¹ Re Castioni [1891] 1 Q.B. at 167

² [1985]A.C. 560.

³ Roger Smith [1975] A.C. 476 at 494C

⁴ H. L. A. Hart, Essays in Jurisprudence and Philosophy (1983) 367 recognises this rule, but seems to me to overlook it at p. 389, where he takes as a slight advantage of the decision in Roger Smith (to be discussed later) that it exempts the person who tries to kill by voodoo (sympathetic magic). This is a case of attempting to commit a crime by inadequate means, and on any view of the law there has never been any reason why it should not be punishable—unlikely as a prosecution may be. If voodoo became a social menace it would have to be controlled, and the law of attempted murder could well be used. For accounts of “voodoo deaths” resulting from fear see correspondence in the B.M.J. in 1965

⁵ Op.Cit., last note, pp. 377–378.

⁶ Hart attaches some importance to a distinction between what is physically impossible and what is logically impossible. Physical impossibility is, conceptually, worlds apart from illogicality, but I doubt the utility of the distinction in relation to attempts, because it can so easily be a matter of words. The “end” of killing a man is physically impossible if you are stabbing a corpse, but it can equally be said to be logically impossible, since “killing a man” means ending the life of a person, and a “corpse” means a body that is already dead. It is physically impossible to handle goods that are not there, so it is physically impossible to handle stolen goods if no goods are there or if the only goods there are not stolen; but the same propositions can be expressed in terms of logical impossibility, given the meaning of the words used.

⁷ [1975]A.C. 476.

⁸ Law Com. No. 102. The commission's recommendation was at first rejected by the Home Office (which did not show the courtesy, not to mention wisdom, of consulting the Law Commission before arriving at its decision); but after a hearing by the Home Affairs Committee of the House of Commons, at which evidence was given by the Law Commission, and in its support by Professor J. C. Smith and myself, the Home Office withdrew its objection. It did so with rather poor grace, preserving its amour-propre in the face of defeat by subjecting the Law Commission's clear draft Bill to amendments which made it slightly less pungent, and continuing its icy attitude towards the commission by again failing to consult the commission on its amendments. However, the Law Commission was (rightly) satisfied that the revised draft was to the same effect as its own.

⁹ [1985] A.C. at 579E. See more on this below at nn.48. 49.

¹⁰ [1975] A.C. at 492D: “I do not think that the present case turns on the test of proximity at all.” It is true, however, that Lord Reid stated that no act could be proximate to an impossible attempt. See Hart’s criticism, op. Cit., p. 387.

Mark Thornton in 25 Cr. L.Q- 301 suggested a qualification upon the “putative fact” approach to attempts that I had propounded in a previous issue of the same quarterly. He proposed that “proximity must be decided on the facts as the agent supposes them to be except for his beliefs concerning proximity itself.” As an instance of the exception, he wrote: “if the agent believes that his act of voodoo will bring about the victim’s death, then this belief is discounted since it is (presumably!) a mistaken belief about proximity.” My answers arc: (1) the voodoo example is a mistake of fact, not a mistake as to the law relating to proximity. (2) It is a mistake as to the adequacy of the means intended to be used, which the law has never regarded as standing in the way of a conviction of attempt. (3) Suppose the defendant fired at a person whom he believed to be within range, but who was out of range. His mistake is precisely analogous to that of the believer in voodoo, but Thornton would not call it a mistake as to proximity, surely.

¹¹ Op. Cit., n.4 above.

¹² See Law Com. No. 102 pp. 3Set seq.; Hart, op. cit., n.4 above, p. 390. A Canadian Act in very similar terms to the Criminal Attempts Act was held by a Canadian court to cover facts like those in Roger Smith: Delaney (1982) 69 C.C.C. (2d) 276.

¹³ A letter to The Times (28 January 1981) in support of the Law Commission was signed by all the Professors of Criminal Law in England and Wales, with the exception of Hogan. The signatories included J. C. Smith, who had atone time written in defence of the steps formula (see below) but who had later become convinced that it did not provide a practical solution. M. D. Cohen in 144 J.P.N. 713 surveyed various suggestions for half-dealing with the impossibility problem, and rightly rejected them all. as the Law Commission did.

¹⁴ [1984]Crim.L.R. 584.

¹⁵ Hart. op. cit., n.4 above at p. 391 stated that “the whole doctrine of ‘impossibility’ has now been eliminated by the Criminal Attempts Act.” The text-writers also accepted this.

¹⁶ Hogan's suggestion was rejected by J. C. Smith for reasons that I find completely convincing: [1985] Crim.L.R. 45.

¹⁷ Also by myself in 135 N.L.J. 337. This was a reply to Hogan; see his rejoinder in 135 N.L.J. 454.

¹⁸ Contrast this jejune treatment of the issue of policy with Hart's careful discussion, op. cit., n.4 above, pp. 338 et seq.

¹⁹ One very partial practical (but not theoretical) solution would be to limit the whole law of attempts to attempts to commit crimes of a specified degree of gravity, as Norway does. The Criminal Attempts Act takes a small step in this direction by eliminating attempts to commit summary offences. But no limitation could curtail the law so drastically as to exclude attempted handling.

²⁰ See the survey of the literature in the Report. Law Com. No. 102. passim. I discussed the subject in Criminal Law: The General Part (C.L.G.P.) 2nd edn (1961) 633 et seq.: Textbook of Criminal Law. 1st ed. (1978) Chap. 17; and also published a brief summary of the argument for the putative fact theory of attempt (answering in anticipation the arguments afterwards used by the Appellate Committee in Anderton v. Ryan) in the leading legal weekly a week before the lords delivered their opinions (135 N.L.J. 337).

²¹ Caldwell [1982] A.C. 341. discussed in my Textbook of Criminal Law. 2nd ed., 108.

²² A.L.I. Model Penal Code s.2.O2(2)(c) (POD).

²³ Before the Act it was held that on an indictment for stealing (say) a mirror from a woman's handbag, where it was not proved that the defendant took the mirror and he was actually after money, he could not be convicted of an attempt to steal money, because under the statute allowing conviction of an attempt on an indictment for the completed crime the defendant can be convicted only of an attempt to commit the crime charged, and here the defendant did not attempt to steal the mirror because he did not want the mirror: Easom [1971] 2 Q.B. 315. this is still the law. The indictment must be for attempting to steal what the defendant wanted, but it may just say “money in a handbag the property of V.” Such a charge, of course, means “money that the defendant believed to be there.” It will now be no answer to say that there was no such property.

²⁴ See Lords Hailsham and Reid in Roger Smith [1975] A.C. at 495–496; Williams, Criminal Law: The General Part, 2nd ed., 639.

²⁵ Farrance [1978] R.T. R. 225 (attempt); a contrary conclusion was reached shortly afterwards by a differently constituted Divisional Court in Neilson [1978] R.T.R. 232. The court in Farrance said that the attempt would have been successful but for “the intervention of a burnt-out clutch.” “intervention” was hardly the right word, since it appears that the clutch had ceased to function before the act charged as an attempt.

²⁶ If the assissin's intention was not to strangle but to shoot his victim from behind, he would be guilty of possessing a firearm with intent to endanger life (Firearms Act 1968. s. 16). so that his act would be illegal on that account.

²⁷ Hensler, 11 Cox 570.

²⁸ Roger Smith [1975] A.C. at p. 494D.

²⁹ Ibid., at pp. 498–499.

³⁰ Of course, no question of impossibility can arise if the end aimed at was possible when the defendant acted, even though it afterwards became impossible. This proposition may partly solve the difficulty presented by the facts in an American case, where D, intending to influence the result of pending proceeding for divorce, attempted to suborn another to give false evidence of an act of adultery by one of the parties to the divorce proceeding; but this act of adultery was not alleged in the divorce pleadings, and therefore was immaterial to the issue. It was held that this was not an attempted subornation of perjury. See Williams, C.L.G.P. 2nd ed., 635. If the pleadings had not been filed at the date of the attempt, the case should on principle have been decided otherwise, because the adultery in question might possibly have been pleaded. On the facts the pleadings had been filed, so it was perhaps a case of impossibility, but in England since the Criminal Attempts Act it should be a punishable attempt unless there is held to be a “missing circumstance” within the rule about to be stated—a dark question to which no answer will here be offered. An alternative view is that it was not in any case an instance of impossibility, because the pleadings could have been amended to allege the adultery in question. See J. C. Smith, op. cit. n.62 below p. 38.

³¹ Percy Dalton (London)Ltd. (1949) 33 C.A.R. 102.

³² The reason why the sellers invoiced the excessive charge per lb was that they apparently misread the regulations. This caused them to add a sum for expenses that they could lawfully have added had they been wholesalers: but they were retailers. However, mistake of the criminal law is no defence, and the sellers were to be considered in the same way as if they knew that the price per lb they invoiced was excessive.

³³ There was no indication that in making the extra charge the sellers reckoned that it would be covered by additional pears in the boxes they sold. They apparently knew there might be some additional pears (the boxes had been imported—and presumably sold to the defendant company—as containing “46 Ib or better”), but they do not seem to have made any investigation of what additional weight was to be expected. So it was pure luck that when charged and when they bethought themselves of weighing the contents of other boxes in the consignment that still remained in their possession, the sellers found that these generally contained two or three lb more than the 46 lb invoiced, which would bring the price charged for the boxful of pears within the permitted limits per lb. The prosecution could not counter this defence by proving that any of the boxes sold did not contain this extra quantity, so the sellers were able to establish a defence to the charge of the full offence.

³⁴ It is unlikely that the sellers positively believed that the boxes contained no more than the weight stated. Since goods sold cannot legally be under-weight, they must frequently be over-weight, if only by a little; and the sellers must have known this. They were, however, indifferent on the point. They did not positively desire the boxes to contain no more than 46 Ib each, nor (since they had misread the regulations) positively desire that the boxes should contain more. They based their price calculations on the boxes containing 46 lb, and were evidently unconcerned whether the boxes might contain a little extra, for which no charge was made. In these circumstances the sellers must be taken to have been reckless as to whether the price at which they sold would be within the law, because they were not certain that sufficient additional pears would be found in every box to make the sale legal. (The notion of recklessness is a bit artificial in this context, because in applying it one has to assume that the sellers knew the price regulations and properly understood them.) So the further question is raised whether recklessness as to a circumstance (in this case the non-existence of additional pears) is a sufficient mental element for an attempt. In particular, can such recklessness by sufficient even though the prosecution did not prove that the necessary circumstance was present (i.e. that there were no additional pears)? These are moot questions that cannot be discussed here. If the answer to them is a negative, the sellers of the pears did not commit an attempt (because they did not intend to commit an offence, but were only reckless as to it); but the court did not put the quashing of the conviction on this ground. The court seems to have thought that an attempt could be committed whether the defendants intended (in the fullest sense) to commit the offence or whether they were in part reckless as to it.

³⁵ By the Sale of Goods Act 1979, s.30(3), when excessive goods are delivered the buyer may keep them and pay for them at the contract rate. But it is likely that on the facts supposed in the text the court would find a common intent that the buyers should not have to pay extra.

³⁶ [1975] A.C. at 495–496.

³⁷ For a further discussion see Hart, op. cit., n.4 above p. 384, and below at nn.48, 49.

³⁸ Ibid., at 497D.

³⁹ [1975] A.C. 476.

⁴⁰ [1985J2W.L.R. 29.

⁴¹ Or, rather, outside the particular law that the defendant is charged as attempting to infringe. A boy under 14, though he cannot be convicted of rape as perpetrator, can be convicted on the same facts of indecent assault. Obviously, this fact cannot affect the boy's immunity from the law of attempted rape.

⁴² [1975] A.C. at 498C.

⁴³ Ibid., at 491–492.

⁴⁴ Ibid., at 500C. So also Lord Morris.

⁴⁵ Hart, op.Cit., n.4 above at pp. 386–387.

⁴⁶ The perspicacity necessary to see this point had already been provided, if the lords had cared to take advantage of it. I put the Bluebeard hypothetical in the N.L.J. article referred to at n.20 above. The article was evidently available to the lords, because Lord Edmund-Davies referred to it in his dissent. The others took no account of it; why not, is speculative.

⁴⁷ See above at n.9; cp. the discussion at n.37.

⁴⁸ The quotation is from p. 580E. The restrictive inference is based on Lord Roskill's previous sentence, which is narrower and must be taken to colour it: the “innocent action” rule applies only to “erroneous belief.” Blubeard made no mistake.

⁴⁹ Suggestions have been made of better wording that the Act might have been employed (Lord Roskill, [1985] A. C.at 580–581; J. C. Smith in [1985] Crim.L.R. 506; Glazebrook in [1985] C.L.J. 340). While I agree that these alternatives would have made it more difficult for the lords to evade the Act, and would be very likely to be effective, I would reply (1) that the existing wording was perfectly clear, and (2) that no wording can defeat a court that thinks it finds some absurd consequences of the legislation and therefore claims the ability to disregard it at pleasure.

⁵⁰ See the Report at p. 54. See also the explicit rejection of the Appellate Committee's decisions at p. 50.

⁵¹ [1978] A.C. 979.

⁵² Wilson [1984] A.C. 242 (included offences). See [l984]” C.L.J. 290.

⁵³ Law Com. No. 102 p. 31.

⁵⁴ See (1945) 23 Can.B.R. 271, 380.

⁵⁵ On this point, as on others, the lords seem to have accepted uncritically the argument of counsel for the defendant, that her “error is not one of fact but one of law or of mixed law and fact” ([1985] A.C. at 568B). However, a certain degree of hesitation about the argument is suggested by the fact that the lords did not repeat it in express terms. In the rare cases where an attempter's mistake relates to a question of criminal law. he may be acquitted on the ground that a mistaken belief in a criminal prohibition cannot create liability: see 135 N.L.J. 505 n.2.

⁵⁶ So also Lords Morris and Dilhorne in Roger Smith [1975] A.C. at 502B, 506B.

⁵⁷ The point would have arisen in Anderton v. Ryan if the defendant had been charged with attempted theft. A receiver of stolen goods is generally guilty of stealing them as well, but if the recorder that Ryan bought was not stolen she presumably got a good title to it by purchase. So the argument might have been advanced that she could not be guilty of attempted theft of what was in fact her own property. In Roger Smith [1975] A.C. at 489 Lord Hailsham said that a charge of theft or attempted theft ought properly to have succeeded against the defendant. It is not clear whether he based this remark on the assumption that Smith took part in the original theft, or on what Smith did by way of intended handling, but anyway it was not a case where the goods had become the defendant's property. Assuming that Smith could have been convicted of attempted theft of the goods (because they were not his property), it would be strange if Ryan could not have been so convicted because she became the owner of the recorder.

⁵⁸ See the Report at p. 50.

⁵⁹ Above at n.25:.

⁶⁰ [1978] A.C. 979.

⁶¹ Francis Bennion, Statute Law, 2nd ed., 229, 237–243.

⁶² See his article in Auckland Law School Centenary Lectures(1983) 25.

⁶³ Cp. the discussion by J. C. Smith in [1985] Crim.L.R. 506. At p. 39 of his article cited in the preceding note, Smith points out that the vital fact that Roger did not know was that the police had taken possession of the goods. When Roger discovered the truth on that point he was already under arrest, and if he was versed in the law he must then have been relieved to know that the plan had failed to the extent of making him safe from liability for handling stolen goods. So in that sense he was perhaps glad to have failed! But it must have been, in W. S. Gilbert's words, “modified rapture.”

The test as worded above presupposes that the attempt failed only because of the error of fact. If it failed for some other reason, one must ask whether the defendant would have regarded himself as succeeding if his plan had succeeded apart from his error of fact.

⁶⁴ [1975] A.C. at 500B. But see Law Com. No. 102 p. 51.

⁶⁵ Above at n.27.

⁶⁶ See his article cited above n.62.

⁶⁷ [1985] Crim.L.R. 505.

⁶⁸ As William Joyce apparently did.

⁶⁹ (1850) 2 Den. 5, 169 E.R. 407. See Williams, C.L.G.P. 2nd ed., 23.

⁷⁰ It was not exactly a case where the defendant believed his shooting to be criminal, but a case where he did not believe facts to exist that would make it legally innocent. Logically, the latter case is a fortiori. It is worse to shoot where the shooter believes that the other party is not subject to arrest than to shoot at him not knowing whether he is subject to arrest or not.