Gross Negligence Manslaughter Revisited: Time for a Change of Direction?

This article postulates that the House of Lords took a wrong turn in Adomako, missing the opportunity to revise the Caldwell/Lawrence guidance on recklessness, to produce a more appropriate determinant of criminal liability for inadvertent conduct causing death. It will be advocated that gross negligence manslaughter is replaced with reckless manslaughter utilising an objective capacity–based test. A proposal that encompasses both acts and omissions will be advanced which is theoretically underpinned by a hybrid theory of culpability. It will be contended that this hybrid theory best represents current approaches to criminally reckless conduct in practice and produces a morally apposite method of ascertaining criminal responsibility where the risk of death was not foreseen.


Introduction
Criminal liability for inadvertent conduct has come under increased scrutiny in the last few decades, particularly with regard to the offence of gross negligence manslaughter. From Stone and Dobinson, 1 to more recently Evans (Gemma), 2 Rose (Honey Maria) 3 and Bawa-Garba (Hadiza), 4 the common law has had difficulty in determining an appropriate line between culpability and exculpation. This article will begin by examining the historical development of inadvertent involuntary manslaughter to illustrate the recurring overlap between reckless and negligent conduct. Given the nature of these terms, some blurring of the boundaries is inevitable in the courts' pursuit of justice; this lack of clear demarcation is because recklessness and negligence are context/reason dependent. It is only when the full context in which the defendant's behaviour arose and his reasons for so behaving are analysed, that we can determine whether the imposition of criminal responsibility in any given instance is appropriate. It is submitted that the current law fails to successfully separate the culpable from those who should not be held criminally responsible. Consequently, a proposal for reform will be postulated to address this issue, underpinned by an appropriate theoretical rationale that is derived from current theories of culpability.

The Development of Inadvertent Manslaughter and a Proposal for Reform
Historically, one of the issues that arises with regard to inadvertent fault in the context of involuntary manslaughter is that the adjectives 'reckless' and 'negligent' have sometimes been used interchangeably in judgments. There was no requirement of awareness or its absence in the origins of this manslaughter offence and, following problems with defining recklessness, it seemed that two erstwhile distinct forms of mens rea had merged. By way of synopsis, from the standpoint of gross negligence manslaughter as enunciated by Brett J. in Nicholls, 5 'wicked negligence' was required, demonstrating a 'wicked mind' in the sense of being reckless and careless. Moral turpitude was essential but adverting to recklessness and carelessness is obfuscating matters, combining two (or three) mental states. 6 In Bateman, 7 Lord Hewart stated that gross negligence required 'such disregard for the life and safety of others as to amount to a crime'.
Advertence to risk was not apparently a factor, gross negligence could encompass both mental states. In some cases, there is little distinction, if any, between disregard and indifference to the welfare of others. Indifference is a term that has often been used in the context of recklessness suggesting a conflation of the mentes reae. This could be because gross negligence was traditionally viewed as the minimum state of recklessness to cover circumstances where there was a lack of foresight of risk. In this way, the lacuna left by a subjective definition of recklessness, where the defendant (D) had to foresee the risk of the particular kind of harm, could be addressed where, in the eyes of the court, policy and justice demanded that the accused was 'deserving' of punishment and must or should have foreseen the risk or deliberately closed his mind to it.
In Andrews v DPP, 8 Lord Atkin developed the guidance from Bateman further opining: [A] simple lack of care . . . is not enough . . . a very high degree of negligence is required . . . 'reckless' most nearly covers the case . . . but it is probably not all-embracing, for 'reckless' suggests an indifference to risk, whereas the accused may have appreciated the risk and intended to avoid it and yet shown such a high degree of negligence in the means adopted to avoid the risk as would justify a conviction. 9 In Stone and Dobinson, 10 Lord Atkin's statement of the law in Andrews was cited with approval by Lane LJ: [I]t is clear from that passage that indifference to an obvious risk and appreciation of such risk, coupled with a determination to run it, are both examples of recklessness . . . Mere inadvertence is not enough. The defendant 5. R v Nicholls [1874] 13 Cox CC 75 D. 6. Recklessness is above negligence in the traditional hierarchy of mentes reae terms and negligence can be divided into 'gross ' and 'simple' negligence. 7. (1925) Cr App R 11. 8. [1937] AC 576. 9. Ibid. at 583. 10. Stone and Dobinson (n 1), Stone and his mistress, D, were deemed to have assumed responsibility for caring for S's sister who died from a combination of anorexia nervosa and their incompetent neglect.
must be proved to have been indifferent to an obvious risk of injury to health, or actually to have foreseen the risk but to have determined nevertheless to run it. 11 Indifference is not a precondition to establishing recklessness and those who incompetently try to avoid a perceived risk may not necessarily be deemed reckless, depending on the reasons for continuing to act and for any incompetence. There could be indifference or callousness evidenced by their failure to take sufficient care to eliminate risk.
Chronologically, the judgments in Andrews and Stone and Dobinson are separated by Cunningham, 12 which utilised a subjective definition of recklessness, requiring D to have foreseen the risk or 'closed his mind' 13 to it. It would seem possible that using the phrase 'indifference to an obvious risk and appreciation of such risk' was opening the door to the Caldwell/Lawrence Direction, discussed below. It is clear that inadvertence alone is insufficient but this mirrors Lord Hailsham's position in Lawrence, 14 who stated that the word 'reckless' applied 'to a person or conduct evincing a state of mind stopping short of deliberate intention, and going beyond mere inadvertence, or, . . . mere carelessness'. 15 In Stone and Dobinson, 16 the conviction for gross negligence manslaughter was based upon 'gross neglect amounting to reckless disregard' for the sister's welfare. The court held that culpability could arise either by indifference to an obvious risk or subjective recklessness. This case, again failing to distinguish between recklessness and negligence, is hard to justify, given that both defendants were of low intelligence. They were not indifferent to the sister's well-being, doing their best to try to help her, and it is unlikely that either defendant was subjectively aware of the serious risk to her health. 17 It appears that foresight of consequences was unnecessary, but it is difficult to see how the defendants fell within the guidance from Lord Atkin in Andrews upon which the appeal court relied. 18 The next relevant milestone was the House of Lords decisions in Caldwell 19 and Lawrence, 20 where Lord Diplock extended the interpretation of recklessness to include inadvertence. He contended that a person would be reckless under the Criminal Damage Act 1971 if: (1) he does an act which in fact creates an obvious risk that property will be destroyed or damaged and (2) when he does that act he either has not given any thought to the possibility of there being any such risk or he has recognised that there was some risk involved and has none the less gone on to do it. 21 In Lawrence, 22 decided on the same day, Lord Diplock applied his 'Model Direction' but the 'obvious risk' under (1) was amended to an 'obvious and serious risk' for offences of reckless driving. Furthermore, an additional element was included as Lord Diplock stated: If satisfied that an obvious and serious risk was created by the manner of the defendant's driving, the jury are entitled to infer that he was in one or other of the states of mind required to constitute the offence and will probably do so; but regard must be given to any explanation he gives as to his state of mind which may displace the inference. 23 11. Ibid This dictum raises the possibility that where D was inadvertent, both capacity and the presence of moral wickedness is required for culpability to be attributed. 24 This is a sound proposal however the requirement of 'moral wickedness' should be replaced with the need for a culpable moral failure, as 'moral wickedness' could be interpreted as requiring evidence of evil or maliciousness, in the modern sense of these terms. There was ambiguity in both of Lord Diplock's judgments: was recklessness to be determined purely objectively, regardless of the cognitive capacity of the accused or was it adaptable to the capacity of the individual defendant? As is known, an objective interpretation was judicially adopted, removing the demarcation that had previously existed between advertence to risk (recklessness) and inadvertent conduct (negligence).
The Model Direction was subsequently applied by the House of Lords in Seymour, 25 which held that recklessness was the most suitable term to signify the level of culpability required for what had previously been termed gross negligence manslaughter. 26 This was not a surprising development as recklessness now included both the advertent and inadvertent risk taker. Although the Caldwell/Lawrence Direction called for an 'obvious' and 'serious' risk of harm, Lord Roskill believed that recklessness as to 'the risk of any degree of harm would suffice'. 27 Clearly that is in line with the traditional view of recklessness, foresight of harm of a particular kind, 28 not the extent of harm that occurs. By way of contrast, gross negligence manslaughter had generally required a risk of death or at least serious harm, 29 apart from the judgment in Stone and Dobinson 30 (above) where Lord Lane suggested that 'an obvious risk of injury to health' was necessary. 31 The extensive approach to manslaughter adopted in Seymour (the risk of any degree of harm was sufficient) would be too broad to sustain without being subject to amelioration by consideration of the capacity of D and the surrounding circumstances. This leeway was potentially provided by the House of Lords in Reid, 32 in holding that the Model Direction was not to be followed ipsissima verba but should be adapted to the facts of the particular case with consideration of any explanation or excuses offered by the defence as to why D should not be found reckless, as stipulated in the Lawrence version of the Direction. 33 In 1993, a group of appeals against conviction for involuntary manslaughter were heard together in the Court of Appeal: Prentice and another; Adomako; and Holloway. 34 As is known, the appeals centred upon the correct approach to be adopted for involuntary manslaughter by breach of duty: gross negligence manslaughter in line with Bateman, or Caldwell/Lawrence recklessness as per Seymour.
The Court of Appeal in Adomako held that the test for recklessness applicable to cases of motor manslaughter was not appropriate to involuntary manslaughter by breach of duty, the test for gross negligence manslaughter should be applied. 35 Lord Taylor found the necessary ingredients of the offence to be (1) the existence of a duty; (2) a breach of this duty causing death; and (3) gross negligence. Preferring not to suggest a standard direction for juries as to what would amount to a finding of gross negligence on their part, his Lordship propounded the following as a non-exhaustive list: a. indifference to an obvious risk of injury to health; b. actual foresight of the risk coupled with a determination nevertheless to run it; c. an appreciation of the risk coupled with an intention to avoid it but also coupled with such a high degree of negligence in the attempted avoidance that the jury consider justifies conviction; d. inattention or failure to advert to a serious risk which goes beyond 'mere inadvertence' in respect of an obvious and important matter which the defendant's duty demanded he should address. 36 Points (a), (b) and (d) were potentially covered by the Caldwell/Lawrence Direction on recklessness. Even (c) would have been accommodated where there was evidence that D was uncertain that he had eliminated the risk entirely as demonstrated by Chief Constable of Avon and Somerset v Shimmen. 37 Moreover, where D had sufficient knowledge and skill and his efforts to entirely eliminate the risk were abysmal, it is likely that a jury would deem him to be reckless anyway. In practice, a judge or jury is unlikely to find D's version of events very convincing in such circumstances. In Reid, their Lordships were sceptical about the existence of any 'lacuna' in the Model Direction, Lord Browne-Wilkinson stating: [t]here may be cases where, despite the defendant being aware of the risk and deciding to take it, he does so because of a reasonable misunderstanding, sudden disability or emergency which render it inappropriate to characterise his conduct as being reckless. 38 It is therefore a matter of logic that where D unreasonably decides he has eliminated such a risk he is deemed reckless, 39 unless there is evidence to rebut that presumption.
The Court of Appeal gave four reasons for preferring gross negligence. The first was that the 'wide definition' of recklessness from Caldwell/Lawrence had caused difficulty for the ordinary lawyer and juror who may have felt that the word 'reckless' had stricter connotations (presumably accounting for capacity). The Court believed it had caused problems for the courts in cases of involuntary manslaughter that would not have happened had gross negligence been the test. 40 The House of Lords, on appeal, rejected defence counsel's cogent arguments that a sole test of negligence was inapt because it could mislead a jury into thinking it equated with the civil test, and the further point that the term 'gross' is unhelpful. This epithet does not provide a mental element and if it is there to signify that a jury must consider both the seriousness of the degree of negligence and a bad attitude on the part of the defendant, it needs greater elucidation. 41 Adopting recklessness would have addressed these objections, but was perhaps rejected purely because of the difficulties the Caldwell/Lawrence formula had caused since its emergence, rather than for any other reason. These difficulties were not simply attributable to ambiguity in Lord Diplock's judgments but more specifically to the poor judicial reasoning and application that followed in their wake.
The Court of Appeal's second reason for favouring gross negligence centred on the perceived difficulty that the Caldwell/Lawrence Direction on recklessness specifically referred to circumstances where the defendant himself had acted to create the relevant risk. Lord Diplock had stated that a person would be reckless if '(1) he does an act which in fact creates an obvious risk . . . and (2) when he does that act . . . ' 42 Adopting a literal approach to this wording, it could not apply to the many of the cases of involuntary manslaughter, other than motor manslaughter, which arise not because of D's commission of 36 an action, but rather because D failed to act to prevent harm when he was placed under a legal duty to act. The Court observed that breach of duty cases involving doctors are in a different category as the risk to the patient's health is already present and this is what causes the doctor to assume the duty of care with consent, often in an emergency situation. 43 Accordingly, it seemed that the Caldwell/Lawrence Direction, due to its formulation, could only apply to cases of commission, leaving omissions to be dealt with utilising the requirements for establishing gross negligence. This is a cogent argument but given that the House had already accepted the Direction was not to apply ipsissima verba, this was hardly an insurmountable hurdle. 44 The third reason was also related to the exact wording of the Model Direction; the 'obvious' risk referred to in the Direction meant that the risk would be obvious to the 'ordinary prudent individual'. In cases where a breach of duty arises the defendant is often an expert in his field and would therefore appreciate risks, or should appreciate risks, which the ordinary prudent person may not. This is another valid argument, but only if we adhere to the exact wording of the Direction; it could easily be modified to take account of this. Where a defendant has special knowledge or expertise and should identify a risk that would not be obvious to the ordinary prudent man, he could still be caught by the subjective limb of the Direction if necessary by a finding that he 'deliberately closed his mind' to it.
However, to produce a more apposite formula, the objective limb of the Direction must be modified ('has not given any thought') to make it clear that the capacity (both general and specific), knowledge and skill of the accused is to be taken into consideration. Accordingly, a doctor could be judged by the standard of a doctor with similar training and experience and in the circumstances existing at the time of the conduct. Applying this proposition to Prentice and Sullman, being unfamiliar with chemotherapy drug administration and only Sullman having experience of performing lumbar punctures, together with other mitigating factors, neither would be deemed reckless. In the Court of Appeal, Lord Taylor had found that on a literal interpretation of the Caldwell/Lawrence Direction such factors could not be considered, whereas a jury could take them into account when deciding if their negligence was gross. 45 This was arguably only an accurate statement of the law if the Direction had to applied literally; the House of Lords in Reid 46 had already ruled that this was not so, a point noted by the appellate court in Adomako. The Court then appear to have ignored the guidance from Reid after citing it. The version of the Direction given in Lawrence 47 expressly allowed for such factors to be taken into consideration. 48 An application of the Lawrence version of the Direction to Stone and Dobinson 49 would find the limited capacity of the defendants to be exculpatory. Similarly, an explicitly modified objective limb of the Direction, as advocated here, contextualised as suggested in Reid, 50 would have the same effect. Yet applying the test of gross negligence to these defendants resulted in their conviction. Perhaps this is because in determining guilt, the focus has been more on how far the accused falls below the standard of a reasonable person in that role, without considering all the circumstances at the time of the actus reus.
The final reason for choosing gross negligence over recklessness was with regard to the situation where D has foreseen a risk and tries to eliminate it in an incompetent manner. 51 Technically, such a person was not covered by the Model Direction unless he realised he had not entirely eliminated it before acting and would fall within this perceived lacuna. 52  would still be caught by the test for gross negligence. However, he could equally have been deemed reckless by closing his mind to the fact that the risk remained. It is doubtful that the lacuna ever existed, and it is possible under the approach to recklessness advocated here that in such circumstances D could be deemed reckless in choosing to eliminate or avoid the risk in the way that he did. Alternatively, he could be simply negligent because he did his incompetent best and honestly thought he had done enough to prevent harm. In the latter case, criminal liability would not be justified. Lord Taylor considered Adomako's case to be an example of the problems that can arise if Lord Diplock's Direction is given to the jury. 53 If a defendant is reckless where he has failed to give any thought to an obvious and serious risk, then Lord Taylor believed a jury would convict Adomako readily on the facts. It is odd that the inadvertent strand of the Direction was applied here because Adomako did try to address the risk he simply failed to correctly diagnose the cause. For Lord Taylor, once the defendant realised there was a serious risk, his response was grossly negligent, circumstances not within the formulation of the Direction but covered by Lord Atkin's dicta in Andrews. 54 This is a fair point, given the courts' interpretation of the Direction, but whether the defendant should have had his conviction upheld is subject to debate. 55 He would be only be convicted under the formulation of the modified direction on recklessness proposed here if his conduct showed an undesirable attitude towards his patient's care, which was not present.
With regard to Holloway, the trial judge had directed the jury in terms of the Caldwell/Lawrence Direction on recklessness. As the defendant had never thought there was a serious risk to the occupants of the house, he clearly fell within its scope. Although the jury were told to take account of the defendant's explanations as to why he failed to identify the risk, as far as the Court of Appeal were concerned, once that state of mind was confirmed (D giving no thought to the risk) any explanations were superfluous-the state of mind was already made out and could not be altered by the reasons why it occurred. This was crucial as it justified adopting gross negligence rather than recklessness as the correct approach; questioning why there was a lack of awareness of the risk would be central to a finding of gross negligence when 'examining the degree of negligence of a skilled man exercising his trade'. 56 However, if D's reasons as to why he should not be deemed reckless are redundant in the Lawrence formulation of the Direction, why were they specifically included in it? It is respectfully submitted that the Court of Appeal erred in this respect.
Adomako was granted leave to appeal by the House of Lords on the question of whether a direction to the jury on gross negligence manslaughter was sufficient without reference to the test for recklessness as formulated in Lawrence or as adapted to the circumstances of the particular case. The appeals against conviction of all the defendants, apart from Adomako, were allowed. 57 The House agreed with the Court of Appeal's decision to reject the Seymour objective recklessness approach to manslaughter and reinstate gross negligence as the appropriate fault term in all such cases of involuntary manslaughter. 58 For Lord Mackay, once a duty of care and causation were established, the question for the jury was whether 'having regard to the risk of death involved' D's conduct was 'so bad in all the circumstances' as to amount to a crime. 59 His Lordship recognised the circularity in the test: D can be convicted of a crime if the jury find his conduct is criminal; but found this acceptable. 60  formulation has been advanced: the real issue for the jury is not whether D's conduct was so bad that it was criminal, but rather 'whether it is bad enough to be condemned as the very grave crime of manslaughter and punished accordingly'. 61 A conclusion can only be drawn in this respect by examining all the circumstances at the time of the actus reus, as proposed here.
The issue as to the extent of the risk involved, now the test was to be that for gross negligence, lacks clarity in Adomako as there was no dissent from the Court of Appeal's position: 62 a risk of injury or death. 63 This would produce a test that was just as broad as that in Seymour (the risk of any degree of harm), and not as restrictive as Caldwell/Lawrence (an obvious and serious risk). Subsequently Gurphal Singh 64 and Misra 65 confirmed that a risk of death was required. Lord Mackay's acknowledgement that a judge may use the term 'reckless' in his guidance to a jury in this type of case with the meaning used in Stone and Dobinson and West London Coroner, Ex parte Gray 66 is unhelpful. Lord Lane's reference in Stone to indifference to an obvious risk or foresight of the risk is reminiscent of the recklessness directions that the House was now seeking to avoid. 67 Moreover, in Ex parte Gray, the judgment is consistent with both Lawrence and Seymour. 68 The Law Commission subsequently noted that in such cases of involuntary manslaughter, it must be shown that D caused the death of another through gross negligence or recklessness, but it was unclear what these fault terms meant or even whether they were describing two separate categories of manslaughter or the same one. 69 It is postulated here that continuing with reckless manslaughter, based upon a modified Caldwell/ Lawrence Direction as advocated in Reid, would have been preferable. The offence of reckless manslaughter, in the Cunningham subjective sense, still exists 70 where D acts in what would be a lawful way (were it not reckless) knowing that there is a high probability that he will cause serious injury but where he does not see it as a virtual certainty. 71 Given that the subjective test for recklessness is in practice, an objective capacity-based test 72 there would have been some congruence. 73 A capacity-based modification to the Caldwell/Lawrence Direction is a preferable alternative to gross negligence. A broader, capacity-based approach to recklessness, taking into consideration D's cognitive capacity and knowledge at the time of the actus reus, 74 including why D failed to see the risk or continued to act despite his appreciation of it, is advocated. If such an examination reveals that D's conduct displays a bad character trait, liability could lie. As far as the capacity of D is concerned, where there is evidence of incapacity exculpation will be dependent upon the 61 extent to which such incapacity is manifest in the circumstances and the extent to which this is faultfree. 75 The remaining obstacle to using a modified Model Direction is adapting it to cover liability for omissions. It is proposed that a person will be deemed reckless for failing to act if he has a legal duty to act and is, or should be, aware of a serious and obvious risk to the victim's welfare and yet fails to act to prevent or ameliorate harm. This is where: (1) he fails to act when there is an obvious and serious risk of death or serious harm to another when he under a legal duty to act [towards that other person], and (2) he either has not given any thought to the possibility of there being any such risk or he has recognised that there was some risk involved and has ignored it or tried to eliminate it in a wholly incompetent manner; and (3) if satisfied that an obvious and serious risk in such circumstances has not been considered or is dealt with in a wholly inappropriate manner by the defendant, the jury are entitled to infer that he has the state of mind required to constitute the offence and will probably do so; but regard must be given to any explanation he gives which may displace the inference.
(i) evidence of a general or specific lack of capacity in the circumstances may be exculpatory; (ii) evidence of a reprehensible attitude will not displace the inference.
A more morally substantive approach to determining culpability could be achieved by asking why the inadvertent defendant did not foresee the risk. 76 Once the reason is proffered, a determination can be made as to the moral blameworthiness of the defendant by applying philosophical principles. With regard to (1), liability will not generally be excluded where the assessment of any risk is absent because of a breach of duty, exculpation will be dependent on why the duty was breached. The proposed formulation extends inadvertent liability to foresight of causing someone serious injury, 77 making it apparently broader than the test for gross negligence which now requires a risk of death. However, it is more restrictive in that without evidence of a reprehensible attitude criminal liability will not be established. The extension of liability to cover an obvious risk of serious bodily harm would provide some symmetry with the law of murder, given the mens rea is an intention to kill or an intent to cause serious harm. It is also in line with recommendations made by the Criminal Law Revision Committee 78 and the Draft Code. 79 The Law Commission, in its proposed reforms of involuntary manslaughter, 80 has opted to abolish common law involuntary manslaughter and replace it with the offences of reckless killing 81 and killing by gross carelessness. 82 The first of these, reckless killing, has the same defect as the draft Code's definition of recklessness; 83 it fails to make clear who decides whether the risk is a reasonable one to take. Lamentably, it also departs from the Code's attempt to restrict the grievous bodily harm rule to 75 circumstances where D was aware that his intention to cause serious harm might cause death. 84 The proposed offence of killing by gross carelessness will be satisfied where a person causes the death of another if: (a) a risk that his conduct will cause the death or serious injury would be obvious to a reasonable person in his position; (b) he is capable of appreciating that risk at the material time; and (c) either-(i) his conduct falls far below what can reasonably be expected of him in the circumstances; or (ii) he intends by his conduct to cause some injury or is aware of, and unreasonably takes, the risk that it may do so. 85 It is clear that this draft provision not only provides a mixture of both advertent and inadvertent conduct but also encompasses intention, recklessness and negligence. It is respectfully submitted that it is too widely drafted and fails to take account of whether D appreciated the risk or why he failed to appreciate it.
It has been suggested that the difference between reckless conduct and negligence is that with the former there is indifference to the risk of harm whereas the latter is a failure to take adequate precautions to ensure a risky act is performed safely, 86 but neither indifference to risks nor awareness of them are necessarily the deciding factors. The real difference is that the term 'reckless' implies condemnation 87 and censure in a way that 'negligence' does not. 'Gross' negligence is synonymous with recklessness and should be deemed to be recklessness, leaving negligence to mean mere inadvertence or everyday carelessness. Once the definition of recklessness is allowed to encompass both the advertent and inadvertent risk taker, the problem with terminology disappears. Such a change would also bridge the existing gap where a negligent act does not result in death as, if inadvertent recklessness is the test, liability could arise where serious physical injury occurs. Having formulated a new provision for reckless manslaughter, it is necessary to underpin this with an appropriate theoretical rationale. This is now examined below.

Theories of Culpability and Inadvertent Conduct
Whether D is advertent or inadvertent to a risk of harm is only one factor that may be relevant to determining the degree of culpability. When applying philosophical principles of culpability to the different approaches to inadvertence, there is clearly a lack of congruence. The three most important theories relevant here are choice theory, character theory and Gardner's 'role theory'. Choice theory, or capacity theory as it is now better known, is an adaptation of Kantian retributivism 88 and the principle of desert underpinning justification for punishment. The basis of this theory is that as a rational moral agent, D should not be punished unless he had both the capacity and a fair opportunity to abide by the law. 89 This respects individual autonomy ensuring that D should only be criminally liable for harm that he is responsible for and culpably causes. A person should only be responsible for matters he can control which are those he freely chooses to do or causes to happen. 90 This restricts influences of chance and luck over which we have no control and which should be irrelevant to culpability and criminal liability. 91 Therefore, 'one is responsible if he could have done otherwise', 92 if he had chosen to. 93 Consequently, this theory cannot successfully account for criminal liability for negligence. It is argued that negligent behaviour could be encompassed where D possessed both the physical and moral capacity to act like a reasonable person, had a fair opportunity to avoid criminal behaviour, 94 but chose not to exercise the relevant capacities. Moore refutes this suggestion that a person can in any real sense 'choose' to behave in a negligent manner, 95 as negligence connotes inadvertent action, if no thought is being given before acting in a particular way how can the subsequent act be 'chosen'? 96 This is a valid objection although negligence is not restricted to the inadvertent, as it encompasses the advertent incompetent too. Incompetence could represent a lack of capacity in the circumstances or a reckless disregard for the health and welfare of the victim. To suggest as a general proposition that on occasions when we do not consciously stop to consider possible risks that we are necessarily choosing to do so is ridiculous. Nobody would try to argue that Adomako chose to ignore the possibility his patient's endotracheal tube had been disconnected, for example.
Choice theorists suggest that it is only where an external crisis occurs that D is excused, providing he behaved as a reasonable person would have done in his situation. This is a draconian position to adopt where conviction for a serious offence like manslaughter may result. Clearly, anyone who lacks the cognitive capacity to rationally consider the situation before acting should be exculpated. It is this capacity element of the theory that is most relevant here-not just in the sense of a general everyday capacity D may possess but the specific capacity he had at the relevant time considering all the circumstances. It is acknowledged that there may be a practical difficulty in determining whether in any given situation, D was actually deprived of the capacity/fair opportunity, given that this may come down to the disputed evidence of expert witnesses, 97 but it would still be an improvement on the current state of affairs.
The next theory for consideration is character theory, the main proponents being Aristotle, 98 Hume, 99 Bayles, 100 Pincoffs 101 and Arenella. 102 Here, criminal culpability is based on the subjective character traits of the individual moral agent, reflecting the principle that only 'bad people' should be punished. 103 For Hudson, '[m]oral virtues and vices . . . are traits which received opinion holds to be traits of character; they are time-tested'. 104 Hume proposed that '[b]lame and punishment are . . . for character traits', but the ability to voluntarily control the character trait is essential. 105 Criminal liability is justifiably imposed where D's action manifests an undesirable character trait, an enduring mental quality 106 that requires correction. Although mental qualities do not have to be voluntary to be blameworthy, punishment requires their voluntary manifestation. 107 If an act does not indicate an undesirable character trait blame would be inappropriate and a person would be excused, 108 as the purpose of punishment is only where its intended use is to alter a person's conduct. 109 Utilising character theory principles, both the advertent and inadvertent agent could be demonstrating bad character flaws but the capacity of the individual defendant would need to be taken into account when determining whether his action or inaction evidenced this. Even though we may punish a single negligent act, such an isolated action does not signify that we are often careless, a bad character trait. Although it has been mooted that a consistently negligent actor demonstrates an attitude of indifference, 110 this is a generalisation as such conduct could equally be attributable to 'awkwardness and stupidity'. 111 Evidence of conduct manifesting a reprehensible character flaw is essential to establishing culpability and where an agent lacks capacity this will be absent.
If the capacity of the accused is irrelevant, it would fit well with Gardner's 'role' theory of culpability. 112 Gardner 113 dismisses the traditional views of both choice and character theorists, instead grounding culpability and responsibility on a role basis. He argues that the actions of D are not merely evidence of D's character but rather constitute it and finds choice theory inextricably linked to character theory. 114 On Gardner's 115 model, responsibility only lies where we are fulfilling a role, for example a specific role such as a doctor, or a non-specific role, that is, a human being, and we fall below an idealised standard of a reasonable person in the role we are fulfilling. All roles have standards of character, skills and knowledge attached to them and D should only be excused if his conduct fell within the boundaries of reasonableness for someone in that role. It is irrelevant whether we have the capacity to achieve this idealised standard; a person's capacity to do better is immaterial. This model fits well with liability for negligence and for non-advertent recklessness (as per Caldwell recklessness which covered both subjective and objective tests within its scope) but is too draconian a stance for imposing criminal liability for manslaughter unless the policy is to privilege harm over culpability. 116 It is possible to extrapolate elements of choice, character and Gardner's 'role' theory to present a composite that would ground culpability for inadvertent conduct causing death based on moral desert. This could be achieved by subjectivising Gardner's theory to take account of the capacity of D. D could be judged on the basis of whether on the particular occasion in question, in performing whatever role, the character of his conduct was morally blameworthy. Scrutiny of D's conduct would consider the specific capacity D had in the existing circumstances. The focus shifts from ascertaining whether D fell far below the standard of the reasonable person performing D's role to why the incident occurred, what circumstances contributed, did the agent make an error of judgment or show a complete disregard for the victim? It is this synthesis of the theories that is advocated here and which forms the basis of the limits of criminal liability for reckless manslaughter. Where D is inadvertent, account should also be taken of the cognitive capacity and attitude of the accused. There may be many reasons why agents fail to be aware of risks. For example, the inexperienced may lack capacity to foresee at least some of the risks obvious to the prudent person in some circumstances. 117 Unless we have learned by experience or have information that risk exists in some particular activity, we are unlikely to think about them. 118 A synthesis of the subjective/objective positions will acknowledge that moral culpability cannot rationally be dependent upon advertence or inadvertence. It is submitted that where D's inadvertence to the risk is the result of morally blameless factors, he should not be criminally responsible for such a serious offence. As others have argued, 119 there are certain character traits and emotions to which we tend to attach blameworthiness, for example, lust, greed, anger and jealousy and it is such characteristics that we rightly expect people to control that, uncontrolled, can lead to culpability. Incapacity should not attract culpability unless it is self-induced. A person may have some capacity to foresee a risk and yet not be able to identify the choices that are then open to them. This does not necessarily evidence a reprehensible character trait. Society expects people to form their opinions based upon reasonable grounds and any moral distinction based upon D's opinion that there was no risk must rely on an assumption that D would have acted differently had he known otherwise and this is not always the case. 120 Where D forms an opinion based upon unreasonable grounds or where he wrongly believes there is no risk, it is likely that the court would find his explanations unconvincing unless there were other factors present that impacted upon his practical reasoning. Any principle which makes it permissible to select between different factors affecting foresight should be grounded in character theory, allowing justice to be done in circumstances where a bad character or attitude was manifested and for agents to be found not criminally liable where it was absent. This will impose an evidential burden on the accused. Once the reason why no thought was given to the risk emerged, it would be relatively straightforward to assess the degree of moral blameworthiness and thus any criminal liability. This approach looks beyond the subjective/ objective dichotomy and add another dimension, why the accused acted as he did, his motivation or emotion behind the actus reus 121 and the context in which the proscribed act occurred.
It is contended here that a capacity-based modification to the Caldwell/Lawrence Direction would have been a preferable alternative to gross negligence. A broader, capacity-based approach to recklessness, taking into consideration D's cognitive capacity and knowledge, 122 and why he failed to see the risk or continued to act despite his appreciation of it, is advocated. If such an examination reveals that D's conduct displays a bad character trait, liability could lie. As far as the capacity of D is concerned, where there is evidence of incapacity exculpation will be dependent upon the extent to which such incapacity is manifest in the circumstances and the extent to which this is fault-free. 123 Where D is found negligent, it is because his conduct is deemed to fall below, or in some cases to grossly deviate from, the standard expected of the reasonable man. This requirement of reasonableness is designed to serve as a balance for competing legal interests: those of the particular defendant and those of society. 124 By focussing on the comparator of this hypothetical reasonable man, the motives, intentions, attitudes and so on, which underpin the defendant's conduct are largely irrelevant, which has the potential for unfairness. This is because, where there are negligence-based convictions, it can be very hard to distinguish between simple mistakes and accidents for which civil liability might be more appropriate, and instances where an agent's stupidity, disregard for others, or arrogance may deservedly attract criminal sanction.
It is accepted that there are two main elements to criminal law, namely harm and culpability. The latter is not established just because D caused harm but because that harm was done culpably, 'even a dog distinguishes between being kicked and being stumbled over'. 125 One view is that liability for negligence can be justified on utilitarian grounds in that it acts as a deterrent in general, even if not on an individual basis. 126 Where the harm caused is serious and the risk of harm is an obvious one, failure to think about or recognise this risk can be as culpable as where D perceives a risk. Often the fault can lie in an agent's failure to notice a particular consideration is relevant. 127 If account were to be taken of any limiting capacities of the defendant, this would circumvent the problem with the Caldwell Model Direction. What is advocated here is that criminal liability would arise from a more objective capacitybased approach to recklessness where there is a serious and obvious risk of harm, which also takes account of why D failed to foresee the risk of harm. The remit of liability for negligence would remain firmly rooted in civil law.
In the context of the prosecution of medical practitioners for gross negligence manslaughter, a lack of flexibility has been identified with a 'one size fits all approach' that fails to adequately distinguish between different medical mistakes, the relevance of moral luck and the vagaries of expert witness testimony. 128 It has been argued that we should not prosecute medical professionals merely for making mistakes and should only convict those who are 'cruel or indifferent doctors' or those 'insufficiently motivated [by] the interests of their patients'. 129 Making a mistake or demonstrating a lack of skill, no matter how tragic the consequence, does not establish this without further evidence.
[O]nce in a while, our lack of information, failure to notice, or forgetfulness results in our underestimating the riskiness of our conduct and causing harm . . . . An injunction to notice, remember, and be fully informed about anything that bears on risks to others is an injunction that no human being can comply with, so violating this injunction reflects no moral defect. 130 There is clearly a serious issue to be addressed in this regard, especially given that healthcare professionals are placed under a duty of care by virtue of their profession and are trying to help their patients, often in a highly pressurised environment. 131 There have been suggestions that either new tailor-made offences should be enacted to cover medically caused death or injury 132 or that only subjectively reckless doctors should be criminally liable. 133 It is contended here that professionals should only be prosecuted where their conduct evinces a level of ineptitude that equates to inadvertent recklessness 134 coupled with evidence of a bad character trait.
Historically, there is a judicial reluctance to permit general personal incapacities to excuse in favour of efficiency and pragmatism 135 but a failure to recognise such general and specific incapacities results in the criminal liability of the non-culpable. It is important that the characteristics of the agent should be taken into consideration when they impact upon his ability to behave as the reasonable person would and when they are not his fault. 136 Additionally, it is essential to ascertain why he failed to advert/avoid the relevant risk. If this failure is a consequence of external factors that are beyond the agent's control, they will not evidence a reprehensible character flaw and criminal responsibility for a serious crime would be inappropriate. Those who inadvertently cause harm may well warrant moral criticism but do not necessarily also deserve the full force of the criminal law to be brought to bear. Although it has been suggested that the criminal law is simply concerned with our fitness for a particular role, 137 this is not the sole determinant of liability as the law targets punishment at those whose behaviour has shown possession of particular vices. The relevant manifested vices are those that cause D to be 'insufficiently motivated to act or not to act by the interests of others', 138 for example, cruelty, dishonesty and indifference. 139 This requirement of a manifestation of one or more of these core vices is necessary for criminal responsibility; where they are absent civil liability is more appropriate. 140 Inadvertence 'does not reliably track the moral vice of insufficient concern that all the other legitimate forms of criminal culpability display'. 141 Any vices which do not show a total lack of regard for the interests of others should not be the concern of the criminal law. 142 Criminal liability represents the State's condemnation, which is only appropriate where D's vices cause harm and demonstrate a lack of concern for the interests of others. 143 Inadvertence alone does not portray this. A lack of skill can be blameworthy but it does not follow that demonstrating a lack of skill automatically deserves criminal liability. 144 The only acknowledged problem with this proposition is that it may be difficult to prove beyond doubt that D showed such disregard or was merely forgetful, preoccupied or distracted.
Applying these principles, there is a case for arguing that Adomako did not deserve criminal punishment. Once he realised his patient was in serious trouble, he did everything he could think of to remedy the situation, showing appropriate concern rather than disregard or indifference. 145 He clearly lacked the skill of a competent anaesthetist on this occasion, but there were many mitigating factors. 146 The contrary position is illustrated by Misra and Srivastava, 147 where the doctors' inaction showed a total disregard for the patient's welfare. 148 Here criminal liability is clearly appropriate and no difficulty arises in establishing the lack of concern required, so perhaps differentiating between the callous and the forgetful, preoccupied or distracted may not be such a problem in practice. Finally, the proposed provision will now be applied to more recent contentious cases to determine its efficacy.

Application of the Proposed Direction to Controversial Cases
As is known, in Evans, 149 D had purchased drugs, brought them into the family home, later discovered that her stepsister was showing signs of suffering from an overdose but failed to summon medical assistance. Her conviction for gross negligence manslaughter was upheld on the basis of a legal duty to act by extending the Miller 150 principle. The utilisation of this principle has already been subject to some trenchant criticism 151 but how would the proposed direction apply? If it is accepted that a duty should exist then (1) is satisfied, and under (2), the risk of harm was recognised, Evans and her mother tried to eliminate it by looking after the victim themselves. As for (3), Evans would then have the opportunity to explain why she should not be deemed reckless. It is common knowledge that drug addicts do tend to try to assist each other rather than seek external assistance 152 and Evans stated that she and her mother had not wanted to get themselves or the victim into trouble, believing they were capable of adequately caring for the victim themselves. There does not appear to be evidence of any lack of capacity in the circumstances leaving the question of whether D's actions manifest a reprehensible attitude towards the victim? It could be argued that by trying to care for Carly, Evans was not displaying a callous disregard for her health and welfare, 153 and therefore should not be criminally liable.
More recently, in Rose, 154 the appeal against conviction for gross negligence manslaughter was upheld but has been subject to much criticism as it exculpates where D is ignorant of any risk only because they have failed to fulfil their duty. 155 One response in defence of this development suggests that the negligently ignorant would be relatively few in number, such behaviour being deterred by the possibility of disciplinary action from relevant healthcare professional bodies. 156 However, it is submitted here that it is exactly defendants like Rose who should be criminally responsible for their breach of duty. Rose demonstrated by her actions a total disregard for both her statutory duties and the health and welfare of her patient. She clearly satisfies (1) if assessing the risk is retrospective. She satisfies (2) as because of her breach of duty she gave no thought to what a proper examination would have revealed-an obvious and serious risk. Her explanations as to her failure would not displace the inference that she was reckless and demonstrated evidence of bad character and a total lack of concern for her patient's welfare. Although the judgment has been praised for following the decision in Rudling, 157 and ensuring the defendant is 'adequately, personally culpable with regard to the death caused', 158 a better view is that such developments incorrectly delineate between criminal liability and exculpation and the courts erred in establishing a prospective test for ascertaining the risk to the patient. 159 In support of Rose, two examples are proffered. In the first, a person fails to carry out their professional responsibilities by examining the patient as in Rose and is consequently unaware of the risk to the patient. This person, unless there is a reasonable excuse for this failure, has deliberately ignored best medical practice but is exonerated precisely because of their ignorance on this account. To state that there is no culpability in these circumstances seem intuitively wrong. 160 Using the proposal advanced here, criminal responsibility would be established as the defendant demonstrated a wanton disregard for the victim's welfare, a reprehensible character flaw. Rose failed to examine the patient's eye, failed to examine the image of the back of the eye, failed to call the patient back for a full examination and falsified records to cover up her wrongdoing. Although the eye condition was rare, there was evidence that she was told the patient had been experiencing headaches and she would have known that examinations can reveal such disorders. This was more than a 'medical mishap'. 161 In the second example, the person performs the examination Rose omitted but because they fail to appreciate the risk to the patient they are appropriately inculpated. It is suggested here that this is not necessarily a disregard of best medical practice and could be an example of a regrettable error. Without evidence that the failure to realise the significance of the examination results is due to a total lack of concern for the patient there should be no criminal liability. Adopting the proposal advanced here that would be the result. On a final point, although Rose follows Rudling, 162 in holding that the risk has to be serious and obvious at the time of the breach, it is difficult to see Rudling was in breach of any duty. On the evidence, experts testified that some doctors might have provided a home visit but others would not. It would be impossible for a GP to see every sick patient, 163 whereas with Rose, she was under a duty to examine those who had eye appointments and the case could have been distinguished on that ground. For the purpose of exculpating Rudling, there was no need to change the test of foreseeability of the risk to a prospective one.
The final case for scrutiny is Bawa-Garba, 164 a junior doctor who had just returned to work after 14 months of maternity leave, who failed to diagnose a case of sepsis in her young patient. The patient had shown a positive response to antibiotics initially, a nurse did not relay relevant symptoms that developed, there was a delay in obtaining blood results and in administering the medication Dr Bawa-Garba had prescribed, there was a failure of the computer system, and this doctor had worked for 12-13 hours that day without any break. These are all factors that support a finding that her failure to diagnose septic shock was not the result of a lack of regard for the welfare of her patient. Applying the provision proposed here, although Dr Bawa-Garba would satisfy (1) and (2), she would not satisfy (3) and would not be criminally responsible.

Conclusion
It is argued here that the appellate courts in Adomako should have adopted a capacity-based approach to reckless manslaughter rather than deciding that gross negligence manslaughter was the correct approach in cases of this nature. The proposal advocated here is able to distinguish more successfully between those who are morally blameworthy and deserving of punishment and those who should be exculpated. In the context of healthcare professionals, it targets conscious departures from good practice as opposed to medical mishaps that have unfortunate consequences. It examines why the inadvertent fail to appreciate a risk and the context in which the proscribed conduct occurred. Where such failure is the manifestation of a reprehensible character flaw criminal responsibility should follow.
This position is clearly grounded in character theory as evidence of a bad character would give rise to criminal liability, it recognises the capacity element in choice theory and is a subjectivised version of 'role' theory. Role theory infers that D is reckless by reference to what a reasonable person performing his role would have foreseen, whereas here it would be referable to the particular agent. If a more objective but capacity-based test for recklessness which includes inadvertence is utilised, (as in practice an objective capacity-based test is already in operation) encompassing both acts and omissions, the offence of gross negligence manslaughter becomes superfluous. From a theoretical standpoint, choice theory alone cannot adequately accommodate liability arising through inadvertence but character theory can assist by highlighting where the inadvertence manifests an indifference to the welfare of others. Consequently, a synthesis of aspects of the choice, character and role theories is the most appropriate basis for grounding culpability and this approach underpins the proposed formulation of liability for reckless inadvertent killing advanced here. A broader, more objective but capacity-based approach which considers both the general and specific capacity of the defendant is needed.