¹ [1987] A.C. 352.

² [1975] Q.B. 27.

³ Zuckerman, in (1976) 92 L.Q.R. 402Google Scholar; Williams, in (1976) 126 N.L.J. 1032Google Scholar; Williams, C.R. in Well and Truly Tried, ed. Campbell and Waller (1983) 295Google Scholar; Healy, in (1986) 136 N.L.J. 423Google Scholar.

⁴ Healy, in [1987] Crim.L.R. 355Google Scholar; Mirfield, in [1988] Crim.L.R. 19 and 233Google Scholar; Smith, J.C. in (1987) 38 N.I.L.Q. 223Google Scholar; Peiris, in (1987) 7 L.S. 279Google Scholar. Zuckerman, in (1987) 103 L.Q.R. 170Google Scholar regretted the quantum of proof placed upon a defendant by the lords. Birch, in [1988] Crim.L.R. 221Google Scholar was slightly ambivalent. She thought the decision “makes the best of a bad job” (p. 232), and felt able to interpret it as serving a useful function until such time as we have a criminal code—apparently accepting that it would be advantageous for a code to change the rules. On the other hand she was less than enthusiastic about Woolmington, regarding this case as a “fly in the ointment” of statutory construction, and saying that “it may some day be necessary to question the courts' extreme devotion to Woolmington” (a devotion that has not impressed me to the same extent). Miss Birch apparently regards questions of mens rea as “the province” of this case (pp. 225, 226); but it is demonstrable that Woolmington applies also to questions as to external elements—both the actus reus in the narrow sense and the external elements of defences like provocation and self-defence). The only writer to accord Hunt an unqualified welcome was Bennion, in [1988] Crim.L.R. 31Google Scholar, giving the remarkable reason that it clarifies the law. How different our perceptions are!

⁵ [1935] A.C. at 481. The principle of this case was expressly extended to common-law defences in Mancini [1942] A.C. 1Google Scholar (provocation), and it has been applied many times since, not only in Britain but throughout the common-law world. “The accused does not need to establish, in the generality of cases, any defence or fact in order to secure an acquittal”: Sparrow (1962) 46 Cr.App.R. at 289Google Scholar.

⁶ See Fletcher, George P., Rethinking Criminal Law (Boston 1978) 524–532Google Scholar.

⁷ See Smith, op. cit., n.4 above.

⁸ A possible, but not very convincing, answer is that parliamentary counsel lack a ready-made formula for placing an evidential burden on the defendant. Such a formula is in fact quite simple; all that is necessary is an enactment that evidence of x adduced by the prosecution is proof of y in the absence of evidence to the contrary. See Hill [1983] 4 C.C.C. (3d) 519 (Sup.Ct.Can.)Google Scholar; J.S. (1986) 30 C.C.C. (3d) 230Google Scholar. The Supreme Court said that the formula shifts the evidential burden to the defendant, but it would be better to say that proof of x satisfies the prosecution's evidential burden in respect of y.

⁹ See Public Prosecutor v. Yuvaraj [1970] A.C. at 922Google Scholar.

¹⁰ In Hunt [1987] at 385D,Google Scholar Lord Ackner objected to this proposal on the ground that “the discharge of an evidential burden proves nothing—it merely raises an issue.” This is the orthodox view, and I agree that, as a matter of the desirable use of language, it is better not to speak of an evidential burden of proof. But the question now under consideration is not whether a particular use of language is proper or undesirably lax, but the meaning that our legislators (that is to say, the framers of our legislation) actually intended by the words they used. If, where an evidential burden lies on the defendant, the jury acquit, this is because they feel at least a reasonable doubt whether the defendant is guilty. It does not appear to be wholly discordant with ordinary language to say that, in such a case, at least a reasonable doubt as to guilt has been established; and if “established” why not “proved”? It may not be the best use of words, but if there is a reasonable possibility that this is how the words “prove” and “proof” have been intended, and if this interpretation diminishes the likelihood of miscarriages of justice, we should accept it.

¹¹ Cmnd. 4991 of 1982, paras. 140–141; Law Com. No. 143 cl. 17. (I do not follow Bennion's doubt about the latter's proposals expressed in [1988] Crim.L.R. 32 n.11.)Google Scholar The C.L.R.C. proposed that two exceptions should be recognised; these could easily be attended to if the main change were made. The committee also pointed out that if the test were applied too leniently by magistrates the prosecution would be able to appeal. The framers of the Draft Code, overendowed with political caution, proposed only to affect future legislation, and in cl. 17(4) made a fatal saving for “any rule of interpretation whereby the burden of proving a special defence is imposed on the defendant on trial on indictment”. See J. C. Smith's change of mind expressed in op. cit., n.4 above at p. 243. The C.L.R.C.'s draft was greatly preferable, but it would be very important that the words “unless the contrary intention appears” in cl. 8(3)(a) should have the word“expressly” inserted.

¹² [1987] A.C. at 367B.

¹³ Numerous examples could be given, but it suffices to cite Clarence Wilson [1984] A.C. 242Google Scholar.

¹⁴ Although statutory reverse-onus provisions are frequent throughout the common-law world, the Supreme Court of Canada has struck them down (in general) under the Canadian Charter of Rights and Freedoms: Oakes [1986] 1 S.C.R. 103Google Scholar; cp. Baumet (1986) 54 C.R. (3d) 176Google Scholar; Grift (1986) 28 C.C.C. (3d) 120Google Scholar; Renner (1986) 29 C.C.C. (3d) 138Google Scholar; but see Gray (1986) 30 C.C.C. (3d) 234Google Scholar as to a qualification. The decision in Oakes was extraordinarily distinguished by the British Columbia Court of Appeal in Burge (1986) 55 C.R. (3d) 131Google Scholar; see the severe editorial note. Sec also, on the general pros and cons, Peiris in (1985) 5 L.S. 284–285.

¹⁵ On the murky origin of this rule and the doubts still surrounding it sec Zuckcrman, op. cit., n.3 above at pp. 410–413; J. C. Smith, op. cit., n.4 above at pp. 238–239.

¹⁶ The Times, 18 02 1987. Cp. the discussion at n.51 belowGoogle Scholar.

¹⁷ The solution offered by the editors of Cross's Evidence, 6th edn 121, is that the peculiar knowledge rule is confined to cases where the charge contains a negative averment(e.g. no licence), and the defendant is peculiarly well placed to give rebutting evidence (i.e. I have a licence) if any is available. But there are major objections to this. (1) The rules stating when the prosecution are required to aver a negative merely relate to pleading, and should not affect the burden of proof. On a charge of assault or murder the prosecution arc not required to plead the negative of self-defence or (for murder) provocation—which according to Cross explains why the burden of negativing these defences rests on the prosecution. But these matters come into issue as soon as the defendant gives sufficient evidence upon them, and the fact that the prosecution were not required to negative them in framing the charge should be irrelevant. (2) No reference seems to have been made to the negative averment theory in Selby's case, above. The issue was whether the defendant failed to report an accident of which he knew. This involved two questions: did he fail to report the accident, and did he know of it? The first is the subject of a negative averment: he did not report. The second issue is not pleaded, because it does not appear in the wording of the statute; but if it were pleaded it would involve an affirmative: he knew of the accident. On the rule in Cross, the defendant should not have carried the burden of proving that he did not know of the accident. But anyway, is it sensible to load the law of evidence with such intricacies? (3) Even if the statement in Cross is right, it does not answer the first objection to the peculiar-knowledge rule stated in the text.

¹⁸ There would also be much to be said for an enactment requiring the withdrawal from the jury of an issue of mental element if the defendant has not given evidence. See Williams, in [1988] Crim.L.R. 97Google Scholar. Some may recoil from this suggestion on the ground that it violates the “right of silence.” But for an innocent defendant it should be preferable to be put under pressure to give evidence than to be subject to conviction without proof of the offence.

¹⁹ Beckford [1987] 3 W.L.R. 611, 3 All E.R. 425Google Scholar (J.C.P.C).

²⁰ Jones (1883) 11 Q.B.D. 118Google Scholar.

²¹ (1865) L.R. 1 C.C.R. 1. This authority was overlooked by J., Bingham in Audley [1907] 1 K.B. at 387Google Scholar.

²² See Zuckerman, in (1976) 92 L.Q.R. 410Google Scholar.

²³ The limitation of the rule to statutory offences and statutory exceptions appears from a careful reading of the section. It runs thus. “Where the defendant to an information or complaint relies for his defence on any exception, exemption, proviso, excuse or qualification, whether or not it accompanies the description of the offence or matter of complaint in the enactment creating the offence or on which the complaint is founded, the burden of proving the exception [etc.] shall be on him.” This clumsy wording needs to be separated out, as follows, “(a) Where the defendant to an information [criminal] relies on any exception etc. whether or not it accompanies the description of the offence in the enactment creating the offence, the burden etc. (b) Where the defendant to a complaint [civil] relies on any exception etc. whether or not it accompanies the matter of complaint on which the complaint is founded, the burden etc.” It will be seen that paragraph (a) seems to assume that there is an enactment creating both the offence and the exception; from this it may be concluded (though the point is not altogether clear) that the section does not apply to common-law offences with statutory exceptions, or to common-law defences to statutory offences. For the history of the section see Zuckerman, in (1976) 92 L.Q.R. 408ff.Google Scholar; J. C. Smith, op. cit. n.4 above.

²⁴ Section 1.13(9).

²⁵ Smith, op. cit., n.4 above, pp. 231ff.

²⁶ See the cases discussed by Lord Griffiths in Hunt, pp. 371–372 (Taylor v. Humphries and Davis v. Scrace). These cases were decided upon the special rule for summary trials; it will be recollected that for trials on indictment it had been held on at least one important occasion that a proviso did not reverse the burden of proof (n.21 above).

²⁷ The point is well brought out by Mirfield, op. cit., n.4 above, p. 28.

²⁸ See the discussion in Cross, Evidence, 6th edn 120ff.

²⁹ See e.g. Guyll v. Bright (1986) 84 Cr.App.R. 260Google Scholar; Leathley v. Drummond [1972] R.T.R. 293Google Scholar; Edwards, n.2 above.

³⁰ Op. cit., n.4 above at pp. 230, 238–239.

³¹ See n.2 above.

³² Above, n.3.

³³ Cmnd 4991 of 1972, para. 140. Lawton L.J. was a member of the committee and concurred in its Report, but it was also he who delivered the judgment in Edwards.

³⁴ Oxford v. Lincoln, The Times, 1 March 1982.

³⁵ Cp. the survey of cases on the criminal codes of South Asia by , Peiris in (1987) 7 L.S. 279Google Scholar. He sums up by saying that his survey “highlights the conceptual weakness of the distinction, with its attendant practical anomalies which have defied judicial ingenuity for almost a century.”

³⁶ Sexual Offences (Amendment) Act 1976 s. l(l).

³⁷ Morgan [1976] A.C. at 214E–FGoogle Scholar.

³⁸ See J. C. Smith's exposition of the illogicalities of the present law: op. cit. n.4 above at pp. 231ff.

³⁹ See Nimmo's case, cited below at n.55.

⁴⁰ (1944) 60 L.Q.R. at 280Google Scholar. See also the study by Jeffries, and Stephan, in (1979) 88 Yale L.J. 1325CrossRefGoogle Scholar.

⁴¹ This was the way in which I put the point in (1982) 2 L.S. 233. I wrote this article partly as a criticism of the doctrine of reasonable mistake (above at n.19), and had cither missed or failed to remember Stone's L.Q.R. note. I am glad to make belated acknowledgment to him for his insight.

⁴² However, the paradox has perhaps a psychological explanation. Even when a person has been execpted from a class in a given respect, we may still, illogically, regard him as being faintly, or possibly, tarred with the same brush as the other members of the class. Some suspicion still clings to him, as the author of the squib perhaps intended it to attach to Hermann. This may be why we so readily suppose that a person who claims to be exceptcd from liability is sufficiently suspect to be made to bear the burden of establishing that he is exceptional.

⁴³ Offences against the Person Act 1861 s.42.

⁴⁴ Loc. cit. above n.40.

⁴⁵ Zuckerman, , in his influential article on this subject ((1976) 92 L.Q.R. 402)Google Scholar, accepted Stone's point that exceptions cannot be gathered from the arrangement of wording in statutes, but argued that the criminal burden should be distributed according to policy (as Stone had done for the civil burden). At first sight, therefore, it may seem that his proposal is open to the same objection on grounds of justice as Stone's. But earlier in his article Zuckerman, had announced an intention of using “burden of proof” to cover both persuasive and evidential burdens, and it appears from his later note ((1987) 103 L.Q.R. 170)Google Scholar that he intended to propose only that the distribution of evidential burdens should be settled according to policy. Perhaps he intended that his “policy” proposal was to be conditional on acceptance of his proposal that criminal burdens on the defendant should be evidential only. In the result, the lords in Hunt accepted the “policy” proposal but applied it to persuasive burdens, thus achieving the worst solution.

⁴⁶ In particular, Reid, Lord in Sweet v Parsley (1970) A.C. at 150CGoogle Scholar expressed a preference for reverse onuses instead of strict liability. He added: “I find it a little surprising that more use has not been made of this method [i.e. reversing the onus by statute]; one of the bad effects of the decision in this House in Woolmington may have been to discourage its use. ”This is a surprising remark to come from Lord Reid, the paragon of the judges of his time. He did not mention what he thought the other bad effects of Woolmington were. In fact Parliament has frequently reversed the probative burden of a negligence issue in regulatory offences.

⁴⁷ [1986] Q.B. 125.

⁴⁸ See below at n.66.

⁴⁹ Per Lord Griffiths at p. 369, Lord Ackner at pp. 379–380.

⁵⁰ [1942] A.C. 1.

⁵¹ Above at n.16.

⁵² Province of Bombay v. Municipal Corpn of Bombay [1947] A.C. at 61–62Google Scholar. Day J. said in 1887: “There are many cases in which such implication would necessarily arise, because otherwise the legislation would be meaningless; that is what I understand by ‘necessary implication.’” The passage is quoted by Bennion, , Statutory Interpretation (1984) 104Google Scholar (1 have corrected a misprint); but the author challenges it because he holds that there is no special rule of necessary implication, even in respect of the Crown. His opinion is idiosyncratic.

⁵³ At p. 380D. Cp. his argument at p. 385C.

⁵⁴ Note 50 above.

⁵⁵ [1968] A.C. 107.

⁵⁶ At p. 374F. Cp. Lord Ackner at p. 382F.

⁵⁷ Mancini [1942] A.C. at 11Google Scholar.

⁵⁸ Note 55 above.

⁵⁹ Lord Griffiths took the seriousness of the offence as a reason for placing the burden of proof on the prosecution (p. 378A). But he also took the fact that the offence was “absolute” as pointing in the same direction. Now, most offences of strict liability are comparatively minor. What is it about such an offence that makes it a strong candidate for the Woolmington umbrella? If an offence is such that we do not mind convicting people who arc morally innocent, lacking any fault, why is it so specially unthinkable that they should be convicted without convincing proof of the actus reus?

Lord Griffiths's proposition about “absolute” offences is hard to reconcile with his approval of Nimmo's case. The offence charged in that case was “absolute”; yet Lord Griffiths approved of the burden of proof being cast on the defendant. (The offence charged in Nimmo was absolute because it did not allow a defence of no negligence—e.g. that the defendant reasonably but erroneously thought that safe means of access were being provided, or reasonably but erroneously thought that no safe means of access were reasonably practicable.)

⁶⁰ Lord Griffiths at p. 374G; Lord Ackner at p. 383E.

⁶¹ See Nimmo's case, [1968] A.C. at 125EGoogle Scholar.

⁶² Lord Griffiths at p. 372F; Lord Ackner at p. 385G.

⁶³ See above at n.38.

⁶⁴ Bennion, in [1988] Crim.L.R. 35Google Scholar, relying from a highly questionable remark by Lord Bridge.

⁶⁵ Lord Griffiths protested that “Parliament can never lightly be taken to have intended to impose an onerous duty on a defendant to prove his innocence in a criminal case, and a court should be very slow to draw any such inference from the language of a statute.” Apparently the court can readily impose a non-onerous duty on the defendant—but how is the judge to know in any instance that discharge of the duty is going to be onerous or easy? And, reassuring as it is to know that Lord Griffiths was concerned about the position of a possibly innocent defendant, why was he so willing to enlarge the power of a court to reverse the burden of proof in pursuance of a policy of facilitating convictions?

A particular version of the policy argument adopted by Lord Ackner (at p. 383F) was that a burden of proof would be impliedly reversed if the statute would otherwise be “unenforceable” or “burdensome” to enforce. But the judges who use this argument do not demonstrate that putting a persuasive burden on the defendant is much more beneficial to the prosecution than putting an evidential burden on him. See J. C. Smith in 38 N.I.L.Q. 240–242.

⁶⁶ [1987] A.C. 378C.