1. The Discipline of Law (London: Butterworths, 1979) p 292.

2. The term ‘trashing’ has been applied to the more exuberant activities of the critical legal studies movement in destroying the cherished tenets of the conventional formal representation of legal materials so as to disclose the lack of any coherent underpinning to legal doctrine. but, as Neil MacCormick has pointed out, the more sober efforts of the movement are not substantially distinct from the efforts of conventional legal scholarship to reveal the existing inadequacies of legal materials as a preliminary to proposing reform: see N MacCormick ‘Reconstruction after Deconstruction: a Response to Cls’ (1990) 10 Ojls 539. in his article taking a critical perspective on recklessness, Alan Nome shows some sensitivity to the issue of overt incoherence in the law, but claims that this can be attributed (by conventional legal scholars) to ‘false judicial reasoning’, whilst maintaining an ‘implicit logic of the legal categories behind the necessary flux of judicial practice’: a Nome ‘Subjectivism, Objectivism and the Limits of Criminal Recklessness’ (1992) 12 OJLS 45. This is a rather generous allowance for inverted logic to supply from incoherent conclusions coherent premises, and effectively transfers the target of purported coherence from the law to legal scholarship.Google Scholar

3. A more polite, though still derisory, image resorted to by Andrew Ashworth in attempting to describe the appellate decisions on intention was playing a piano accordion: Editorial [1986] Crim LR 1.

4. The XII Tables were placed in the Roman forum c 450 BC.

5. Cf the observation of Diplock Lj, in R v Mowatt [1968] 1 QB 421 at 426: ‘The function of a summing-up is not to give the jury a general dissertation upon some aspect of the criminal law, but to tell them what are the issues of fact on which they must make up their minds.Google Scholar

6. Kenny's Outlines of Criminal Law (1st edn, 1902) J W C Turner (ed) (Cambridge: Cambridge University Press, 16th edn, 1952) p 186. Although Jeremy Horder has provided forceful arguments against the historical basis for the correspondence principle as an aspect of ‘ideal subjectivism’ in ‘Two Histories and Four Hidden Principles of Mens Rea’ (1997) 113 LQR 95. these do not encompass the crucial aspect of a subjective approach to recklessness at issue here: taking into account the individual circumstances of the defendant in determining whether he had actual awareness of the risk. Horder's preoccupation with the correspondence principle as a target distracts his attention from this key point — in his perception of differing viewpoints between Turner and Kenny (at 114–115, 117); of a conflict between Mowatt and Cunningham (at 114); and even of the impact of Cunningham itself (at 117–118). On this key requirement of actual awareness of the risk by the defendant all these authorities are at one — and this does not detract from the possibility of disagreement as to the issue of what is the subject matter of the appreciated risk required for each offence - nor denigrate Horder's suggestion that the correspondence principle is not the appropriate answer.

7. Caldwell, R. v.. [1981] 1 All ER 961 at 964.Google Scholar

8. It is possible to define words at both of these levels, such that there is consistency at the general level whilst conflict between meanings at the level of particular instances. Take, for example, ‘a cry’, which at the general level can be defined as a loud utterance, but within particular instances can be variously, an expression of pain, a call for help, etc. I consider this in greater detail in relation to ‘claim’ in ‘More Comments on Rights and Claims’ (1991) 10 Law and Philosophy 271 (also found as ch IV of A Halpin Rights and Law - Analysis and Theory, (Oxford: Hart Publishing, 1997)). The mere identification of a general synthesising definition does not close our inquiry unless it is at that level that our concerns are expressed.Google Scholar

9. The same error is perpetrated by Lord Hailsham in Lawrence, R.V.. [198] 1 All ER 974 at 978. In following Lord Diplock in taking the ordinary language, or dictionary meaning, Lord Hailsham asserts that the word, though varying in pronunciation over the centuries, has borne ‘the same meaning’ (emphasis added). But in the words immediately following, which amount to an attempt to amplify that meaning, Lord Hailsham simply informs us that it is ‘applied to a person or conduct evincing a state of mind stopping short of deliberate intention, and going beyond mere … carelessness’. Again, the device of a range is used which might cover a number of disparate instances without any attempt at identifying a synthesising feature. More remarkably, Lord Hailsham's range ends before ‘mere carelessness’, but carelessness is to be found within the range of dictionary meanings given for recklessness, as Lord Diplock points out in the same case (at 981).Google Scholar

10. The exercise was performed by the Court of Criminal Appeal in Cunningham in suggesting how the jury should have been directed in that case: ‘he foresaw that the removal of the gas meter might cause injury to someone but nevertheless removed it’ ([1957] 2 All ER 413 at 415).Google Scholar

11. [1981] 1 All ER 961 at 964.Google Scholar

12. More rigorously, it could also be taken back to the main subject of the preceding sentence, ie the word ‘recklessness’, but as this is found qualified in the relative clause with its popular or dictionary meaning, this amounts to the same thing.

13. The clause underwent a remarkable transformation on the very same day in the House of Lords, when the speeches in Lawrence were given, which unanimously purported to follow Caldwell, yet provided in the corresponding clause a presupposition that the risk would have been apparent to ‘an ordinary prudent individual’, making a switch from subjective to objective ([1981] 1 All ER 974 at 982). The subjective form of the original clause in Caldwell was ammunition for academic commentators who sought to put a qualified subjectivity on the Caldwell direction — for a recent survey of the discussion, see D W Elliott ‘Endangering Life by Destroying or Damaging Property’ [1997] Crim LR 382 at 383–387; and for persistence in the qualified subjectivity reading, see L H Leigh ‘Liability for Inadvertence: A Lordly Legacy?’ (1995) 58 MLR 457 at 462–463. I consider below the further mischievous impact of the clause in Elliott v C.Google Scholar

14. .[1981] 1 All ER 961 at 966.Google Scholar

15. Ibid at 967.

16. [1981] 1 All ER 974 at 982, 983.Google Scholar

17. [1983] 2 All ER 1005.Google Scholar

18. It was followed by the Court of Appeal in R V Malcolm R (1984) 77 Cr App R 334 and in R v Bell [1984] 3 All ER 842.Google Scholar

19. Above n 13.

20. Ibid.

21. [1983] 2 All ER 1005 at 1010.Google Scholar

22. Ibid at 1011.

23. Ibid at 1011.

24. Ibid at 1011.

25. As in fact occurred before the magistrates, who were persuaded that it was the proper reading of the Caldwell test.

26. [1983] 2 All ER 1005 at 1011–1012.Google Scholar

27. See above text accompanying n 11.

28. ‘The correct reading is provided by Lord Ackner in Reid, R. v.. [1992] 1 WLR 793 at 805 but, remarkably, their Lordships in that case fail to discuss Elliott v C (despite its having been cited in argument).Google Scholar

29. It would also have been impossible for Lord Diplock himself to make the subjective-objective switch in the presupposition clause in Lawrence (see above n 13).

30. It should not be overlooked that the first of Lord Diplock's favoured instances unequivocally adopts the Kennyl Cunningham approach to recklessness, but not as a definition, for this would knock out the other instance. The ability of this strong subjective type of recklessness to trail alongside whatever is made of Lord Diplock's second instance is uncontroversial, and it is common practice to focus on the second instance as distinctively bearing the Caldwell brand.

31. Reid, R. v.. [1992] 1 WLR 793.Google Scholar

32. ‘The Caldwell direction had been regarded as a model one, in the sense of governing how juries should be directed, for criminal damage in Elliott v C [1983] 2 All ER 1005, particularly by Goff LJ at 1010–1011. The Lawrence direction had been held by the Court of Appeal to be a model one for reckless driving in R v Madigan (1982) 75 Cr App R 145 at 148. Both these positions are overruled by Reid. Given the switch in Lord Goff s own attitude from Elliott v C to Reid, it is again surprising that no mention is made by their Lordships in the latter case of the former.Google Scholar

33. [1992] 1 WLR 793 at 796 (Lord Keith), 805 (Lord Ackner), 813–814 (Lord Goff), 819 (Lord Browne-Wilkinson, also concurring with Lord Goff); Lord Roskill concurring with all the other speeches. Much is made in these speeches of Lord Diplock referring to his direction as ‘an appropriate instruction’ (emphasis supplied by Lord Goff at 813). That this did not bear the indefinite connotation that their Lordships seek to place upon it is evident from the full description that Lord Diplock gives: ‘an appropriate instruction to the jury on what is meant by driving recklessly’ (emphasis added), rather than an appropriate instruction for the circumstances of this particular case: see [198] 1 All ER 974 at 982.Google Scholar

34. ‘Lord Browne-Wilkinson does state that he does not propose to address Caldwell in his own comments ([1992] 1 WLR 793 at 817) but does concur with Lord Goff (at 816), who, in common with Lord Ackner, does embrace Caldwell (at 807, 804). Both Lord Keith and Lord Roskill concur with Lord Ackner and Lord Goff.Google Scholar

35. Ibid at 805, 807, 814, 817.

36. Ibid at 796, 799, 806, 813, 814–815, 816, 819.

37. Ibid at 796, 804, 807, 815, 819.

38. Ibid at 807, 815.

39. See above nn 35 and 36.

40. To give only one prominent example, the conflict between Stephenson and Caldwell.

41. This may involve stipulating one or a number of legal meanings. So long as it is clear which meaning(s) apply to which offence, this still effectively removes the problem.

42. Elliott v C differs from Cunningham in stipulating two legal meanings for recklessness: leaving the Cunningham meaning for offences involving malice, and adding a hard objective meaning for criminal damage (and other offences?).

43. The correct inference to draw from this is, I think, that Reid has overruled Elliott v C. This conclusion is supported more specifically by dicta of Lord Keith at 796 and Lord Ackner at 805 (see below n 51 and text at n 52): for comment see Leigh (above n 13) at 465. But the picture is not as clear as it might have been due to the remarkable omission of their Lordships in Reid to make an explicit reference to Ellioft v C: see further above nn 28 and 32.

44. [1992] 1 WLR 793 at 796.Google Scholar

45. Ibid at 806.

46. Ibid at 811, 813.

47. Ibid at 813, 816.

48. Ibid at 819.

49. Lawrence, R. v. [1981] 1 All ER 974 at 982: ‘… the defendant was in fact driving the vehicle in such a manner as to create an obvious and serious risk.’ Note also the preceding paragraph for further amplification of these words.Google Scholar

50. See Smith and Hogan, Criminal Law (London: Butter worths, 8th edn, 1996) 64; A Ashworth Principles of Criminal Law (Oxford: Clarendon Press, 2nd edn, 1995) 177.

51. Lord Keith, Proposes a wide formula: ‘where his capacity to appreciate risks was adversely affected by some condition not involving fault on his part’ ([1991] 1 WLR 793 at 796).Google Scholar

52. See further below n 90.

53. Though not the other cases that Lord Browne-Wilkinson mistakenly runs together.

54. For further discussion, see Smith and Hogan (above n 50) pp 69–70.

55. [1937] AC 576.

56. Ibid at 593. For further discussion, see below n 69 and text thereat.

57. Clear evidence of this state of confusion can be found in the observation by Lord Taylor CJ in Prentice, R. v. [1993] 3 WLR 927 at 936 (considered in the following section): ‘the effect… has been to create conflicting approaches and uncertainty… The diversity of views is illustrated by the stances adopted in these three appeals which have not been consistent even among counsel for the Crown on the one hand and those for the defence on the other.’.Google Scholar

58. [1994] 3 WLR 288.Google Scholar

59. Ibid at 296.

60. Ibid at 296, 297.

61. Ibid at 296, 297–298.

62. ‘The Court of Appeal's decision in Prentice [1993] 3 WLR 927, in which Adomako joined three other appeals involving cases of gross negligence manslaughter.Google Scholar

63. R v Andrews [1937] AC 576.

64. Seymour, R. v. [1983] 2 AC 493 — to be overruled in Adomako so as to remove the motor manslaughter exception.Google Scholar

65. [1993] 3 WLR 927 at 936–937.Google Scholar

66. In Lord Atkin's famous passage: ‘Simple lack of care such as will constitute civil liability is not enough: for purposes of the criminal law there are degrees of negligence: and a very high degree of negligence is required to be proved…’ ([1937] AC 576 at 583).

67. [1993] 3 WLR 927 at 942; cf at 950–951.Google Scholar

68. Significantly (at 936), Lord Taylor suggests that the ‘modifying effects’ of Reid have ‘brought the Lawrence/Caldwell approach closer to the Andrews gross negligence test’.

69. Lord Atkin in Andrews [1937] AC 576 at 583; Lord Taylor CJ in Prentice [1993] 3 WLR 927 at 933, 936, 947. There may be more to analyse here than simply closing the culpable sector of the lacuna, though that is certainly included. Part of the complexity relates to which risk we make the subject of the test. As Lord Taylor points out (at 936), the initial risk to the patient in an emergency department is not created by the doctor's conduct but rather causes the doctor to respond. Nothing the doctor does can amount to taking that recognised risk. However, if the doctor's response is incompetent, then that creates a further risk to the patient. If the doctor is ignorant of his incompetence and so fails to consider this second risk, or if (less likely) the doctor is aware of it but goes ahead anyway, then either way he can be caught by the Diplock test. But there may be cases where the doctor is aware of the risk to the patient of adopting the wrong procedure, seeks to avoid it but, with gross negligence, still gets it wrong. This is the standard lacuna case that would still be caught by the gross negligence test. There is a further nice point to raise about the lacuna analysis itself. Can one not always proceed to a further level of risk, which the defendant has failed to consider? In the example considered, the doctor is aware of (1) the risk to the patient's health which brought him into the hospital, and (2) the further general risk of his choosing the wrong treatment, but fails to consider (3) the risk obvious to a competent doctor of choosing that particular wrong treatment which he erroneously believes to be correct. Leigh (above n 13) at 463 and 466–467, also concludes that the lacuna does not exist on the grounds of: (a) Lord Diplock's lack of a comprehensive test for recklessness; (b) recklessness in the assessment of there being no risk (rather than in the subsequent course of conduct) - differing in perspective though not in impact from the view suggested here.

70. [1993] 3 WLR 927 at 936.Google Scholar

71. As must be implicit in Lawrence in dealing with an ordinary prudent motorist. Cf the Law Commission proposal in this respect, which in the proposed gross carelessness variant of involuntary manslaughter makes explicit reference in cl 2(2) to the accused's actual knowledge and any skill or experience he professed to have: Legislating the Criminal Code: Involuntary Manslaughter (Law Com no 237, HC 171, 1996).

72. [1993] 3 WLR 927 at 937.Google Scholar

73. This repeats Lord Atkin's view in Andrews [1937] AC 576 at 583. Of the four states of gross negligence given by Lord Taylor ([1993] 3 WLR 927 at 937, 952): (a) and (b) amount to the two limbs of Caldwell recklessness with the important qualification of indifference added to (a); (c) covers the lacuna subject to there being a high degree of negligence; and (d) covers gross negligence in performing a duty. The list is explicitly stated to be not exhaustive. For further comment, see G Virgo ‘Back to Basics - Reconstructing Manslaughter’ (1994) 53 CLJ 44 at 46ff.

74. [1993] 3 WLR 927 at 937.Google Scholar

75. The four cases, (above n 73) are offered ‘without purporting to give an exhaustive definition’ but it is still an attempt at definition. The point is reinforced by the subsequent observation on the distinction between two of the cases identified (at 952).

76. [1993] 3 WLR 927 at 947.Google Scholar

77. [1994] 3 WLR 288 at 292. The argument mounted by counsel for Adomako ([1995] 1 AC 171 at 175–177) suggested that the ‘gross’ qualification on negligence was insufficient to capture the requisite state of mind and could be satisfied by what was regarded as a particularly bad case of ‘ordinary human error or thoughtlessness’. The state of mind required for recklessness comprised ‘actual disregard of, or indifference to’ the risk - which, given counsel's reliance on the Seymour adoption of the Caldwell-Lawrence test for manslaughter, seems to be an attempted paraphrase of that test. Counsel for the Crown argued (at 180) that the Culdwell-Lawrence test was not appropriate. Ironically, in the three successful appeals in the Court of Appeal the allegiances of defence and prosecution towards the two conflicting tests had been the other way round.Google Scholar

78. [1994] 3 WLR 288 at 297.Google Scholar

79. Ibid at 297.

80. See above, in particular nn 73 and 75 and text thereat.

81. [1994] 3 WLR 288 at 296, 297.Google Scholar

82. If this last sentiment is fully adopted, it is difficult to see what remains of case law, given that it is precisely the enunciation of general principles out of particular contexts on which it is based — on which, see J Finnis Natural Law and Natural Rights (Oxford: Clarendon Press, 1980) p 269.

83. [1994] 3 WLR 288 at 297 — in the answer to the certified question.Google Scholar

84. Ibid at 298. Further obfuscation also remains as to the current status of the lacuna, entrenching that found in Reid, since the advance made by the Court of Appeal in Prentice in dealing with this depended on a recognition of the distinction between gross negligence and recklessness.

85. [1995] 1 Cr App R 157.Google Scholar

86. Ibid at 167.

87. Ibid at 168.

88. Ibid at 163 (emphasis added).

89. See above, text at nn 8–9 and 11–12. The confusion in appellation (from which easily proceeds an error of substance) is common. Lord Mackay in Adomako [1994] 3 WLR 288 at 297 refers to ‘the definition of recklessness in R v Lawrence’.Google Scholar

90. See above, text at n 21 and following n 22; and for criticism of the failure to draw on Lord Diplock's insistence on equivalent culpability, Leigh's third proposition (above n 13) at 462. the contrast in culpability is also evident in Hobhouse Lj's judgment where, on the same page, he cites Lord Diplock's requirement in Miller that the risk ‘must be one that would be obvious to anyone’ (emphasis added) followed by the justices' determination in Elliott v C that the risk ‘would not have been obvious to her’ ([1995] 1 Cr App R 157 at 166). One must assume that directions based on such expansive requirements are dealing with the class of people with a normal capacity to appreciate risk, and accordingly applicable only to such persons. Otherwise, the existence of a single person to whom the risk was not obvious would render them nugatory.Google Scholar

91. [1995] 1 Cr App R 157 at 169. Since he had set fire to hay on which his friends were sleeping in a barn and had in evidence admitted to having taken the risk that they might not have woken up in time and got up, the case is quite different from Elliott v C, in which the justices found the defendant had not the capacity to appreciate the risk to her own life (at 160–161 and 166).Google Scholar

92. [1995] 1 Cr App R 157 at 165, 169. See above n 45.Google Scholar

93. A singular dictionary definition should not be equated with a singular meaning for the word. Dictionary definitions frequently contain a range of disjunctive meanings for the word being defined, as the entries for a great number of common words reveal.

94. Notable further examples are provided by the disputes that have occurred as to what form of recklessness is required for lesser assaults and rape.

95. See entries in the Oxford English Dictionary.

96. See the problems faced by Lord Goff on this point in Reid [1992] 1 WLR 793 at 812, 815. His final resort to a requirement of mens rea (‘a certain state of mind’) at 815 leaves the job unfinished, since the requisite state of mind has not been properly defined. See also the comments of Lord Atkin in Andrews [1937] AC 576, who traces the exclusion of ‘mere inadvertence’ as a historical development (at 581–582) but does not consider that the use of mens rea is helpful in making the distinction (at 583).Google Scholar

97. If done with sufficient indication that the grossness is an evaluation of all the circumstances that the particular defendant was in, this need not be insensitive to the capacities of particular defendants to deal with risk: see the Law Commission proposal in Law Com no 237, discussed above n 71. The qualification to be made by the tribunal of fact in this proposal is rendered by ‘far below what can reasonably be expected of him in the circumstances’: cl 2(1)(c)(i). It also overcomes the lacuna problem: see above n 69 and text thereat.

98. For further comment, see Gardner, S ‘Manslaughter by Gross Negligence’ (1995) 111 LQR 22.Google Scholar

99. A relatively recent recognition of this connection is to be found in Lord Diplock's equation of intention with foresight of a probable outcome of one's conduct in R v Hyam [1975] AC 55 at 86. He finds the common element to be ‘willingness to produce the particular evil consequence’. This is inaccurate, for the latter state of mind may not include this element (see further A Halpin ‘Intended Consequences and Unintentional Fallacies’ (1987) 7 OJLS 104 at 108). But both states of mind do share a willingness to produce the risk of the evil consequence, or, as I have put it, a conscious decision to expose to harm. It is clear that this lesser element did once suffice for intention in the criminal law. As Lord Edmund-Davies noted in Caldwell [1981] 1 All ER 961 at 970, Austin recognised the modem (Cunningham) recklessness as intention: see J Austin Lectures on Jurisprudence (London: John Murray, 5th edn, 1885 reprinted 1929) vol I, pp 428–429, where intention is defined as involving ‘a state of consciousness’ as to a future event whose probability may be rated ‘higher or lower’. This is seen to include subjective recklessness in the example that follows of pistol practice resulting in the undesired wounding of a passer-by, whom I am conscious ‘may chance to be there’. Significantly, Austin informs his students that ‘every wrong … supposes intention or negligence’ (p 344), marking a division based wholly on the presence or absence of consciousness of the potential harm in the mind of the wrongdoer (p 428), for which an independent notion of recklessness is unnecessary. Austin's paradigm of the pistol practice, incidentally, is evidence against Horder's depiction of ‘the malice principle’ as requiring conduct directed at an interest of the victim (above n 6) at 96. The use of intention to embrace subjective recklessness is also found in Kenny (above n 6) in the first edition, p 148, prior to its separation in subsequent editions in the form approved by Cunningham..

100. See above the latter part of n 96.

101. Cf the early proposals for codification which drew up definitions of ‘recklessness’ and ‘heedlessness’: Codification of the Criminal Law (Law Corn no 143, Hc 270, 1985) in cl 22 of the draft code, which did not survive the revised proposals, and A Criminal Code for England and Wales (Law Com no 177, Hc 299,1989) in cl 18 of the draft code. However, para 8.20 of the commentary on that proposal holds out the possibility of ‘providing further key terms’. the need for such an additional fault term was proposed following Reid by Leigh, L H ‘Recklessness after Reid’ (1993) 56 MLR 208 at 212, 217–218, and both recklessness and gross carelessness are found in the proposals in Law Com no 237 - though their relationship is unclear (see below n 104).CrossRefGoogle Scholar

102. Interesting comparative material on the use of conscious exposure to harm as the fundamental requirement for mens rea is provided by the new Israeli Penal Code (in force since August 1995). the text of the draft and enacted code together with papers from a colloquium on the draft proposal are to be found in (1996) 30 Israel Lr. There is a general requirement of ‘awareness’ in the mens rea provision in s 20 of the Code (ibid at 13), which is constructed to cover both intention and variants of recklessness. for further discussion of more detailed problems raised by this provision, see B Burkhardt ‘Some Questions and Comments on What is Called “the Mental Element of the Offence”’ ibid 82 at 84–93 (the article also contains comparative references to the English proposed code in Law Corn no 177 and the German draft E 1962). for general discussion in favour of restricting recklessness to cases where there is awareness of risk, see J Brady ‘Recklessness’ (1996) 15 Law and Philosophy 183. This position suffered a severe setback in Reid, in which it was abandoned mid-argument by counsel ([1992] 1 WLR 793 at 800, 804): for comment see S Gardner ‘Recklessness Refined’ (1993) 109 LQR 21.Google Scholar

103. [1991] 3 WLR 914 at 924, upholding (on that point) the Court of Appeal in R v Spratt [1990] 1 WLR 1073 at 1082–1083. In their note prior to the Lords' decision, Ashworth and Campbell pointed out the existence of other authority on the point derived from cases relating mistake to the negation of mens rea: A Ashworth and K Campbell ‘Recklessness in Assault — and in General?’ (1991) 107 LQR 187 at 190. This is an important implication of the choice of recklessness to take into account, which as the authors observe (at 188) was lacking in Caldwell. For arguments in favour of a broader approach to recklessness for the assault offences, see Clarkson, C M V ‘Violence and the Law Commission’ [1994] Crim LR 324 at 330–331.Google Scholar

104. The switch from conscious exposure to harm to culpable exposure to harm can be made in two ways: the first is to argue that some form(s) of inadvertent exposure to harm are equally as culpable as conscious exposure; the second is to acknowledge the greater culpability of the latter but to argue that the threshold of culpability also accommodates the former. There is a vacillation between these two positions in Law Corn no 237 (above n 71) — contrast paras 4.12–4.16, which justify bringing in inadvertence by arguing for an equivalence in culpability, with para 5.7, which justifies a separation of the offences of reckless killing and killing by gross carelessness on the basis of ‘a clear distinction, in terms of moral fault’ between advertence and inadvertence. It is important to clarify which of the two positions is being maintained, since the second position effectively lowers the threshold of culpability in a way which renders redundant argument over the special status of conscious exposure to harm in determining liability for serious criminal offences.