the hypnotised subject does not exhibit outward manifestations of his condition. Indeed, objective manifestations of hypnosis are virtually non-existent unless the subject is, at the time of the offence, connected to machines evaluating his blood pressure and respiration rate. 144 There is also merit to Morse's final suggestion that the theory of irrationality could more readily encompass cases of dissociated behaviour than lack of volition itself. In Burgess, 145 the courts followed a similar approach to Morse by holding that a certain dissociative disorder, namely, sleepwalking, is capable of constituting insanity. The court stated that episodes of sleepwalking are caused by an internal tendency to sleepwalk, and thus constituted a 'disease of the mind' for the purpose of the insanity defence. Although this may seem bizarre, it is not beyond the purview of the courts to interpret hypnosis in a similar fashion by declaring that 'hypnotic suggestibility' constitutes an internal factor for legal insanity. As Wilson and others purport, '[t]he status of hypnotic influences, which is generally assumed to cause simple automatism, may have to be reassessed in light of the decision in Burgess. As with somnambulism, not everybody is susceptible to the type of hypnotic influences which, it is claimed, may impel a person to commit a crime'. 146

Departing from the Volitional Theory
As the preceding discussion illustrates, this area is one of notorious difficulty. Moore himself admits that there is 'no certain answer', 147 and Morse, though somewhat favouring an excuse, begins and concludes his discussion by stating that overall, he is 'undecided'. 148 Thus, it may be beneficial to look elsewhere to form a conclusive opinion. Specifically, this paper concludes by considering the provisional proposals suggested by the Law Commission of England and Wales in order to decipher whether hypnosis occupies a legitimate space within a reformed automatism defence. The Commission recommend abolishing the rules which govern automatism and replacing them with a new statutory defence, framed as follows: Where the magistrates or jury find that the accused raises evidence that at the time of the alleged offence, he or she wholly lacked the capacity to control his or her conduct, he or she shall be acquitted unless the prosecution disproves the plea to the criminal standard. 149 For present purposes, it is important to note that the Commission have refrained from adopting terms like 'will' or 'volition' to describe when behaviour is voluntary/involuntary. 150  paper, the Commission do not seem to fully explain their departure from these traditional terms, aside from a handful of brief comments highlighting that the aforementioned phrases are 'not very helpful' 151 and 'not in ordinary usage'. 152 While it is acknowledged that these comments carry a certain amount of weight, it is submitted that a more thorough critique would place their departure upon a firmer basis. There are a number of more general concerns associated with defining voluntariness in terms of volition or 'what is willed', 153 some of which have been discussed above. It is likely that these critiques implicitly motivated the Commission to avoid framing their reformed automatism defence in this particular manner.
The first fundamental flaw with defining voluntariness in terms of exerting one's will is that it cannot adequately account for omissions. 154 For instance, if a parent deliberately fails to provide their child with sufficient nourishment, resulting in the child's death, it would be uncontentious to say that the parent's omission was voluntary and they should bear responsibility for their neglectful behaviour. 155 The issue, however, is that a proponent of the volitional theory would be forced to conclude that the parent's omission was involuntary. After all, the will is simply not involved when a person fails to do something. 156 This lacuna is an inevitable drawback of the volitional theory, because it solely attempts to explain when action is voluntary at the expense of other types of behaviour. 157 Clearly, refusing to hold the parent responsible in this kind of scenario would be an absurd result. A supporter of volition as the hallmark of voluntariness may attempt to avoid this difficulty in one of two ways. First, he may argue that certain omissions are really disguised actions. For instance, he may insist that the human statue or the gymnast holding a difficult pose are both flexing their muscles in order to remain still, and thus, both are covertly committing willed bodily movements. 158 The problem with this approach is glaringly obvious, for the idea of 'flexing one's muscles' is not even applicable to real-life scenarios in which liability is typically imposed for omitting to act. In essence, to regard a failure to act as a 'willed bodily movement' is to stretch ordinary language usage beyond reasonable bounds.
Second, one may try to argue that although omissions are not bodily movements, they can still be willed. In other words, there is a difference between, for instance, the person who deliberately omits to pay their taxes because of their political views and the person who simply forgets. It could be argued that the former is a voluntary or willed omission while the latter is not. 159 Nonetheless, this argument is also unpersuasive. As Wilson purports, 'to characterise a voluntary omission as a willed failure to act is conceptually absurd. And, of course, it is inconsistent with the numerous crimes where liability for an inadvertent omission is well recognised'. 160 Thus, if a worker kills a passer-by because he legitimately forgets to secure heavy machinery, he will still be held criminally liable despite the fact that he did not 'will' for the equipment to be left unfastened; 'I forgot' is generally not a legally recognised excuse. 161 A further objection to the volitional thesis is that it cannot adequately account for 'automatic' behaviour. For instance, the common phenomenon known as highway hypnosis occurs when a person drives quite competently for some distance but later reports no memory of having consciously done so. 162 This is comparable to habitual and other routine-like behaviour. In such cases, it is somewhat artificial to speak of a person as intentionally engaging in 'willed bodily movement', due to the inherent lack of any real thought involved in this type of conduct. Despite this, habitual behaviour provides no defence of involuntariness, due to the ease in which the habitual actor can turn his attention to what he is doing if the need to do so arises. As such, it may be difficult for Moore's theory to distinguish between the genuinely dissociated agent, and the agent who merely acts dissociatively.
Perhaps the most intractable flaw of the volitional thesis is a part scientific, part philosophical problem. As noted above, the volitional theory espouses that we hold actors responsible for their wrongdoing because in furtherance of something they desired, they caused their body to move in a particular way, with their volition acting as the executive lever. This represents the view that mind and body are distinct with volition (inside the mind) causing the body to move. However, from what we know of the natural world, it is difficult to discern how mental processes of any kind can legitimately cause bodily movement. 163 Mind and matter are not supposed to interpenetrate. Matter is subject to the laws of physics. There can be no exception. 164 We cannot will a cup to fall off the table because only physical forces can impart movement, whether those forces be a gust of wind or another moving body. It seems to follow that we cannot also will a human arm to move-even if it is our own arm-because both cups and arms are subject to the same laws of motion. If our arm moves something physical moves it. It may be force exerted by someone else or, more usually, the body's own biochemical messengers-neurons. The same laws of physics dictate that we cannot will our neurons to move, however much we may try.

The Control Principle
As demonstrated above, the Commission's proposed alternative focuses on the notion of 'control'. In their words, an accused satisfies their reformed automatism defence where he or she 'wholly lacked the capacity to control his or her conduct'. This emphasis on control seems to echo the views of various legal scholars. To provide several instances of this trend, note the similarity of the following passages: 'persons are responsible and deserve punishment only for those states of affairs over which they exercise control' (Husak); 165 'involuntariness should be defined in terms of a lack of control, in the sense of a total inability to restrain one's conduct' (Yeo); 166 'whether she was conscious or unconscious, what is essential to the denial of responsibility for a defendant's involuntary conduct is that she was unable to control that behaviour and to prevent it from occurring' (Simester). 167 Aside from articulating the importance of control, all three of the preceding statements seem to rest on the same bedrock notion of criminal responsibility, namely, that a person 'could have done otherwise'. 168 One may ponder whether this distinction holds any significance. After all, whether one defines involuntariness in terms of a lack of willing or a lack of control, these are merely two different approaches to explaining the same concept. Indeed, some would suggest that 'volition' and 'control' are mere synonyms. 169 Nonetheless, by addressing the issues examined above, it will be submitted that the control principle is to be preferred. Control is not only a more coherent theoretical basis for which to explain hypnotically induced behaviour, it is also more generally reflective of our current system of criminal law. 170 The first issue concerned the volitional theory's lack of applicability to omissions. In stark contrast, the control principle can readily account for omissions-based liability; the parent's omission was voluntary because providing the child with sufficient sustenance was within their control. 171 This theory explains why the parent is held responsible despite the absence of willing, and irrespective of whether the failure to fulfil their duty was inadvertent or deliberate. Similarly, the second general concern can be easily dealt with in a synonymous manner; the person acting on 'automatic pilot' is held responsible because it is within their bodily control to turn their attention back to reality.
The final issue revolved around the part philosophical, part scientific, mind-body problem. Without delving too deeply into the rich and tortuous literature, it suffices to say that the control principle avoids this complex problem, for it does not rest upon stipulating that criminal responsibility attaches to a specific mental process which causes bodily movement. Of course, talk of 'control' and 'could have done otherwise' raise their own philosophical issues of whether we are actually ever free to do otherwise or whether all of our actions are pre-determined-otherwise known as the free will versus determinism debate. 172 However, as the Commission acknowledge, to 'deny that we are free enough to at least bear responsibility for our actions would undo the entire foundation of the criminal justice system'. 173 As such, while the control principle is certainly not void of its own philosophical qualms, it is at least compatible with our current system of criminal law.
The final argument is specifically tailored to the aims of this paper: Can hypnosis constitute involuntariness for the purposes of legal automatism? As the preceding discussion illustrates, there are problems with trying to explain hypnotically induced behaviour simply in terms of whether or not it was willed. Specifically, it leads to the bizarre conclusion that incredibly complex and entirely purposive behaviour is not produced by volition. Vincent Chiao appears to concur with this statement when he suggests that, 'it is not at all clear that we should assimilate conduct that so strongly supports a complex, purposive interpretation to conduct that is akin to tripping and falling down a flight of stairs'. 174 Does the concept of control fare any better? It is submitted that the control principle provides an escape from this predicament, achieving the conclusion that Moore wants but without the controversy attached. Specifically, hypnotically induced behaviour is involuntary not because it is unwilled, but rather, because the subject lacks control over their will. This position is echoed by Corrado, who, it should be stressed, uses the words 'will' and 'choose' as synonyms: The mere fact that an action is the result of an act of will does not entail that it is voluntary. For example, if an agent is hypnotised, and while under the influence, chooses to do something that he would not otherwise have done, then (if hypnotism operates as we commonly believe it does) the agent could not have chosen otherwise, and his act is not voluntary. Yet it was willed by him. 175 This is a purely conceptual point, but Corrado's approach seems the most cogent: it is not the willing of an act that makes it voluntary; it is the willing of it while at the same time being able to will otherwise. 176 In other words, whether one has control over their conduct.
Of course, this still leaves open the possibility that in cases in which we possess no control over our will, there is simply no volition present at all. As previously discussed, this is Moore's way out of the problem. For Moore, volitions must be 'responsive to one's desires, beliefs, and intentions . . . to perform their resolving function', 177 ie the agent must be able to will an alternate course of action for action to be considered willed at all. However, from a practical point of view, an answer to these questions should suggest ways in which the various possibilities might be verified in court. As Corrado purports, 'Moore's answer absolutely collapses on this point': Which of these two questions would be more fruitful when put to an expert: (1) "Was the accused in a condition that deprived her of control over her choices and volitions?" (2) "Was the accused's volition unresponsive to the bulk of her beliefs and desires?" 178 The first admits of an answer that is more or less verifiable: 'Here is how the agent was treated beforehand; people who have been treated like that have been shown to be incapable of choosing anything but what they have been directed to choose'. In contrast, the complex terminology of the second question seems incapable of proof.

Conclusion
The aim of this paper has been to formulate a coherent definition of voluntariness that is able to accommodate instances of complex, purposive conduct, such as that committed under hypnosis. It has been suggested that the Law Commission of England and Wales, by abandoning the doctrine of volition in favour of the more contemporary control principle, has made a significant step forward in this respect. A lack of control more accurately explains the defect present in hypnotically induced individuals, as well as providing a simple and straightforward response to a number of well-known issues that have long plagued the volitional theory of action.
From a theoretical standpoint, it may thus be plausible to contend that a defendant operating under hypnotic influences should be permitted to rely on a control-oriented defence of involuntary conduct. Unfortunately, from a practical perspective, it is unlikely that such a defence will be successfully utilised in the foreseeable future. There would be numerous evidential difficulties if a hypnosis defence were to arise, such as the extraordinary ease in which hypnotic influence can be feigned, and the general difficulty of determining the precise extent to which a defendant subjected to such influences has 'lost control'. Ultimately, the future of this defence will depend upon advancements by those medical and scientific researchers whose ambition it is to untangle the complex strands of this intriguing and niche phenomenon. Until then, the law will continue down a path of substantial injustice; denying a defence to a group of defendants who are often more victims than perpetrators.