Woolmington v DPP

Counsel for the Director of Public Prosecutions advanced consequentialist arguments of an alarmist nature: he predicted great uncertainty if section 6 (1) is interpreted as I have suggested. That is not how the matter struck the Lord Chief Justice in R v Director of Public Prosecutions, Ex parte Kebilene [2000] 2 AC 326 , at 341A-B; see also my judgment at 367-368C; Lord Slynn of Hadley, 362A-B; and Lord Cooke of Thorndon was in general agreement: at 372C. Moreover one is reminded of the unfounded predictions that the 1998 Act would cause chaos in our legal system. A healthy scepticism ought to be observed about practised predictions of an avalanche of dire consequences likely to flow from any new development. My view is that, if section 6 (1) is interpreted as I have suggested, the orderly development of convention principles in our country will be advanced.

It follows that I would hold that the House has jurisdiction to rule on the substantive issues.

The approach of the common law to the presumption of innocence was memorably stated by Viscount Sankey LC in Woolmington v D.P.P. 1935 AC 462 , 481 to be that "Throughout the web of the English criminal law one golden thread is always to be seen, that it is the duty of the prosecution to prove the prisoner's guilt. . . " The Woolmington principle was, however, subject to Parliament legislating to the contrary. It is a fact that the legislature has frequently and in an arbitrary and indiscriminate manner made inroads on the basic presumption of innocence. Ashworth and Blake (The Presumption of Innocence in English Criminal Law, 1996 Crim. L.R. 306, at 309) found 219 examples, among 540 offences triable in the Crown Court, of legal burdens or presumptions operating against the defendant. They observed that no fewer than 40% of the offences triable in the Crown Court appear to violate the presumption. In 1972 a most distinguished Criminal Law Revision Committee had observed that "we are strongly of the opinion that, both on principle and for the sake of clarity and convenience in practice, burdens on the defence should be evidential only": Eleventh Report, Evidence (General) Cmnd 4991 of 1972, para 140 . Nevertheless, the process of enacting legal reverse burden of proof provisions continued apace.

In the meantime the human rights movement came into existence. The foundation of it was the Universal Declaration of Human Rights (1948), which has been the starting point of subsequent human rights texts. In article 11(1) it provided: "Everyone charged with a penal offence has the right to be presumed innocent until proved guilty according to law . . ." Borrowing this language almost verbatim, article 6.2 of the European Convention for the Protection of Human Rights and Fundamental Freedoms (1950) provided: "Everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law". Article 14.2 of the International Covenant on Civil and Political Rights (1966) which was signed by the United Kingdom in 1966 is to the same effect. Nevertheless, and despite the right of petition to the European Court of Human Rights created for the United Kingdom in 1961, there was no constraint in our domestic law to legislative incursions on the presumption of innocence. But by the 1998 Act Parliament has provided that, subject to the ultimate constitutional principle of the sovereignty of Parliament, inroads on the presumption of innocence must be compatible with article 6.2 as properly construed. If incompatibility arises, the subtle mechanisms of the 1998 Act come into play.

In H M Advocate v McIntosh, P.C. (5/2/2001) Lord Bingham of Cornhill recently referred to the judgment of Sachs J of the South African Constitutional Court in State v Coetzee [1997] 2 LRC 593 . It is worth setting out the eloquent explanation by Sachs J of the significance of the presumption of innocence in full [para 220 at 677]:

Counsel for the appellant submitted that the defence put forward by the appellant under section 28 is an ingredient of the offence under section 5 (3). His argument was that knowledge of the existence and control of the contents of the container is the gravamen of the offence for which the legislature prescribed a maximum sentence of life imprisonment. The contrary argument advanced on behalf of the Director of Public Prosecutions relied on the observation of Lord Woolf CJ in the Court of Appeal [2001] 2 WLR at 221F that "What the offence does is to make the defendant responsible for ensuring that he does not take into his possession containers which in fact contain drugs." Taking into account that section 28 deals directly with the situation where the accused is denying moral blameworthiness and the fact that the maximum prescribed penalty is life imprisonment, I conclude that the appellant's interpretation is to be preferred. It follows that section 28 derogates from the presumption of innocence. I would, however, also reach this conclusion on broader grounds. The distinction between constituent elements of the crime and defensive issues will sometimes be unprincipled and arbitrary. After all, it is sometimes simply a matter of which drafting technique is adopted: a true constituent element can be removed from the definition of the crime and cast as a defensive issue whereas any definition of an offence can be reformulated so as to include all possible defences within it. It is necessary to concentrate not on technicalities and niceties of language but rather on matters of substance. I do not have in mind cases within the narrow exception "limited to offences arising under enactments which prohibit the doing of an act save in specified circumstances or by persons of specified classes or with specified qualifications or with the licence or permission of specified authorities"; R v Edwards [1975] QB 27; R v Hunt [1987] AC 352; section 101 of the Magistrates' Courts Act 1980. There are other cases where the defence is so closely linked with mens rea and moral blameworthiness that it would derogate from the presumption to transfer the legal burden to the accused, e.g. the hypothetical case of transferring the burden of disproving provocation to an accused. In R v Whyte (1988) 51 DLR 4th 481 the Canadian Supreme Court rejected an argument that as a matter of principle a constitutional presumption of innocence only applies to elements of the offence and not excuses. Giving the judgment of the court Dickson CJC observed (at 493):

It is now necessary to consider the question of justification for the legislative interference with the presumption of innocence. I am satisfied that there is an objective justification for some interference with the burden of proof in prosecutions under section 5 of the 1971 Act. The basis for this justification is that sophisticated drug smugglers, dealers and couriers typically secrete drugs in some container, thereby enabling the person in possession of the container to say that he was unaware of the contents. Such defences are commonplace and they pose real difficulties for the police and prosecuting authorities.

That is, however, not the end of the matter. The burden is on the state to show that the legislative means adopted were not greater than necessary. Where there is objective justification for some inroad on the presumption of innocence the legislature has a choice. The first is to impose a legal burden of proof on the accused. If such a burden is created the matter in question must be taken as proved against the accused unless he satisfies the jury on a balance of probabilities to the contrary: The Eleventh Report of the Cr.L.R.C., para 138. The second is to impose an evidential burden only on the accused. If this technique is adopted the matter must be taken as proved against the accused unless there is sufficient evidence to raise an issue on the matter but, if there is sufficient evidence, then the prosecution have the burden of satisfying the jury as to the matter beyond reasonable doubt in the ordinary way: The Eleventh Report of the Cr.L.R.C., para 138. It is important to bear in mind that it is not enough for the defence merely to allege the fact in question: the court decides whether there is a real issue on the matter: The Eleventh Report of the Cr.L.R.C., para 138. A transfer of a legal burden amounts to a far more drastic interference with the presumption of innocence than the creation of an evidential burden of the accused. The former requires the accused to establish his innocence. It necessarily involves the risk that, if the jury are faithful to the judge's direction, they may convict where the accused has not discharged the legal burden resting on him but left them unsure on the point. This risk is not present if only an evidential burden is created.

The principle of proportionality requires the House to consider whether there was a pressing necessity to impose a legal rather than evidential burden on the accused. The effect of section 28 is that in a prosecution for possession of controlled drugs with intent to supply, although the prosecution must establish that prohibited drugs were in the possession of the defendant, and that he or she knew that the package contained something, the accused must prove on a balance of probabilities that he did not know that the package contained controlled drugs. If the jury is in doubt on this issue, they must convict him. This may occur when an accused adduces sufficient evidence to raise a doubt about his guilt but the jury is not convinced on a balance or probabilities that his account is true. Indeed it obliges the court to convict if the version of the accused is as likely to be true as not . This is a far reaching consequence: a guilty verdict may be returned in respect of an offence punishable by life imprisonment even though the jury may consider that it is reasonably possible that the accused had been duped. It would be unprincipled to brush aside such possibilities as unlikely to happen in practice. Moreover, as Justice has pointed out in its valuable intervention, there may be real difficulties in determining the real facts upon which the sentencer must act in such cases. In any event, the burden of showing that only a reverse legal burden can overcome the difficulties of the prosecution in drugs cases is a heavy one.

A new realism in regard to the problems faced by the prosecution in drugs cases have significantly reduced their scope. First, the relevant facts are usually peculiarly within the knowledge of the possessor of the container and that possession presumptively suggests, in the absence of exculpatory evidence, that the person in possession of it in fact knew what was in the container. This is simply a species of circumstantial evidence. It will usually be a complete answer to a no case submission. It is also a factor which a judge may squarely place before the jury. After all, it is simple common sense that possession of a package containing drugs will generally as a matter of simple common sense demand a full and adequate explanation. Secondly, the statutory provisions enabling a judge to comment on an accused's failure to mention facts when questioned or charged has strengthened the position of the prosecution: section 34 of the Criminal Justice Act 1994 . Thirdly, I turn to the fears centred on the ability of an accused in a drugs case to manipulate the system by providing a mixed statement containing a self-serving explanation that he did not know what was in the package. The perceived difficulty is that the whole statement may be introduced as evidence and he may not testify. In the leading case of Duncan (1981) 73 Cr App R 359, Lord Lane CJ observed (at 365):

Returning to the relative merits of the transfer of a legal burden on an important element or issue to the accused, as opposed to the creation of a mere evidential burden, there have been noteworthy developments in England and in cognate legal systems. In R v Director of Public Prosecutions, Ex Parte Kebilene [2000] 2 AC 326 in the Divisional Court Bingham LCJ had no doubt that, in the context of a serious offence (terrorism), a reverse legal burden of proof provision on a matter central to the wrongdoing alleged against the defendant would breach article 6.2. On the appeal to the House a majority suggested that, once the 1998 Act was in force, reverse legal burden provisions may have to be interpreted as imposing merely an evidential burden on the defendant. Responding to Kebilene Parliament enacted the Terrorism Act 2000 which in section 118(1) and (2) provides that the reverse onus of proof is satisfied if the person adduces evidence which is sufficient to raise an issue with respect to the matter unless the prosecution can prove the contrary beyond reasonable doubt. Comparative experience in constitutional democracies underlines the vice inherent in transfer of legal burden provisions, and the utility, in appropriate contexts, of evidential presumptions. This distinction has been explored in depth in the context of the presumption of innocence contained in section 11(d) of the Canadian Charter of Rights and Freedoms in judgments of the Canadian Supreme Court: see in particular R v Oakes (1986) 26 DLR (4th) 200; R v Whyte (1988) 51 DLR 4th 481; R v Downey (1992) 2 SCR 10; R v Osolin [1993] 4 SCR 595; and Hogg, Constitutional Law of Canada, 4th ed., 1997, 1178-1183. For present purposes the interest lies not in the results of these decisions, but in the approach enunciated in respect of reverse burden provisions and evidentiary presumptions. The view has prevailed that if by the provisions of the statutory presumptions, an accused is required to establish, that is to say prove or disprove, on a balance of probabilities either an element of the offence or an excuse, then it contravenes section 11(d). Such a provision would permit a conviction in spite of a reasonable doubt: Osolin, 29, per Cory J (majority judgment.) On the other hand, a permissive or evidentiary presumption from which a trier of fact may (as opposed to must) draw an inference of guilt will not infringe section 11(d): Osolin, ibid. The same point emerges from jurisprudence of the South African Constitutional Court: State v Mbatha [1996] 2 LRC 208; State v Manamela [2000] 5 LRC 65. In Manamela the majority of the South African Constitutional Court held that a reverse burden provision in respect of handling recently stolen goods was incompatible with a constitutional presumption of innocence. On the other hand, an evidential burden requiring the accused to explain his possession of the goods would not have amounted to a violation of the constitutional right of silence. The majority observed [at para 49]:

In these circumstances I am satisfied that the transfer of the legal burden in section 28 does not satisfy the criterion of proportionality. Viewed in its place in the current legal system section 28 of the 1971 Act is a disproportionate reaction to perceived difficulties facing the prosecution in drugs cases. It would be sufficient to impose an evidential burden on the accused. It follows that section 28 is incompatible with convention rights.

The question is therefore whether, in accordance with section 3 of the 1998 Act, it is possible to read section 28 in a way which is compatible with convention rights: see paragraph 44 of my speech in R v A (No. 2) [2001] 2 WLR 1546 for a detailed explanation of the import of section 3(1). Specifically, the question is whether the words "to prove" in section 28(2) and "if he proves" in subsection (3) may be read as placing only an evidential burden on the accused as Professor Glanville Williams suggested in "The Logic of 'Exceptions'" [1988] C.L.J. 261, at 264-265. If the answer is in the affirmative the burden of proof rests on the prosecution to disprove beyond reasonable doubt the defence. In Kebilene I described this as a respectable argument: 370G. Lord Slynn agreed: 362A. Lord Cooke of Thorndon regarded the distinguished author's view as a possible meaning under section 3. Specifically, Lord Cooke stated that "unless the contrary is proved" can be taken to mean "unless sufficient evidence is given to the contrary" [2000] 2 AC at 373G. I respectfully adopt Lord Cooke's observation. Applying section 3 I would therefore read section 28 (2) and (3) as creating an evidential burden only. In particular this involves reading the words "prove" and "proves" as meaning giving sufficient evidence. I am in agreement with the observations of Lord Hope of Craighead in paragraph 77 of his speech.

My Lords, this is a case of an accused found in possession of two kilograms of cocaine worth over £140,000. It must be comparatively rare for a drug dealer to entrust such a valuable parcel of drugs to an innocent. In any event the appellant's detailed story stretches judicial credibility beyond breaking-point. Even if the judge had directed the jury in accordance with law as I have held it to be the appellant's conviction would have been a foregone conclusion. I would dismiss the appeal.

On 9 April 1999 in the Crown Court at Warrington the appellant was convicted of possessing a controlled drug of class A with intent to supply contrary to section 5(3) of the Misuse of Drugs Act 1971. He was sentenced to seven years' imprisonment. He appealed against his conviction, but on 31 July 2000 the Court of Appeal (Criminal Division) (Lord Woolf CJ, Rougier and Bell JJ) dismissed his appeal: [2001] 2 WLR 211 . The principal issue in the appeal, which was heard together with two other appeals which raised the same issue but from which on 20 December 2000 an appeal committee refused leave to appeal to this House, was the effect of the Human Rights Act 1998 on various statutory provisions which fell into a familiar category. These are provisions which, as Lord Woolf CJ put it at p 215D-E, provide a benefit to a defendant who is being tried for a criminal offence but require him to prove certain facts which the statute specifies before he can obtain that benefit.

The Human Rights Act 1998 received the Royal Assent on 9 November 1998 but the majority of its provisions were not in force at the date of the appellant's trial. Sections 18 , 20 and 21 (5) and 22 came into force on the passing of that Act: section 22 (2). The other provisions of the Act came into force on the days appointed by the Secretary of State by order under section 22 (3). The majority of its provisions were brought into force on 2 October 2000: Human Rights Act 1998 (Commencement No 2) Order 2000 (SI 2000/1851). Lord Woolf CJ said that, in giving its judgment, the Court of Appeal had assumed that the Act was in force at the time when it gave judgment. He also said that, although it had been accepted by all parties that because of section 22 (4) together with section 7 and section 8 of the 1998 Act the court had to approach the safety of the conviction as if the Act had been in force when the judge summed up, the court had reservations as to whether Parliament could have intended such a result: [2001] 2 WLR 211 , 222C D-E.

Now the appellant appeals to your Lordships' House. All the provisions of the 1998 Act are in force, so it is no longer necessary to make any assumptions. But among the issues of general public importance for which a certificate was given by the Court of Appeal under section 33(2) of the Criminal Appeal Act 1968 was the following:

But there are two other issues which are also of general public importance. The first is whether it is an essential element of the offence of possession of a controlled drug under section 5 of the Misuse of Drugs Act 1971 that the accused knows that he has a controlled drug in his possession. The second is whether it is compatible with article 6(2) of the European Convention for the Protection of Human Rights and Fundamental Freedoms for a trial judge to direct a jury, under reference to the defence which is available under section 28 of the 1971 Act, that there is a legal, or persuasive, burden on the accused to prove that he neither believed nor suspected that the substance in question was a controlled drug.

The second issue can only affect the appellant's case if he succeeds on the third issue as to retrospectivity. But the appellant does not need to succeed on the third issue if he is to succeed on his first issue. If he is right on the first issue, which has nothing to do with his article 6(2) Convention right, it must follow that there was a fundamental defect in the summing up by the trial judge.

The Crown case against the appellant depended upon the evidence of police officers who said that on 25 November 1998 they were in the railway station at Runcorn. Prior to the arrival of the London to Liverpool train at 1.30pm they saw two men in the reception hall. One of them was carrying a duffle bag. When the train arrived the appellant got off the train, crossed the bridge and went into the booking hall. The two men who had been waiting outside came in to the booking hall and approached him. One of them said "Steve", and the appellant acknowledged that this was his name and shook hands. All three men then left the booking hall and went to two cars which were in the car park. The appellant and one man got into one car and the other man got into another. About two minutes later the appellant returned to the reception area with a duffle bag. He tried to make a call from his mobile phone. He then went into a phone kiosk, where the police went up to him, identified themselves as police and asked him what he had in the bag. The appellant said, "I don't know, I've just been paid to pick it up". The bag contained two kilogrammes of cocaine at 76% purity which was worth over £140,000.

The appellant's defence was that he did not know what was in the duffle bag. He said in evidence that he had had a phone call the previous evening from a man named John who was in the business of printing tee shirts. He later met John and received from him an envelope which he assumed to contain money. He said that he had in the past driven up to Liverpool to drop off money for him. He got onto the train thinking that he was going to Liverpool. While he was on the train John phoned him and told him to get off at Runcorn. When he got off the train he walked over to a man named Jebb whom he knew and to whom he had also spoken when he was on the train. What normally happened when he met Jebb was that he gave the money to him and took back from him a few packages of tee shirts. On this occasion Jebb was with another man, who handed the appellant the duffle bag. The appellant said that when he asked, "What's this?" Jebb said that it was a bit of scrap for John and then that it was scrap gold. When he was in the car he looked into the bag and Jebb said that it was scrap jewellery but told him to stop rooting about in the bag. When the appellant said that he was just having a look Jebb threatened him, pointing to something under his left arm. The appellant thought that he was going to be shot. As he got out of the car Jebb said that he was to take the bag or he would shoot him and his girlfriend. He went to the telephone to warn his girlfriend and to ask John why he had to take the duffle bag. His case was that at no stage did he suspect that the bag contained controlled drugs and that in any event he was acting under duress.

The trial judge (HHJ Hale) told the jury in the course of his summing up that the real issue in the case was that raised by the defendant who said that, while he accepted all that the prosecution had to prove, he had a defence because he did not know what was in the bag. He gave the following direction:

The jury's verdict shows that they must have held that the appellant had failed to show on a balance of probabilities that he did not know or believe or have reason to suspect that the duffle bag contained controlled drugs and that they were also sure that he was not acting under duress. I think that it is unclear what they would have made of the case if they had been told that proof of knowledge that the bag contained controlled drugs was an essential element of the offence of possession which the prosecution had to prove to the required standard so that they were sure of what was being alleged. On the other hand I take a different view as to what the position would have been if they had been told that that it was for the appellant to provide evidence to support his defence of lack of knowledge that the product or substance was a controlled drug but that this was an evidential burden only which did not require him to prove anything. In that event the direction as to where the burden of proof lay would have been indistinguishable from that which the trial judge gave as to the defence of duress. As the jury rejected that defence the weight to be attached to which depended crucially upon what they made of the appellant's evidence, it is safe to assume that they would have reached the same decision with regard to this defence also. I have no doubt that Mr Owen QC for the appellant was right not to take the opportunity which was offered to him to contend otherwise.

Before I turn to the first issue I must say something about the structure of the Misuse of Drugs Act 1971 with particular reference to the question of possession. It is necessary for me to do this in order to set the scene for an examination of the first and second issues.

Section 5 of the 1971 Act is one of a group of sections containing various restrictions relating to controlled drugs. These include restrictions on their importation and exportation and their production and supply. Section 5 restricts the possession of controlled drugs. The leading provision is that in subsection (1) which provides that, subject to regulations for the time being in force, it shall not be lawful for a person to have a controlled drug in his possession. The offences are created by subsection (2), which provides that it is an offence for a person to have a controlled drug in his possession, and subsection (3), which provides that it is an offence for a person to have a controlled drug, whether lawfully or not, with intent to supply it to another in contravention of section 4(1) of the Act. Schedule 4 to the Act, which deals with the prosecution and punishment of offences, lays down more severe penalties for a section 5(3) offence when it is prosecuted on indictment than it does for a section 5(2) offence.

There are also differences in regard to the penalties for a section 5(2) offence which is prosecuted summarily. These depend upon whether the drug involved is a class A, B or C drug. As my noble and learned friend Lord Steyn has pointed out, the maximum sentence for the section 5(3) offence is one of life imprisonment. But that is so only in the case of a class A drug. In the case of a class B drug the maximum sentence is 14 years imprisonment. In the case of a class C drug it is 5 years imprisonment. In practice the sentences which are imposed for this offence are determinate rather than indeterminate sentences. But it is right to say that the offence sometimes attracts very long periods of imprisonment. The periods of imprisonment vary according to the class of drug involved.

The Act also provides a person who is found to be in possession of a controlled drug with a number of defences. Section 5(4) contains a defence which is available in the case of a charge of simple possession under section 5(2) of the Act. It is a defence which proceeds on the assumption that the person knew or suspected that the thing which was in his possession was a controlled drug. Section 28 deals with the defences of lack of knowledge.

The Act does not contain a definition of "possession", except that section 37(3) provides that for the purposes of the Act the things which a person has in his possession shall be taken to include any thing subject to his control which is in the custody of another. But it is well settled that the expression embraces both a factual and a mental element. The factual element is that of control, as section 37(3) indicates. Unless the thing is in the person's control, albeit while it is in the custody of another, it cannot be said to be in his possession. The mental element is that of knowledge. It is the extent of the knowledge which has to be established that is in dispute in this case. This is the point raised by the first issue. The directions which were given to the jury by the trial judge were to the effect that, while it was necessary for the prosecution to prove that the person knew that the thing was in his control, it was not necessary for the prosecution to prove that he knew that the thing was a controlled drug.

I shall have to examine the provisions of the 1971 Act and the meaning of the word "possession" further when I am dealing with the first and second issues. For the time being it is sufficient to notice that at the time of the trial the directions which were given to the jury by the trial judge were in accordance with the law as it was understood to be at that time. According to this understanding, the mental element is satisfied if it is proved that the person knew that he had the thing in or subject to his control or, if it was in a container such as a bag, that it contained the thing which upon examination was found to be a controlled drug.

As the Lord Justice General (Rodger) said in Salmon v H M Advocate, 1999 JC 67, 78E, after a careful review of the English and Scottish authorities including R v McNamara (1988) 87 Cr App R 246, the prosecution discharge their initial burden by proving that the accused knew that there was something in the bag and that it contained something which turned out to be the controlled drug and that the bag and its contents were under his control. It is not necessary for the prosecution to prove that he knew that the thing was in law a controlled drug for him to be found to be in possession of it. Then there are the statutory defences. If the accused says that he did not know or suspect or have reason to suspect that the bag contained the substance which turned out to be a controlled drug (section 28(2)) or that he did not know or suspect or have reason to suspect that the substance or product was a controlled drug (section 28(3)(b)(i)), the judge's task is to direct the jury to consider whether they are satisfied, on the balance of probabilities, that the defence has been made out: Salmon v H M Advocate , p 79E-F, G-H. That, as I have said, is what the law was understood to be at the date of the trial.

As I have just observed, the directions which the trial judge gave on this matter were in accordance with the law as it was understood to be at the time of the trial. Mr Owen QC for the appellant submitted that this understanding was wrong. Relying to a great extent on the speech of Lord Reid in Warner v Metropolitan Police Commissioner [1969] 2 AC 256 and your Lordships' decision in B v Director of Public Prosecutions [2000] 2 AC 428 , he said that the mental element in the offence of possession of a controlled drug was not satisfied unless the prosecution proved that the defendant knew that the substance or product in his possession was a controlled drug. He said that the offences described in section 5 of the 1971 Act required proof of possession not of a container or an article of whose character the defendant was unaware but of a controlled drug. Knowledge that it was a controlled drug must be taken to be an essential element in the mens rea of the offence.

I would reject this argument. I consider the settled law to be correct on this point. As far as the Misuse of Drugs Act 1971 is concerned, there are two elements to possession. There is the physical element, and there is the mental element. The physical element involves proof that the thing is in the custody of the defendant or subject to his control. The mental element involves proof of knowledge that the thing exists and that it is in his possession. Proof of knowledge that the thing is an article of a particular kind, quality or description is not required. It is not necessary for the prosecution to prove that the defendant knew that the thing was a controlled drug which the law makes it an offence to possess. I observe that Mr Owen did not submit that it was necessary for the prosecution to prove that the defendant was aware that the thing was a class A, B or C drug, as the case may be, although the class into which the drug falls will usually be relevant to any sentence he may receive.

The long title of the 1971 Act states that it is an Act to make new provision with respect to dangerous or otherwise harmful drugs. It was enacted against the background of two important decisions of this House relating to legislation which it repealed: the Drugs (Prevention of Misuse) Act 1964 and the Dangerous Drugs Act 1965. In Warner v Metropolitan Police Commissioner [1969] 2 AC 256, in which the appellant had been convicted of an offence contrary to section 1 of the 1964 Act, it was held by the majority (Lord Reid dissenting) that the prosecution had only to prove that the accused knew of the existence of the thing and that it was in general not a defence for him to say that he believed the thing to be something else such as scent and not drugs. Lord Pearce said at p 305F:

In Sweet v Parsley [1970] AC 132 the appellant was convicted of a contravention of the Dangerous Drugs Act 1965 in that she was concerned in the management of premises which were used for the smoking of cannabis although this was a farmhouse which she visited infrequently and the prosecutor conceded that she was unaware that the premises were used for that purpose. The conviction was quashed on the ground that the offence of which she had been convicted was not an absolute offence. Lord Reid observed at p 150B-D that there were at least two possible ways in which the public scandal of persons being convicted on a serious charge who were in no way blameworthy could be avoided without placing on the prosecutor the full burden of proving mens rea in cases where to do so would lead to many acquittals which were unjust:

The structure of the 1971 Act shows that what Parliament decided to do was to follow what Lord Pearce said was needed to satisfy the meaning of the word "possession" and to adopt the first of the two methods suggested by Lord Reid for avoiding the conviction of those who are not blameworthy. This is indicated most clearly by the words used to describe the defences mentioned in section 5(4) of the Act. This subsection provides:

The defences afforded by this subsection depend upon proof that the accused who is proved to have had a controlled drug in his possession acted as he did "knowing or suspecting it to be a controlled drug". The burden of "proving" that this is what he knew or suspected is placed on the accused. The inclusion of these words in each of paragraphs (a) and (b) of subsection (4) confirms that proof that the controlled drug was in the possession of the accused does not depend upon proof that the accused knew or suspected that the thing was a controlled drug. If it did, the words which I have quoted would have been omitted as they refer to something which, on this hypothesis, the prosecution would have to establish in every case in order to prove that the accused had the controlled drug in his possession. The words of the statute are wholly inconsistent with the appellant's argument.

Further confirmation that this was the approach selected by Parliament can be found in the wording of section 28. Subsection (2) of this section provides:

The function of this provision, which has often been overlooked, was explained by the Lord Justice General (Rodger) in Salmon v H M Advocate , 1999 JC 67. At pp 73H-74A and 74D-F he said:

The contrast which is drawn in these passages between the facts that the Crown must prove and the state of knowledge of the accused is to be found in the language of section 28(2) itself. It demonstrates that what Parliament chose to do was to define the offence in such a way as to require the Crown to prove the facts from which, in appropriate cases, the inference could be drawn that the accused was in possession of the thing which, upon examination and analysis, was shown to be a controlled drug. It left it to the accused to raise the question of lack of knowledge as a defence. In B v Director of Public Prosecutions [2000] 2 AC 428 , 460C-D, Lord Nicholls of Birkenhead said that the starting-point for a court as to the necessary mental element in the case of statutory offences is the established common law presumption that mens rea is an essential ingredient unless Parliament has indicated a contrary intention either expressly or by necessary implication. I would hold that the necessary contrary intention has been indicated in the present case.

I do not think that it is surprising that Parliament made that choice in view of the difficulties which the prosecution would face if it had to prove in every case that the accused knew that the thing was a controlled drug. Taken to its logical conclusion, a requirement to prove mens rea as to the gravamen of the offence would extend to proof of knowledge that it was a controlled drug of the class alleged, as different penalties apply to each class. The legislation has clearly not gone that far, as section 28(3)(a) shows. As it is not a defence for the accused to prove that he neither knew nor suspected nor had reason to suspect that the substance or product in question was the particular controlled drug alleged, it is plain that this is not something that the prosecution need establish. Proof of possession of the substance or product is sufficient. Strict liability follows, subject to the defences which are provided for by the statute

In most cases possession of a container such as a bag or a tin will enable the inference to be drawn that the accused was in possession of its contents, and in most cases where the substance or product is out in the open such as where it is found on the accused's mantlepiece or at his bedside there will be other facts and circumstances from which that inference can be drawn. The problem arises in regard to proof that he knew that the thing in the container, on the mantlepiece or at his bedside was a controlled drug. The fact that the tablet or powder was a controlled drug may be capable of being proved only after careful examination and analysis. Inferences can be drawn if it is found in the company of other material which is used in connection with the supplying or use of controlled drugs. But if it is found on its own and its appearance is all that there is to go by, it may be very difficult for the prosecution to prove that the accused knew that it was a controlled drug.

I think that there are sound reasons of policy for construing the legislation in such a way as not to put the initial burden of proving knowledge of that fact on the Crown. On this issue therefore I would reject the appellant's argument. But this brings me to the question whether the burden which then rests on the accused to raise the question of his knowledge as a defence is a persuasive burden - in which case he must establish his defence on the balance of probabilities - or an evidential burden only which leaves the burden of proof throughout on the Crown.

As I have said, the view hitherto has been that the burden on the accused is a persuasive burden. The wording of section 28(2) and (3), in which the words "to prove" and "if he proves" are used, supports this view. The ordinary meaning of these words is that there is a persuasive burden that must be discharged. But the appellant has raised the question whether this reading of the words used in these subsections is compatible with his article 6(2) Convention right. This is the second issue.

Article 6(2) of the Convention provides that everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law. There is an important question as to whether a statutory provision which transfers the burden of proof to the accused can ever be compatible with that presumption. But for the purposes of this case it is necessary only to answer the particular questions which have been raised. They are (a) whether the provisions of sections 28(2) and 28(3)(b)(i) of the 1971 Act, which according to the ordinary meaning of the words used require the accused to prove the defences mentioned there on the balance of probabilities, are incompatible with the Convention right; and (b), if so, whether they can be read and given effect to under section 3 of the 1998 Act in a way which is compatible with it.

Mr Owen made it clear that the arguments which he presented on these questions were not directed to the defences which are mentioned in section 5(4) of the 1971 Act. This was because section 5(4) relates to things which the accused must establish if he wishes to avoid conviction but are not an essential element of the offence: see R v Edwards [1975] QB 27, 39-40. In that case the Court of Appeal said that, where an enactment prohibits the doing of an act save in specified circumstances or by persons of specified classes or with specified qualifications, it is for the defendant to prove that he was entitled to do the prohibited act. In R v Hunt [1987] AC 352, 375 Lord Griffiths said that he had little doubt that the occasions upon which a statute would be construed as imposing a burden of proof on a defendant which did not fall within that formulation would be exceedingly rare. It was to cases falling outside that formulation that Mr Owen directed his argument.

The section 5(4) defence has not been raised in this case, but I would not wish to be taken as accepting that exceptions of that kind are always immune from challenge on Convention grounds. As I see it, there are three distinct questions, and all three questions need to be asked and answered.

The first question is whether, upon the construction of the enactment, the defence is an exception of the kind described in R v Edwards. The second is whether the language used by Parliament, according to its ordinary meaning, has modified the golden thread rule as described by Viscount Sankey LC in Woolmington v Director of Public Prosecutions [1935] AC 462 , 481. This rule requires that, subject to the defence of insanity and to any statutory exception which transfers the burden of proof in the case of a particular offence laid down in an enactment, the prosecution must always prove its entire case beyond reasonable doubt. This question too is a question of construction. In a case of a provision such as that found in section 5(4), where the words used are "it shall be a defence for him to prove", the answer to it is plain on the face of the enactment. A provision which takes this form is understood to be an express statutory exception to the golden thread rule.

But there is a third question, which was the subject of some debate in the light of the Hunt case but has now been brought right out into the open by sections 3 (1) and 6 (1) of the Human Rights Act 1998 . It used to be whether placing the burden on the accused by the particular statute can be justified by broader considerations of policy: see Peter Mirfield, The Legacy of Hunt [1988] Crim L R 19; D J Birch, Hunting the Snark; the Elusive Statutory Exception [1988] Crim L R 221; Peter Mirfield, An Ungrateful Reply [1988] Crim L R 233. It can now be expressed in the language which is appropriate to an examination of the Convention rights.

Section 3 (1) of the 1998 Act provides that, so far as it is possible to do so, primary and secondary legislation must be read and given effect in a way which is compatible with the Convention rights. I should now like to explain how, as I see it, this important and far-reaching new approach to the construction of statutes should be employed consistently with the need (a) to respect the will of the legislature so far as this remains appropriate and (b) to preserve the integrity of our statute law so far this is possible.

The first point, as I said in paragraph 108 of my speech in R v A [2001] UKHL 25 , is that the effect of section 3 (1) is that the interpretation which it requires is to be achieved only so far as this is possible. The word "must", which section 3 (1) uses, is qualified by the phrase "so far as it is possible to do so". The obligation, powerful though it is, is not to be performed without regard to its limitations. Resort to it will not be possible if the legislation contains provisions, either in the words or phrases which are under scrutiny or elsewhere, which expressly contradict the meaning which the enactment would have to be given to make it compatible. The same consequence will follow if legislation contains provisions which have this effect by necessary implication. Further justification for giving this qualified meaning to section 3 (1) is to be found in the words "read and give effect." As the side note indicates, the obligation is one which applies to the interpretation of legislation. This function belongs, as it has always done, to the judges. But it is not for them to legislate. Section 3 (1) preserves the sovereignty of Parliament. It does not give power to the judges to overrule decisions which the language of the statute shows have been taken on the very point at issue by the legislator.

The second point, as I said in paragraph 110 of my speech in R v A , is that great care must be taken, in cases where a different meaning has to be given to the legislation from the ordinary meaning of the words used by the legislator, to identify precisely the word or phrase which, if given its ordinary meaning, would otherwise be incompatible. Just as much care must then be taken to say how the word or phrase is to be construed if it is to be made compatible. The justification for this approach to the use of section 3 (1) is to be found in the nature of legislation itself. Its primary characteristic, for present purposes, is its ability to achieve certainty by the use of clear and precise language. It provides a set of rules by which, according to the ordinary meaning of the words used, the conduct of affairs may be regulated. So far as possible judges should seek to achieve the same attention to detail in their use of language to express the effect of applying section 3 (1) as the parliamentary draftsman would have done if he had been amending the statute. It ought to be possible for any words that need to be substituted to be fitted in to the statute as if they had been inserted there by amendment. If this cannot be done without doing such violence to the statute as to make it unintelligible or unworkable, the use of this technique will not be possible. It will then be necessary to leave it to Parliament to amend the statute and to resort instead to the making of a declaration of incompatibility.

As to the techniques that may be used, it is clear that the courts are not bound by previous authority as to what the statute means. It has been suggested that a strained or non-literal construction may be adopted, that words may be read in by way of addition to those used by the legislator and that the words may be "read down" to give them a narrower construction that their ordinary meaning would bear: Clayton and Tomlinson, The Law of Human Rights, para 4.28, p 168 (Oxford, 2000). It may be enough simply to say what the effect of the provision is without altering the ordinary meaning of the words used: see Brown v Stott 2000 JC 328, 355B-C, per Lord Justice General Rodger. In other cases, as in Vasquez v The Queen [1994] 1 WLR 1304, the words used will require to be expressed in different language in order to explain how they are to be read in a way that its compatible. The exercise in these cases is one of translation into compatible language from language that is incompatible. In other cases, as in R v A , it may be necessary for words to be read in to explain the meaning that must be given to the provision if it is to be compatible. But the interpretation of a statute by reading words in to give effect to the presumed intention must always be distinguished carefully from amendment. Amendment is a legislative act. It is an exercise which must be reserved to Parliament.

The haphazard way in which reverse burden of proof provisions have been introduced into legislation by Parliament has been identified and persuasively criticised: Andrew Ashworth and Meredith Blake, The Presumption of Innocence in English Criminal Law [1996] Crim L R 306. As they say, at p 314, nothing could be clearer than the 11th Report of the Criminal Law Revision Committee, Evidence (General) , Cmnd 4991 (1972), para 140 where the Committee stated: "we are strongly of the opinion that, both on principle and for the sake of clarity and convenience in practice, burdens on the defence should be evidential only." It is generally accepted that Woolmington changed the law as to the burden of proof in the case of common law defences such as self-defence and non-insane automatism: Professor J C Smith, The Presumption of Innocence (1987) 38 NILQ 223, 226; Vasquez v The Queen [1994] 1 WLR 1304, 1309G-H. The same approach has been taken in Scotland to where the onus lies in the case of all common law pleas and defences other than the plea of diminished responsibility and the defence of insanity: Lambie v H M Advocate , 1973 JC 53 (incrimination); Ross v H M Advocate , 1991 JC 210 (non-insane automatism).

The lack of clarity and the inconvenience of applying a different rule to defences created by statute is obvious in the present case. Section 28(4) of the 1971 Act provides that nothing in that section shall prejudice any defence which it is open to a person when charged with an offence to which that section applies to raise apart from that section. In this case the appellant did raise such a defence. It was his defence of duress. That defence was intimately bound up with his defence under the statute, as it depended entirely upon what the jury made of his evidence. But the trial judge had to direct the jury that the onus as regards the defence of duress rested on the prosecution. The jury were not told why there was a difference as to where the onus lay. There was no need for this information to be given to them. But it would not be surprising if they found it hard to maintain a clear distinction between the two positions as to onus when they examined the evidence.

There is no doubt that it is possible, in the light of section 3 (1) of the Human Rights Act 1998 , to read sections 28(2) and 28(3) of the 1971 Act in such a way as to impose no more than an evidential burden on the accused. As it is a rule of construction, the exercise which section 3 (1) prescribes makes it necessary to identify the words used by the legislature which would otherwise be incompatible with the Convention right and then to say how these words are to be construed according to the rule to make them compatible. But in this case there is no difficulty. As Lord Cooke of Thorndon said in R v Director of Public Prosecutions, ex parte Kebilene [2000 ] 2 AC 326, 373G:

In Vasquez v The Queen [1994] 1 WLR 1304 the Privy Council were asked to consider the question whether section 116(a) of the Belize Criminal Code which placed the burden on the defendant to prove extreme provocation as a defence to murder contravened the defendant's right under 6(3)(a) of the Constitution of Belize to be presumed innocent until he was proved guilty. Applying the principles described in Attorney-General of The Gambia v Momodou Jobe [1984] AC 689, 700 by Lord Diplock and in Attorney-General of Hong Kong v Lee Kwong-kut [1993] AC 951, 962 by Lord Woolf, the Board held that section 116(a) was in conflict with the Constitution and that it had to be modified to conform to it. The words "if either of the following matters of extenuation be proved on his behalf" were to be construed as though they read "if there is such evidence as raises a reasonable doubt as to whether": p 1314D-E, per Lord Jauncey of Tullichettle. It was by this means that Belize was brought into line with the other Commonwealth countries of the Caribbean, where the onus of proof of unprovoked killing was placed on the prosecution. It provides a good example of the use of an interpretative obligation of the kind that has now been written into our domestic law by section 3(1) of the 1998 Act.

More recently, in Michael Yearwood v The Queen [2001] UKPC (June 2001) the Board held that section 239 of the Grenada Criminal Code, which is in the same terms as section 116 of the Belize Criminal Code, had to be read and given effect to in a way that was compatible with the provisions for the protection of the fundamental rights and freedoms to which every person is entitled under the Constitution of Grenada, and in particular with section 8(2)(a) of the Constitution which entitles a person who is accused of a criminal charge to the presumption of innocence. It held that the words "are proved on his behalf" in section 239 must be read and given effect to as if for those words there were substituted the words "are the subject of such evidence as to raise a reasonable doubt."

Of course, the fact that it is possible for a statutory provision to be read in this way does not mean that it must be so read. The first question is whether, leaving aside section 3(1), there would be a breach of the Convention. For the reasons which I sought to explain in R v Director of Public Prosecutions, ex parte Kebilene [2000] 2 AC 326 , 383-388, I do not think that a reverse onus provision will inevitably give rise to a finding of incompatibility. In Salabiaku v France (1988) 13 EHRR 379 at p 388, para 28 the European Court of Human Rights said:

Mr Owen said that the court was not concerned in the Salabiaku case with a provision applicable to a person charged with a serious criminal offence which placed the burden of proof on him with respect to an essential element of it. That is true, but I do not think that this deprives it of value as a statement of principle. What it means is that, as the article 6(2) right is not absolute and unqualified, the test to be applied is whether the modification or limitation of that right pursues a legitimate aim and whether it satisfies the principle of proportionality: Ashingdane v United Kingdom (1985) 7 EHRR 528 ; see also Brown v Stott [2001] 2 WLR 817 . It is now well settled that the principle which is to be applied requires a balance to be struck between the general interest of the community and the protection of the fundamental rights of the individual. This will not be achieved if the reverse onus provision goes beyond what is necessary to accomplish the objective of the statute.

The statutory objective is to penalise the unauthorised possession of dangerous or otherwise harmful drugs. But the statute recognises, among other things, that it would be wrong to penalise those who neither knew nor suspected nor had reason to suspect the existence of some fact alleged by the prosecution which it is necessary for the prosecution to prove if he is to be convicted of the offence charged (section 28(2)) or that the substance or product in question is a controlled drug (section 28(3)(b)(i)). That being so, it is hard to see why a person who is accused of the offence of possessing a controlled drug and who wishes to raise this defence should be deprived of the full benefit of the presumption of innocence. The systems of control and prosecution might well be in jeopardy if there were to be an initial onus on the prosecution to establish that the accused knew these things. The right to silence and the covert and unscrupulous nature of drug-related activities must be taken into account in the assessment as to whether a fair balance had been achieved. But we are not concerned here with the initial onus. As I have said in my answer to the first issue, the prosecution do not need to prove that the accused knew that the thing in his possession was a controlled drug. This is a matter which must be raised by the defence.

The choice then is between a persuasive burden, which is what the ordinary meaning of the statutory language lays down, and an evidential burden, which is the meaning which it is possible to give to the statutory language under section 3(1) of the 1998 Act. If the evidential burden were to be so slight as to make no difference - if it were to be enough, for example, for the accused merely to mention the defence without adducing any evidence - important practical considerations would suggest that in the general interest of the community the burden would have to be a persuasive one. But an evidential burden is not to be thought of as a burden which is illusory. What the accused must do is put evidence before the court which, if believed, could be taken by a reasonable jury to support his defence. That is what Professor Glanville Williams envisaged when he was giving this meaning to the words "unless the contrary is proved": "The Logic of 'Exceptions'" [1988] CLJ 261, 265. It is what the Judicial Committee envisaged in Vasquez v The Queen [1994] 1 WLR 1304, 1314G-H and in Michael Yearwood v The Queen [2001] UKPC . It is what the common law requires of a defendant who wishes to invoke one of the common law defences such as provocation or duress.

The practical effect of reading section 28(2) and section 28(3) as imposing an evidential burden only on the accused and not a persuasive burden as they have been understood to impose hitherto is likely in almost every case that can be imagined to be minimal. In Salmon v H M Advocate , 1999 JC 67, 75C-D, the Lord Justice General said this as to the effect on the accused of the persuasive burden:

It is worth noting in this connection that Parliament itself has recently recognised the force of the argument that as a general rule statutory provisions which require the accused to prove something as a defence to the offence with which he has been charged should be read and given effect to as if they imposed only an evidential burden on him and not a probative one. The Terrorism Act 2000 contains several provisions which say that it shall be a defence for a person charged with an offence to prove something. For example, section 57(2) provides that it shall be a defence for him to prove that his possession of an article was not for a purpose connected with the commission, preparation or instigation of an act of terrorism. But section 118(2), which applies to a number of provisions in the Act including section 57(2) which say that it is a defence for a person to prove something, provides:

Section 53 (3) of the Regulation of Investigatory Powers Act 2000 is to the same effect. It provides a defence to the offence of possession described in section 53 (2). It places the onus of proving the contrary beyond a reasonable doubt on the prosecutor if sufficient evidence of that fact is adduced to raise an issue with respect to it. It is not unreasonable to think that, if Parliament were now to have an opportunity of reconsidering the words used in section 28 (2) and (3) of the 1971 Act, it would be content to qualify them in precisely the same way.

I would therefore read the words "to prove" in section 28 (2) as if the words used in the subsection were "to give sufficient evidence", and I would give the same meaning to the words "if he proves" in section 28 (3). The effect which is to be given to this meaning is that the burden of proof remains on the prosecution throughout. If sufficient evidence is adduced to raise the issue, it will be for the prosecution to show beyond reasonable doubt that the defence is not made out by the evidence. The question whether these provisions must be read and given effect to in that way in this case depends on the issue of retrospectivity, to which I now turn.

The appellant's argument on this issue was presented under two main heads, which were stated as alternatives. The first was based upon the provisions of section 6 (1) of the Human Rights Act 1998 , read together with those of section 7 (1)(b) and section 22 (4). The second was based upon the provisions of section 6 (1) of the Act without reference to section 22 (4). Section 6 (1) provides the starting point for each of these two alternative heads of argument. It provides:

Each of these two alternatives raises questions about the structure of the 1998 Act. So I must now try to describe that structure, with particular reference to the issue about the extent to which its provisions were intended to be retrospective.

It is first necessary to have regard to the Convention. Article 13 of the Convention provides that everyone whose rights and freedoms as set forth in the Convention are violated shall have an effective remedy before a national authority. This article is not one of the Convention rights to which effect is given by the 1998 Act, but this was no accident. As I observed in Brown v Stott [2001] 2 WLR 817 , 847B, the reason which was given for its omission from the articles set out in Schedule 1 to the Act was that sections 7 to 9 of the Act were intended to lay down an appropriate remedial structure for giving effect to the Convention rights as defined by section 1 (1) of the Act . The state's obligation to provide an effective remedy before a national tribunal in the event of a violation of the Convention rights is part of the background against which the provisions of subsections (2) to (4) of section 22 were enacted.

The relevant subsections of section 22 are in these terms:

These provisions provide important guidance as to the application of the Act to events occurring or legislation passed prior to its coming into force. As a general rule the approach which was taken by Parliament was to apply the Act prospectively as from the date when the relevant provisions were brought into force and not retrospectively. The purpose of the government, as the White Paper which introduced the Bill explained, was to see "rights brought home": Rights Brought Home: The Human Rights Bill, Cm. 3782 (1997), para 1.19. The provisions of sections 7 to 9 of the Act which deal with proceedings, judicial acts and judicial remedies were for the most part designed to apply to events occurring or legislation passed after the Act came into force. In general it was as from the date of the coming into force of the 1998 Act that the rights were intended to be brought home.

But this not the whole story, as can be seen from an examination of section 3 and section 22 (4) of the Act . Section 3 (2)(a) provides that the interpretative obligation which section 3 (1) lays down applies to primary and secondary legislation "whenever enacted". Section 22 (4) provides that section 7 (1)(b) applies to proceedings brought by or at the instigation of a public authority "whenever the act in question took place". It is these provisions which open the door to the appellant's first alternative argument on retrospectivity.

This argument depends upon section 22 (4), and it was presented in two ways. That subsection applies only to proceedings brought by or at the instigation of a public authority. The appellant's first approach was to apply the subsection to the prosecuting authority. The second was to apply it to the court as a public authority.

Had the appellant's complaint been that the prosecutor had acted in a way which was inconsistent with his article 6(2) Convention right the statutory route would have been easy to follow. The certified question refers to "an investigating or prosecuting authority", and there is a reference to the "prosecuting authority" in the relevant paragraph of the statement of facts and issues. The Crown Prosecution Service, which was the prosecuting authority in this case, is a public authority within the meaning of sections 7 and 22 (4) of the 1998 Act. Prosecutions brought by the CPS are proceedings brought by a public authority within the meaning of section 22 (4). So section 7 (1)(b) applies to proceedings brought by the CPS whenever the act in question - that is, the act on its part which is made unlawful by section 6 (1) - took place.

As for section 7 (1)(b), it enables a person who claims that a public authority has acted in a way which is made unlawful by section 6 (1) of the Act to rely on the Convention right or rights concerned in any legal proceedings. The expression "legal proceedings" in this paragraph is defined in subsection (6) of section 7 , which provides:

The effect of section 22 (4) is to limit the extent to which section 7 (1)(b) can be applied retrospectively. It can be used retrospectively in proceedings brought by or at the instigation of public authorities - that is to say, to enable a person to rely on the Convention right or rights defensively. This may be done at the stage of any appeal against the decision of any court. So the retrospective use of section 7 (1)(b) is permitted by section 22 (4) at each stage of an appeal, including an appeal to your Lordships' House. The appeal is treated by the Act as if it were part of the same legal proceedings as those brought by or at the instance of the public authority, irrespective of the person at whose instance the appeal is brought. But it is plain that section 7 (1)(b) may not be used with retrospective effect in proceedings brought against a public authority. That is the effect of the concluding words of section 22 (4).

This was the route for the giving of retrospective effect to a Convention right which I had in mind when I said in Reg v Director of Public Prosecutions, ex parte Kebilene [2000] 2 AC 326 , 375D:

Mr Starmer sought to rely on this passage in support of his argument. But in that case the court was being asked to review an exercise of the discretion of the Director of Public Prosecutions in giving his consent to the prosecution of the applicants under section 16A of the 1989 Act as amended. There was no doubt that the court was dealing there with proceedings brought by or at the instigation of a public authority. Nor was it in question that it was his act in consenting to the proceedings that had led to the alleged incompatibility with the applicants' Convention right. The situation was the same in Wallbank v The Parochial Church Council of Aston Cantlow and Wilmcote (Case No.: A3/2000/0644), 17 May 2001(CA). In that case it was conceded that the questions whether the Parochial Church Council was a public authority within the meaning of section 6 of the 1998 Act and, if so, whether its action in serving notice upon the defendants was unlawful by reason of article 1 of the First Protocol, read either alone or with article 14 of the Convention, were open to the defendants to raise although the impugned notice was served on 12 September 1994 - long before the 1998 Act was brought into force. That too was a case where a potential victim was seeking to rely on an alleged breach of a Convention right in legal proceedings brought by a public authority within the meaning of section 22 (4). It was a case where a public authority, which was what the Parochial Church Council was held to be, was seeking to rely on acts of its own which were lawful under the domestic law at the time when they were done but were incompatible with the Convention right.

In so far as the question raised by this issue relates to a breach of Convention rights by the prosecuting authority, I would answer it in the affirmative. I consider that an accused whose trial took place before the coming into force of the 1998 Act is entitled to rely in an appeal after the coming into force of that Act on an alleged breach of his Convention rights by the prosecuting authority. But that is not the situation in the present case. The proceedings were brought by a public authority, but it is not the bringing of those proceedings nor any other act on its part that is said to have led to the alleged incompatibility. The appellant's case is that it was the directions that the trial judge gave in his summing up which were incompatible with his article 6(2) Convention right. That was an act of the court. It was not an act of the prosecuting authority.

The appellant's alternative argument based upon section 22 (4) and section 7 (1)(b) seeks to apply those provisions to enable effect to be given to the obligations in sections 3 (1) and 6 (1) retrospectively to the acts of the trial judge when he was delivering his summing up in the trial court, or alternatively to the acts of the appeal court when it was considering the safety of the appellant's conviction in the light of that summing up. Section 6 (3)(a) provides that in that section "public authority" includes a court or tribunal. There is no doubt therefore that, if section 6 (1) of the Act had been in force at the time of the trial, the obligation which is set out in that subsection would have been binding on the trial judge. He would also have been obliged by section 3 (1) to read and give effect to section 28 of the 1971 Act in a way which was compatible with Convention rights.

I am unable to construe section 22 (4) read with section 7 (1)(b) in this way. I do not have any difficulty in reading the expression "public authority" in these provisions as including a court or tribunal. The meaning which is given to that expression for the purposes of section 6 by section 6 (3) must be read in to sections 7 (1) and 7 (6) and section 22 (4), as all these provisions are interlinked. But a court or tribunal is not a public authority by or at the instigation of which proceedings are brought. Section 7 (1) contemplates proceedings to which a public authority is a party. A court or tribunal is not a party to the proceedings which are brought before it in its judicial capacity. In my opinion the words of the statute are incapable of supporting this part of the appellant's argument.

Mr Starmer said that the interpretative obligation in section 3 (1) applies to the provisions of the 1998 Act in exactly the same way as it applies to any other statute. I agree. Section 3 (2)(a) provides that the section applies to primary legislation whenever enacted. So it applies to the 1998 Act, even although that Act was enacted before section 3 itself was brought into force on 2 October 2000. Where section 3 (1) applies it may have the effect of altering the effect of prior legislation retrospectively. But the obligation which section 3 (1) lays down is to read and give effect to legislation compatibly with Convention rights "so far as it is possible to do so". This will not be possible if the legislation expressly contradicts the meaning which it would have to be given if it is to be made compatible.

Section 22 (4) provides expressly that subsection 7 (1)(b) does not apply to an act taking place before the coming into force of that section "otherwise" than in the case of proceedings "brought by or at the instigation of" a public authority. I would hold that this provision expressly contradicts Mr Starmer's argument. In any event I am unable to identify any respect in which it is incompatible with any of the Convention rights mentioned in section 1 (1) of the 1998 Act. A deliberate choice was made by Parliament as to the extent to which section 7 (1)(b) could be given effect to retrospectively in order to provide a person whose Convention rights have been violated with an effective remedy. But article 13 of the Convention, to which section 7 gives effect, is not one of the Convention rights mentioned in section 1 (1). I do not think that it is open to the court to make a different choice from that which was made by Parliament. To do so would not be to construe the enactment in the way which section 3 (1) contemplates. It would be to do something which it does not permit, which is to legislate.

The appellant's second alternative argument on retrospectivity is based entirely upon the provisions of section 6 (1). His proposition is simply this. Section 6 (1) provides that it is unlawful for a public authority to act in a way that is incompatible with a Convention right. A court, including an appellate court, is a public authority: see section 6 (3)(a). The House of Lords is a public authority when it is acting in its judicial capacity: see section 6 (4). It is unlawful for the House of Lords in that capacity to act in a way that is incompatible with the appellant's article 6(2) Convention right. Section 6 is now in force. This means that the House must give effect to that Convention right in the hearing of any appeal. Giving effect to it involves examining the question whether the summing up by the trial judge was incompatible with the Convention right, irrespective of the question whether section 3 (1) of the 1998 Act was in force at the time of the summing up. If it is found that the summing up was incompatible the House must take account of that fact in its consideration of the question whether the appellant's conviction was unsafe. In my opinion this argument is inconsistent with the scheme of the 1998 Act.

What section 6 (1) declares to be unlawful is an "act" of a public authority. The 1998 Act tells us that an act for this purpose includes a failure to act: section 6 (6). In contrast to section 3 (1) which is expressed in terms of an obligation, section 6 (1) is expressed in terms of a prohibition. It imposes a duty on a public authority not to act in a way which is incompatible with a Convention right. The word "act" suggests that the prohibition is concerned with conduct. Conduct by a public authority is usually preceded by some process of reasoning or decision-taking, but it is the act rather than the decision which precedes it which is the subject of the prohibition. When the word is applied to things done by a court, it is not difficult to apply the prohibition to the conduct of the proceedings. They must not be conducted in a way which is incompatible with the article 6 Convention right to a fair and public hearing within a reasonable time by an independent and impartial tribunal. But what of the court's decision-taking function and the process of reasoning that leads up to it? Peter Mirfield, Regulation of Investigatory Powers Act 2000 (2): Evidential Aspects [2001] Crim L R 91, 92-93 has suggested that it is open to question whether these aspects of what a court does are included within the word "act".

I would be inclined to give the word "act" a broad and purposive meaning in this context, for the reasons indicated by Lord Wilberforce in Minister of Home Affairs v Fisher [1980] AC 319 , 328E-329A. It seems to me that, it is not only the manner of the decision-taking exercise that is brought under scrutiny. I would hold that the prohibition applies also to the interpretative obligation in section 3(1). The court is prohibited from carrying out the "act" of statutory construction otherwise than in accordance with that obligation. A decision which is based on the application and development of the common law also is an act by the court, so I think that it must follow that this too is subject to the prohibition in section 6(1). Thus it is unlawful for a court to conduct a hearing in a way which is incompatible with a person's article 6 Convention rights. But the prohibition in section 6(1) also affects matters of substance. So it will be unlawful within the meaning of section 6(1) for a court to determine a criminal charge on an interpretation of a statute which ignores the interpretative obligation in section 3(1), or on a proposition of law which is incompatible with a Convention right. It will be unlawful in the Convention sense for an appellate court to do likewise. The remedy for decisions of that kind lies in the domestic law which provides for, or sets limits to, an appeal against or the review of decisions of the lower courts: see section 9(1). The defendant's Convention rights are protected so long as the basis upon which the appeal process itself is conducted is not incompatible with the Convention rights. Thus far I am at one with the appellant's argument.

The appellant maintains that it would be incompatible with his article 6(2) Convention right for an appellate court not to read and give effect to section 28 of the 1971 Act in a way that was compatible with it. As soon as section 3(1) was brought into force the interpretative obligation was binding on all courts irrespective of the date when the legislation was enacted. I agree that it would have been binding on the trial court had the section been in force at the date of the trial. But there is nothing in the 1998 Act to indicate that that subsection is to be applied retrospectively to acts of courts or tribunals which took place before the coming into force of section 3(1). The provisions of section 22(4) are to the contrary. There would have been no point in enacting that section 7(1)(b) was to have retrospective effect in the way which that subsection provides but not otherwise if appellate courts were to be obliged by section 6(1) to give retrospective effect to that subsection in all cases where they were required to adjudicate upon acts by courts or tribunals as public authorities.

In my opinion the position was correctly stated by Sir Andrew Morritt V-C in Wilson v The First County Trust Limited (No 2) (Case No B2/1999/1073), 2 May 2001, when he said in paragraph 20:

I would hold that the trial judge was right to tell the jury that it was not necessary for the prosecution to prove that the appellant knew that the thing that was in the bag was a controlled drug. I would also hold that a direction that it is for the accused to prove his defence of lack of knowledge under section 28 of the 1971 Act on the balance of probability is not compatible with the article 6(2) Convention right. But this finding does not provide the appellant with a remedy under section 6(1) of the 1998 Act, as the relevant provisions of that Act were not in force at the time of his trial. Had it been necessary to do so, I would have held that his conviction was not unsafe on the ground that the jury would have reached the same result if a direction had been given to them which was compatible with the Convention right. I would dismiss the appeal.

The appellant was convicted of an offence under section 5(3) of the Misuse of Drugs Act 1971. That section provides:

The appeal raises firstly a question about the essential elements of the offence of possession of a controlled drug, and secondly certain questions relating to the European Convention on Human Rights and Fundamental Freedoms and the Human Rights Act 1998 . I turn first to the construction of sections 5 and 28 of the Misuse of Drugs Act 1971.

The focus of the argument in relation to the statutory offence contained in section 5(3) has been on the element of possession. In that respect the argument may also extend to the offence set out in section 5(2), simple possession of a controlled drug. But what is exactly comprehended in the concept of possession may vary in accordance with the context in which the word is used. It may be accompanied by other expressions which may colour the substance of what is intended, as for example the expression "possession, custody or power" ( B v B [1979] 1 All ER 801). There may be some qualification, adjectival or adverbial, attached to the language, such as for example "actual" possession, or "knowingly" possess, which may attract a particular meaning. But even if the word stands alone it does not necessarily follow that the meaning is in every context identical or that precisely the same elements have to be proved in order to establish possession of the thing with which the offence is concerned. The context may well colour the meaning. In the present case it is the analysis of the concept of possession under section 5 of the 1971 Act which is essentially in issue. While it may well be safe to state in a general way that possession in the context of the criminal law comprises the elements of control and knowledge, the critical question of the substance of the knowledge which is one of the ingredients of the offence may not be open to the same answer in the context of different statutory provisions.

This observation relates particularly to the existence of any mental element in the offence. Some care may have to be taken in the present context with the use of the expression "mens rea". While the label is convenient as referring to the mental element or elements which may be comprised in any particular offence it may be difficult to find any single common definition of it applicable to all cases. It may indeed, as Sir J.F.Stephen ( History of the Criminal Law , 1883 Vol 2, 95, quoted in Gordon, Criminal Law 3rd ed. 7.08) observed, mean:

The first of the certified questions asks whether it is an essential element of the offence of possession of a controlled drug that the accused knows that he has a controlled drug in his possession. This question requires consideration of the ingredients of possession for the purposes of section 5(2) and (3). In my view the word has the same meaning which it has always had in the context of the legislation relating to controlled drugs, that is to say that the accused has control of the thing and knows that he has it in his control. That these two elements are necessary can be found in the decision in Reg v Warner [1969] 2 AC 256 which concerned the Drugs (Prevention of Misuse) Act 1964 and there is nothing to show that Parliament intended any other meaning to be given to the word in the later legislation ( Reg v Ashton - Rickhardt [1978] 1 WLR 37, 42). The meaning of possession for the purposes of the Misuse of Drugs Act 1971 is now well established. The first of the two elements, control, involves a physical control. The concept is enlarged by section 37(3) of the Act which states that "things which a person has in his possession shall be taken to include any thing subject to his control which is in the custody of another". The second element involves that the defendant knows that the thing in question is under his control. He need not know what its nature is, but so long as he knows that the thing, whatever it is, is under his control, it is in his possession ( Reg v Ashton-Rickardt [1978] 1 WLR 37). The prosecution requires to prove that the defendant had control of the thing, that he knew he had control of it, and that the thing was the controlled drug which the prosecution allege it to have been. In addition, of course, for section 5(3) the prosecution must prove that the defendant had the intent to supply the thing.

The particular question is whether it is necessary for possession of a controlled drug that the accused knows that the thing is of the nature which it is alleged to be. The answer to that question is in my view evident from other provisions of the Act, in particular sections 5(4) and 28(3)( b )(i). Section 5(4) applies to proceedings for an offence under section 5(2) where "it is proved that the accused had a controlled drug in his possession". The offence under section 5(2) is one of having a controlled drug in his possession unlawfully, that is to say, contrary to any regulations under section 7 authorising the possession of controlled drugs (see section 5(1)). The defence which is made available by section 5(4) expressly includes the ingredient that the accused knew or suspected that what he possessed was a controlled drug. It deals with cases where it is proved that the accused had a controlled drug in his possession but that "knowing or suspecting it to be a controlled drug" he possessed it for certain specified purposes. That that fact may come in after it has been proved that he had a controlled drug in his possession demonstrates that such knowledge or suspicion is not an ingredient of possession. The words which I have quoted would be unnecessary if the knowledge or suspicion was already an ingredient of the possession which has been proved. Section 28(3)( b )(i) deals with cases where the accused neither believed nor suspected nor had reason to suspect that the substance or product in issue was a controlled drug. In such a case he is to be acquitted. But that provision would be unnecessary if proof of possession of a controlled drug necessarily entailed proof of knowledge that the thing in question was a controlled drug.

The matter is one of the construction of the particular legislation. A contrast may be made with DPP v Brooks [1974] AC 862 . That case concerned a statutory provision in Jamaica which provided that "every person who…has in his possession any…ganja;…shall be guilty of an offence". The Privy Council held that the prosecution had to prove that the defendant knew that the thing which he had in his van was ganja. That view was reached in light of long-standing authority on the meaning of the provision and the particular wording of the Jamaican statute. There the statute was silent on the matter of the precise knowledge of the accused.

I note that the word "knowingly" is not found in section 5(3). It is a word expressly used in section 8 and I believe that its absence in section 5(3) does have some significance. In Sweet v Parsley [1970] AC 132 , 149, Lord Reid observed that it is:

In some cases the drug may be found on the accused's person or in his house in a state or situation from which it is immediately obvious that it is a controlled drug. In such cases it may well be that that no question about the accused's knowledge of the nature of the substance may arise. A more difficult problem arises where the drug in question is within some container, such as a box or a bag. But the law here is clear to the effect that if the defendant is in possession of the container and knows that there is something in it, he will be taken to be in possession of the contents of the container. That was clearly affirmed in McNamara ([1988] 87 Cr App R 246)). Where the drug is in a container, it is sufficient for the prosecution to prove that the defendant had control of the container, that he knew of its existence and that there was something in it, and that the something was in fact the controlled drug which the prosecution alleges it to be. The prosecution does not require to prove that the accused knew that the thing was a controlled drug.

The appellant placed some reliance upon in B (A Minor) v Director of Public Prosecutions [2000] AC 428. In that case this House affirmed that mens rea was an essential element of every criminal offence unless Parliament expressly or by necessary implication provided to the contrary. The case was one of an alleged breach of section 1(1) of the Indecency with Children Act 1960. In B what was required to be proved by the prosecution was the absence of a genuine belief on the part of the defendant that the victim was 14 years old or more. The section was silent on the matter of knowledge or belief, and the House held that the offence must be taken to include the mental element of genuine belief. The importance of the decision to my mind is that it has affirmed a rule of general application to all statutory offences, not simply to statutory sexual offences. One must then approach statutory offences with the understanding that there is necessarily a mental element in the offences there contained, unless there is a clear indication to the contrary.

But I do not consider that B is of particular assistance in the present case. The mental element of the offence under section 5(3) is partly built into the concept of possession and the element of the intent to supply, and partly set out in section 28 of the Misuse of Drugs Act 1971. The mental element respecting knowledge or suspicion of the nature of the thing has been expressly catered for in section 28. It only arises for consideration on the assumption that the facts which the prosecution requires to prove in order to obtain a conviction have been proved. If the accused did not know or suspect or have reason to suspect any of those facts, then section 28(2) provides him or her with a defence. I entirely agree with the recognition by the High Court of Justiciary in Salmon v HM Advocate 1999 JC 67 of the priority to be given to subsection (2) of section 28 over subsection (3). Subsection (3) expressly only comes into play if it is proved that the substance or product in question was the controlled drug which it was necessary for the prosecution to have proved it to have been. Section 28 provides an escape for the defence by adding a qualification to the strict operation of the definition of possession. It affords a defence to an accused person where no defence had previously existed under the earlier legislation ( Reg v Ashton-Rickardt (1978) 1 WLR 37, 43). If knowledge or ignorance of the accused is brought into issue, section 28 opens up that further mental element in the possession in the particular case. But when notice is taken of all these elements in the statute there is no room to infer any additional element of guilty intention in section 5(3).

The second of the certified questions asks whether section 28(3) of the Misuse of Drugs Act 1971 is compatible with Article 6(2) of the Convention. This question raises in the first place a consideration of the nature of the burden of proof imposed on the accused by section 28 of the Act. It should be noticed at the outset that in some cases this problem may not arise at all since no question of the accused's knowledge or ignorance that the thing was a controlled drug may be put in issue. To prove the possession of a controlled drug it is sufficient for the prosecutor to prove that the defendant had control of a thing, that he knew of the existence of the thing, and that the thing was in fact a controlled drug. The evidence may stop at that point and the offence, so far as possession is concerned, may be thereby established. In such a case there is no question of any burden on the accused under section 28. But in some cases more elaboration may be required. That is the qualification upon the scope of section 5(3) which is imported by the words "subject to section 28". In cases where section 28 arises the offence is not established simply by proof of possession. These more elaborate cases may arise particularly where the drug is within a container and, while it is clear that the container contains something, it is not evident from looking at the container what its contents may be.

Section 28(2) uses the words "it shall be a defence for the accused to prove…". It is tempting to hold that this implies merely what is referred to as an "evidential" burden, that is to say, the adducing of evidence sufficient to put the matter in issue. That course is recognised in the case of a variety of common law defences. Here we are dealing with a statutory offence. But even there it may sometimes be possible to preserve the whole burden upon the prosecution, although whether that can be done or not depends on a construction of the particular language used in the statute. In Reg v John [1974] 1 WLR 624 for example in relation to a statutory offence of a failure to provide a specimen "without reasonable excuse", it was recognised that while it was for the defendant to state the excuse it was for the prosecution to negative it.

In the ordinary case, where the presumption of innocence has been recognised at common law long before its embodiment in the Convention, I should be slow to construe a criminal provision so as to impose a persuasive burden upon him. While the Act uses the word "prove" it is perfectly possible to construe that as implying an evidential burden rather than a persuasive one. Section 5(2), the offence of simple possession, is qualified by section 5(4). The essential ingredients for a conviction under section 5(2), as I have already mentioned, are that the accused should have had control of the drug, knew that he had control of it, and that it was in fact a controlled drug, even although he did not know that the thing he possessed was a controlled drug. Section 5(4) applies where "it is proved that the accused had a controlled drug in his possession". It applies then where there was control and knowledge of the existence of the thing, even although the accused did not know that the thing was a controlled drug which in fact it was. Section 5(4) continues that "it shall be a defence for him to prove (a) that, knowing or suspecting it to be a controlled drug" he took possession of it to prevent another from committing an offence. I did not understand it to be contended that that provision imposed a persuasive burden on the accused. But if that provision does not do so it is not immediately easy to see why section 28 should do so, when it also only comes into operation when the ingredients of the offence have all been proved. In neither case is one dealing with the essential ingredients of the statutory offence.

But this approach appears to run counter to what has become generally recognised as the proper construction in England. That approach is fortified by the language of section 28(2) that "it shall be a defence for the accused to prove…". The ordinary meaning of those words imply a persuasive burden of proof even although the accused has only to establish his defence to the standard of a balance of probabilities. There is some force in the argument that the word "prove" is not apt to describe an evidential burden in which the accused is not required to prove anything but simply raise an issue. The legislation may well have been inspired by the observations of Lord Reid in Warner and in Sweet v Parsley . In the former case he observed (p 280:

I proceed accordingly on the basis that section 28(2) imposes a persuasive burden of proof on the accused. But it is to be noted that the practical effect of the burden on the defendant may not be very significant. I agree with the observations of the Lord Justice-General in Salmon v HM Advocate 1999 JC 67, that section 28 does not require that the accused gives evidence. There may, for example, only be evidence of surprise on the part of the defendant on the discovery of the presence of the controlled drug. Or there may only be the exculpatory part of a mixed statement which he gave to the police at some stage. On the other hand, if the burden is a persuasive one the defendant may be ill advised in practice to rest his defence at trial on such a single slender piece of evidence alone. To be persuaded even on a balance of probabilities a jury might well require fortification of such an isolated indication of the defence.

It is only at this stage in the argument in the present case that considerations of the Convention, and in particular of Article 6(2) of the Convention, arise. The appellant contends that the offence contained in section 5 (3) involves the imposition on the accused of a burden of proof such as would be contrary to the provisions of Article 6 of the European Convention on Human Rights and Fundamental Freedoms, and in particular Article 6(2) which states that "everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law". The route adopted by the appellant involves a direct invocation of the Human Rights Act 1998 . But a problem immediately arises. The Human Rights Act 1998 only came into full effect on 2 October 2000. The appellant's trial and conviction took place before that date. Indeed the determination of his appeal occurred before that date. So he has to show that nevertheless he can invoke the provisions of the Act .

In approaching the problem of the retrospectivity of the Human Rights Act 1998 it is to be remembered at the outset that the Act did not incorporate the rights set out in the Convention into the domestic laws of the United Kingdom. The purpose of the Act , as set out in its preamble, was "to give further effect to rights and freedoms" guaranteed under that Convention. The Convention rights have not become part of the constitution so as to obtain any superiority over the powers of Parliament or the validity of primary legislation. The Act requires in section 3 that the language which Parliament has used be construed so as to secure that it is compatible with the Convention rights, so far as it is possible to do so. Section 4 of the Act , which enables the courts to pronounce a declaration of incompatibility, expressly provides in subsection (6) that such a declaration does not affect the validity, continuing operation or enforcement of the legislation declared to be incompatible. But it is left to the discretion of a minister to make a remedial order under section 10 when an incompatibility is identified. One principle achievement of the Act is to enable the Convention rights to be directly invoked in the domestic courts. In that respect the Act is important as a procedural measure which has opened a further means of access to justice for the citizen, more immediate and more familiar than a recourse to the Court in Strasbourg.

The questions then arise: when did Parliament intend that this new procedure should be available, and in respect of what acts should it operate? Parliament had to resolve a question of procedural importance in which considerations of practicality and convenience should be significant. The substance of the Convention rights remained unchanged. For many years they had been directly accessible in Strasbourg. It was considered desirable to allow a period of time to elapse after the passing of the Act and its full effectiveness. One reason for that was no doubt the practical reason of enabling all those concerned with the application of the Act to prepare for the responsibility of that work. Parliament also had to consider how far, if at all, acts done prior to the Act coming into full effect ought to fall within the scope of the new domestic jurisdiction. The line had to be drawn at some point and Parliament had to decide where the line should be drawn for the purposes of the jurisdiction of the courts after the Act came into effect. How Parliament decided to do that is a question of construction of the Act . But in approaching that question it is in my view proper to keep in mind the essentially procedural nature of the question.

Two routes were suggested by the appellant by which he might invoke the provisions of the Act even although his trial, conviction and appeal had all occurred before the Act came fully into effect. One route is by section 22 (4). That section applies to the right provided by section 7 (1)( b ) of someone who claims that a public authority has acted in a way incompatible with a Convention right (and so under section 6 (1) has acted unlawfully) to rely on the Convention right concerned in any legal proceedings provided that he is the victim of the unlawful act. Section 22 (4) extends the application of section 7 (1)( b ) to proceedings brought by or at the instigation of a public authority whenever the act in question took place. So an act contrary to the Convention done by a public authority before 2 October 2000 may be relied upon by the victim in legal proceedings brought by or at the instigation of a public authority from 2 October 2000. If the appellant is seeking to found on section 7 (1)( b ) he must qualify as, to quote the opening words of section 7 , "a person who claims that a public authority has acted (or proposes to act) in a way which is made unlawful by section 6(1)".

The first question is to identify that public authority. Plainly the public authority here cannot be this House, because the House has not yet acted at the stage of the hearing and it will at that stage be far from certain whether the House is proposing to act in an unlawful way. The prosecutor may qualify as a public authority. But the act complained of is not the bringing of the prosecution but the act of the trial judge in directing that a persuasive burden of proof lay on the appellant to establish a defence under section 28(2) or of the Court of Appeal in refusing his appeal. The public authority here must presumably be the trial judge or the Court of Appeal. So the proposed application of these provisions here is to the effect that since the present proceedings are an appeal arising out of a criminal prosecution brought by a public authority (the public prosecutor), the appellant is claiming that a public authority (the trial judge or the Court of Appeal) has acted in a way which is incompatible with his Article 6(2) right, and that he may rely on that right in these proceedings.

I do not find it necessary to determine the question whether the public authority referred to in section 7 (1) and in section 7 (6) and in section 22 (4) requires to be the same public authority. It may be that the public authority referred to in section 7 (1) can be a different public authority from that in sections 7 (6) and 22 (4). I would proceed upon two other considerations. First, section 22 (4) relates only to section 7 (1)( b ). So section 7 (1)( a ) does not apply to an act taking place prior to 2 October 2000. In so far as the act for the purposes of section 7 (1)(a) is a judicial act, the proceedings must be one or other of those set out in section 9 (1). The first of those is the exercising of a right of appeal. So it would appear that a remedy by way of appeal under section 7 (1)( a ) is not to apply to acts prior to 2 October 2000. It does not appear likely that there should be a retrospectivity under section 7 (1)( b ) in respect of an appeal heard from 2 October 2000 against a decision given prior to that date, if there is none under section 7 (I)(a).

Secondly, in terms of section 22 (4) the only acts contravening the Convention prior to 2 October 2000 which can be considered are acts on which reliance may be placed in terms of section 7 (1)( b ). Section 22 (4) states expressly that "otherwise that subsection does not apply to an act taking place before the coming into force of that section". That subsection applies to "legal proceedings" as defined in section 7 (6), that is both "proceedings brought by or at the instigation of a public authority" and an appeal. But the provision admitting reliance on earlier acts in section 22 (4) is limited to "proceedings brought by or at the instigation of a public authority". The use of the same language as was used in section 7 (6) is significant. In my view the intention is that section 22 (4) should not extend to the other kind of "legal proceedings" mentioned in section 7 (6), namely an appeal. I am not persuaded that section 22 (4) can avail the appellant.

The second route which is suggested by the appellant is by a simple application of section 6 (1). Here the House in its judicial capacity is to be seen as the public authority. The argument runs that it is bound to apply the Convention in the present appeal. Its decision must be compatible with the Convention. This is not strictly a point of retroactivity at all. The argument focuses upon an act taking place from 2 October 2000.

The usual understanding of the appeal process is that the correctness of the decision appealed against should be determined in accordance with the law as it stood when the case was decided by the lower court. But on the appellant's approach it would seem that any case either of a civil or a criminal nature, decided according to the law as construed in the ordinary way prior to 2 October 2000, if an appeal was brought so as to be heard after 2 October 2000, would require to be decided by the application of a rule of construction, namely section 3 of the 1998 Act, which was not obligatory on the lower court. But that involves giving an undue extension to the effect of section 3 . In my view section 3 only became obligatory on courts on 2 October 2000. The rule of construction which it expresses applies to all legislation whenever enacted. But there is nothing to show that it was intended by section 3 that the meaning given to a statutory provision by a court prior to 2 October 2000 should be changed in the event of an appeal against that decision being heard on or after that date. It is suggested that the Human Rights Act 1998 should itself be subject to the interpretative regime contained in section 3 . But I am not persuaded that there is any "Convention right" which requires appeals to be determined in that way. It is to be remembered that Article 13 of the Convention is not included within the "Convention rights" for the purposes of the 1998 Act.

The appellant sought to confine the scope of his argument to criminal cases only, but I am not persuaded that there is justification for making that distinction in the Act . Even if one looks only to criminal proceedings it does not seem to me likely that Parliament intended to disturb the usual understanding of the appeal process and allow matters of alleged incompatibility with the Convention rights, which could always have been raised in Strasbourg, to be opened up by the taking of an appeal to a higher court on or after the date on which the Act came into effect. In general Acts of Parliament should not be read as operating so as to affect things done prior to their coming into effect. I see no reason why that principle should not apply to the Human Rights Act 1998 . If a departure from the usual course was intended I would expect that to have been clearly stated.

Furthermore, if section 6 (1) is to be construed as requiring courts of appeal to apply the Convention to acts which occurred prior to 2 October 1998, that would not seem to me to be consistent with the careful and precise provision which Parliament did make for the extent to which acts prior to 2 October 1998 could be relied upon. Section 22 (4) is expressly limited to one situation. It relates only to section 7 (1)(b), it applies only to proceedings brought by or on behalf of a public authority, and it applies to acts "whenever the act in question took place", that is to say whether before or after the coming into force of the Act . Section 22 (4) then expressly states with regard to section 7 (1)( b ) that "otherwise that subsection does not apply to an act taking place before the coming into force of that section". As I have already decided, section 7 (1)( b ) does not allow a person to rely on a Convention right allegedly breached prior to 2 October 2000 in an appeal heard on or after that date. So the argument reaches the result that under section 6 (1) the appeal court is bound to take account of that Convention right, although the appellant may not rely upon it. That does not seem to me to be a likely interpretation of the intention of Parliament in passing section 6 (1) and it does not seem to fit comfortably with the express provision made in section 22 (4).

I accept that section 6 (1) imposes an obligation on a court, including, to use the language of section 6 (4), this House in its judicial capacity, to act in conformity with the Convention. But section 6 (1) does not provide any remedy. It simply explains that, where a public authority acts in a way which is incompatible with a Convention right, that act is unlawful. That word might mean that the act might be the basis for an offence. But that is obviously not the meaning to be given to it here. Indeed section 7 (8) states that "Nothing in this Act creates a criminal offence". Another meaning which may be ascribed to the word "unlawful" is that it is to be unenforceable. But if one breaks off at that point one is left without any means of obtaining any remedy. It seems to me that section 6 cannot be taken in isolation but has to be read and understood along with the rest of the Act and in particular with the sections which immediately follow it.

As I have already noted section 7 (1) makes provision for the proceedings which may be used by a victim. Para ( a ) enables him to bring proceedings against the authority in the appropriate court or tribunal. Para ( b ) enables him to rely on the Convention right in any legal proceedings, and those include (subsection (6)), proceedings brought by or at the instigation of a public authority and an appeal But we are here dealing with a judicial act. The Act specifically provides in section 9 that there are only three ways in which proceedings under section 7 (1)( a ) may be brought. These are "(a) by exercising a right of appeal; (b) on appeal, an application for judicial review, or (c) in such other forum as may be prescribed by rules". I have some difficulty in applying section 9 to a decision of this House but if there is a difficulty there it would not justify the construction of section 6 which is sought by the appellant in the present case. In my view section 6 is part of the series of sections on public authorities and cannot be isolated so as to be independent of them. The better view to my mind is that the only way in which an act prior to 2 October 2000 can come within the scope of the Act is by section 22 (4).

The construction of the 1998 Act which I have preferred is given some support by reference to the corresponding situation which arose in New Zealand in Minto and Cuthbert v Police ((1990-92) NZBORR 208). In that case appeals were taken against convictions for offences of breach of the peace committed by people staging a demonstration. The offences and the convictions occurred prior to the coming into force of the Bill of Rights Act 1990. The appeals were heard after that Act came into force and the appellants sought to found on the right to freedom of peaceful assembly guaranteed by section 16 of that Act. It was held that section 16 should not be given such a retrospective effect. Robertson J said in reference to the appellant's counsel (p 214):

In my view accordingly it is not open to the appellants to invoke the provisions of the Human Rights Act in the present appeal. But it is nevertheless appropriate to consider for the future whether the provisions of section 28 are compatible with Article 6(2) of the Convention.

It may be noted at the outset that cases may well occur under section 5 of the Misuse of Drugs Act where no problem of compliance with Article 6(2) will arise. In the case of the possession of controlled drugs Parliament has provided for the mental element of the offence in two respects. First, there is a mental element built into the offence itself in the matter of possession. An ingredient of that element, as I have already stated, is knowledge of the existence of the thing in issue In the simplest case the only mental element which may arise is knowledge of the existence of the thing which is in fact a controlled drug. That is an essential of possession and the prosecution is required to prove that the accused had that knowledge. In some cases the inquiry may not have to go any further. Any consideration of the knowledge or ignorance of the accused about the nature of the thing which he is proved to have possessed may not arise. Indeed in some cases it may be indisputable that if there was possession the accused must have known the nature of what he possessed. In such a case no question of that kind of knowledge requires to be explored. If the inquiry stops at that point, the offence will have been proved. The onus of proof remained upon the prosecution and no onus has ever passed to the accused. In such circumstances there can be no contravention of Article 6(2) of the Convention. But the present case is one where the accused relied upon section 28(3)( b )(i) and since I am proceeding upon the basis that the burden was on him to prove his assertion of ignorance under that provision there is a real question whether the offence prescribed in section 5(3), subject as it is to the provisions of section 28, contravenes Article 6(2) of the Convention in cutting across the presumption of innocence.

It is necessary first to say a word about the application of that Article. It states "Everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law". There is no doubt that it does apply to the present proceedings. Article 6(2) applies to any person charged with a criminal offence. So the appellant may invoke the benefit of the provision. But the imposition of a burden on the defendant does not necessarily involve a contravention of the Convention. Some guidance on the problem can be obtained from the jurisprudence of the ECHR. First, one must look not only at the form of the legislation but also at its substance and effect ( X v UK App No 5124/71, July 1972). As regards the identification of the essentials of an offence the Privy Council have recognised that a similar approach should be taken. What is decisive is "the substance and reality of the language creating the offence rather than its form" ( A-G of Hong Kong v Lee Kwong Kut [1993] AC 951, 969). Secondly:

Where a rebuttable presumption is placed on the accused that may still not breach the Convention right. In AG v Malta (App 16641/90) the statutory provision made a director of a company guilty of an offence committed by the company "unless he proves that the offence was committed without his knowledge…." That was held by the domestic court not to be unreasonable since otherwise it would be impossible in the majority of cases to prove its case against a company. The Commission declared the application inadmissible. They proceeded on the consideration that the presumption created by the provision was rebuttable, that the Maltese courts enjoyed a genuine freedom of assessment and that there was no indication that the provision was applied to the applicant in a manner incompatible with the presumption of innocence.

But the decision may have to turn upon the particular circumstances. The focus of attention paid by the European Court of Human Rights tends to be directed at the particular circumstances of the case before them. Considerations which have weighed with the Court in deciding that in particular cases that there was no infringement have included a consideration of the whole evidence and such matters as these: that while the accused was deemed liable for the offence, he had the opportunity to put extenuating circumstances before the court and was entitled to be acquitted if he succeeded in establishing a case of force majeure, unavoidable mistake, or necessity ( Salbiaku , and Pham Hoang ), that the domestic courts had also identified in the case an element of intent, even although they were under no legal obligation to do so ( Salabiaku ), and that the courts had also made a careful discrimination between the two charges originally brought against the accused and acquitted him of one of them while convicting him on the other ( Salabiaku ). But it would be of little assistance for the future to decide the present case simply on its own particular facts. Some more general opinion ought to be possible.

Reasons can readily be adduced to support the imposition of the burden of proof on the accused in the present context. Firstly, the question whether the accused was ignorant or had no reason to suspect that what he possessed was a controlled drug is a matter very much within his own knowledge. There are sound practical reasons for imposing the burden on him to prove his ignorance. Secondly, the proof may be relatively easy for him, as I have already noted. Thirdly, there is a serious consideration of the public interest in the discouragement of what is well recognised as a grave social evil, the unlawful distribution of controlled drugs. Fourthly the knowledge of the defendant of the nature of what he possessed is brought in as a defence, not as an ingredient of the offence. In some cases it may never arise. It can be strongly argued that a transfer of a persuasive burden of proof onto the defendant under section 28 could be compatible with Article 6(2).

But while it might seem reasonable for such considerations to let the accused bear the burden of proof I do not consider that such a result can be justified when one weighs the considerations of what is, or at least may be, at stake for the accused and the interests of the public. As I have already noted, in order to be acceptable a presumption must fall within limits which "take into account the importance of what is at stake and maintain the rights of the defence". If the matter is approached as one of generality one can make no useful distinction here between the various classes of drugs which may be involved. In the most serious cases the accused may face a sentence of life imprisonment. A strict responsibility may be acceptable in the case of statutory offences which are concerned to regulate the conduct of some particular activity in the public interest. The requirement to have a licence in order to carry on certain kinds of activity is an obvious example. The promotion of health and safety and the avoidance of pollution are among the purposes to be served by such controls. These kinds of cases may properly be seen as not truly criminal. Many may be relatively trivial and only involve a monetary penalty. Many may carry with them no real social disgrace or infamy.

In Sweet v Parsley (1970) AC 132 at 148 Lord Reid observed that "there has for centuries been a presumption that Parliament did not intend to make criminals of persons who were in no way blameworthy in what they did". The advent of the Human Rights Act 1998 has certainly sharpened a consciousness of the human right which is embodied in the presumption of innocence and invites a closer scrutiny of what Ashworth and Blake ( The Presumption of Innocence in English Criminal Law , (1996) CLR 314) have described as a large-scale derogation from basic principle They quote the advice of the Eleventh Report of the Criminal Law Revision Committee that "both on principle and for the sake of clarity and convenience in practice, burdens on the defence should be evidential only". The 1998 Act should encourage a reconsideration of a trend which has for over a decade been exposed to powerful criticism.

While it may be that offences under section 5 of the Misuse of Drugs Act may be described as regulatory they can lead to the most serious of consequences for the accused. Of course trafficking in controlled drugs is a notorious social evil, but if any error is to be made in the weighing of the scales of justice it should be to the effect that the guilty should go free rather than that an innocent person should be wrongly convicted. By imposing a persuasive burden on the accused it would be possible for an accused person to be convicted where the jury believed he might well be innocent but have not been persuaded that he probably did not know the nature of what he possessed. The jury may have a reasonable doubt as to his guilt in respect of his knowledge of the nature of what he possessed but still be required to convict. Looking to the potentially serious consequences of a conviction at least in respect of class A drugs it does not seem to me that such a burden is acceptable.

But I have no difficulty in finding the solution by an application of section 3 of the Human Rights Act 1998 . It requires no straining of the language of section 28 to construe the references to proof as intending an evidential burden and not a persuasive one. Indeed, as I have already stated, it would be a construction to which I would in any event have inclined, even without the added compulsion of the Human Rights Act . The construction seems to me to be something which is well within the scope of what is "possible" for the purposes of section 3 . In rewriting the legislation as it did in the Act of 1971, Parliament must have recognised that in fairness to the accused provision must be made for the case where he was ignorant of the nature of the thing which he possessed and that the law as laid down by the majority of this House in Warner was unduly harsh. It seems to me that the proper way by which that harshness should be alleviated is to recognise that the accused should have the opportunity to raise the issue of his knowledge but to leave the persuasive burden of proof throughout on the prosecution. Respect for the "golden thread" of the presumption of innocence deserves no less.

I am not persuaded that this approach gives rise to any practical problem. In some cases no issue may arise at all about the accused's knowledge of the nature of thing which he is alleged to possess. Section 28 then will not come into play and it would only distract a jury from the point in issue to give any direction, or certainly any detailed direction, about it. When the matter of the accused's knowledge (including in that word for convenience the matters of suspicion and reason to suspect which are detailed in section 28) does arise, such burden as there is on the defence is discharged by the defence expressly raising knowledge as an issue. Where that occurs the judge will require to explain the substance of section 28 to the jury but will also simply remind them that the burden of proving guilt is throughout on the prosecution. If the jury are satisfied on the matter of possession but are left with a reasonable doubt on the matter of his knowledge or suspicion of the existence of a fact which the prosecution has to prove, such as, for example, the existence of the controlled drug, then they should acquit. If the jury are satisfied beyond reasonable doubt that the accused possessed the substance or product in question but are not satisfied beyond reasonable doubt that he knew that it was a controlled drug (or suspected or had reason to suspect that it was) then again they should acquit him. They can only convict if they are satisfied beyond reasonable doubt that the prosecution has proved possession of the controlled drug and, if the issue is raised, that the lines of defence set out in section 28 are without foundation.

I do not consider that this understanding of the offences in question in the future should cause any serious problem for any past or current cases. Ultimately what is in issue is the fairness of the trial. No doubt in many cases an unfair trial in contravention of Article 6 will constitute an unsafe conviction (see for example R v Togher 9 Nov 2000 R v Forbes 14 Dec 2000). But an unfairness is not always fatal to an conviction. In particular in the present context a direction that there is a persuasive burden on the accused to establish a defence under section 28 will not necessarily lead to an unsafe conviction. If there is doubt about guilt then the conviction must be held to be unsafe. But if there is no doubt about guilt it is not every case where an unfairness can be identified that will necessarily and inevitably lead to a quashing of the conviction.

The present appeal has been more concerned with generalities than with the particular facts and circumstances of the particular case which has been used as a vehicle for raising them. If one turns to the facts here, which I need not repeat as they have been narrated by others of your Lordships, I am not persuaded that there is any ground for doubting that the verdict of the jury, who rejected the defence of duress, would also have rejected the defence under section 28 even had it been explained to them that the burden of proof remained throughout on the prosecution. I see no room for quashing the conviction.

For the foregoing reasons I would dismiss the appeal.

The appellant, Steven Lambert, was convicted on 9 April 1999 in the Crown Court at Warrington after a trial before Judge Hale and a jury of the offence of possession of a controlled class A drug with intent to supply, contrary to section 5(3) of the Misuse of Drugs Act 1971 and was sentenced to seven years' imprisonment.

In summing up to the jury the trial judge succinctly summarised the facts relied on by the prosecution:

In the course of his summing up the trial judge directed the jury pursuant to section 28(2) and (3) of the 1971 Act that the burden of proof was on the appellant to prove on the balance of probabilities that he neither believed nor suspected nor had reason to suspect that the substance in the bag was a controlled drug.

Article 6(2) of the Convention for the Protection of Human Rights and Fundamental Freedoms provides:

However, the Court of Appeal accepted (though with reservations) that, as agreed by all counsel, the Human Rights Act 1998 , which, apart from four sections, came into force on 2 October 2000, more than a year after the date of the appellant's conviction at the Crown Court, did have a retrospective effect so far as article 6 was concerned. Therefore the first question which arises on this appeal is whether the 1998 Act has a retrospective effect. If it does not, the applicant cannot rely on Article 6(2) in the English courts and his appeal must fail.

Section 6 of the 1998 Act provides:

Mr Starmer further submitted that the Crown could not counter this argument by relying on section 6 (2)(a) or (b) and contending that the House itself was bound to give effect to the provisions of section 28 of the 1971 Act, because under section 3 of the 1998 Act the House could read and give effect (as the Crown accepted) to section 28 in a way compatible with Article 6(2).

The argument which Mr Starmer skilfully advanced is a powerful one, but I have come to the conclusion that it should not be accepted. It is a well established principle that no statute should be construed so as to have a retrospective operation unless its language is such as plainly to require such a construction. In addition, save as to the proceedings described in the first part of section 22 (4), I consider that that subsection supports the view that the 1998 Act is not to have a retrospective effect. In my opinion an Act has a retrospective effect if it operates to make unlawful or unsafe a conviction which was lawful and safe at the time it was imposed. This is the result for which the appellant contends. Before the commencement of the 1998 Act he had been lawfully convicted and his conviction was safe. But he submits that the effect of section 6 of the 1998 Act, coming into operation after his conviction, is to make the conviction unlawful and unsafe. In my opinion it is no answer for the appellant to maintain that he is only concerned with the lawfulness of a decision taken by the Appellate Committee on a date after 2 October 2000. I consider that this argument does not alter the reality that if the House were to quash the conviction it would be giving a retrospective effect to section 6 .

Nor do I think that the appellant can derive assistance from section 3 of the 1998 Act which provides:

Mr Starmer's second main submission relied on the combined effect of sections 6 , 7 and 22 . Under section 7 (1)(b) a person who claims that a court has acted in a way which is made unlawful by section 6 (1) may rely on the Convention right concerned in any legal proceedings. Section 22 (4) provides that paragraph (b) of subsection (1) of section 7 applies to proceedings brought by or at the instigation of a public authority whenever the act in question took place. Therefore the submission was that the appeal before the House was "legal proceedings" and that the appellant could rely on Article 6(2) because the prosecution ("proceedings") was brought against the appellant by the Director of Public Prosecutions, a public authority, notwithstanding that the conviction by the Crown Court took place before the 1998 Act came into force.

In my opinion this argument fails because section 7 (6) provides:

Although the 1998 Act was passed on 9 November 1998, its sections, with the exception of sections 18 , 21 (5) and 22 , did not come into force until 2 October 2000, the day appointed by the Secretary of State pursuant to section 22 (3). I consider that this delayed commencement of the great majority of the sections in the Act suggests that it was the intent of Parliament that the Act , (save in the limited circumstances set out in section 22 (4)), should not operate retrospectively, but that the sections were only to operate in respect of acts taking place subsequent to 2 October 2000 after public authorities in the United Kingdom had had the opportunity to ensure that their actions and procedures would not violate Convention rights. I think that Parliament did not intend to bring about the situation which was a cause of concern to Rose LJ, the Vice-President of the Criminal Division, in R v Kansal [2001] All ER (D) 311 (May) and which he described in paragraph 24 of the judgment of the Court of Appeal:

In reaching my conclusion on this difficult issue of retrospectivity I have taken into account the opinions in favour of retrospectivity expressed in R v Director of Public Prosecutions Ex parte Kebilene , by Lord Bingham of Cornhill CJ in the Divisional Court and by Lord Steyn in this House, and the opinions against retrospectivity expressed by Lord Hobhouse of Woodborough in Ex parte Kebilene and by Waller LJ in Parker v Director of Public Prosecutions . I have also taken into account the judgment of the Court of Appeal in Wilson v The First County Trust Ltd (2nd May 2001).

Mr Starmer relied on the following passage in the judgment of Sir Andrew Morritt V-C at paragraph 17:

Therefore I consider that this appeal should be dismissed on the grounds that the appellant's conviction was safe prior to 2 October 2000 and the provisions of the 1998 Act do not require an appellate court giving its decision after 2 October 2000 to declare it unsafe. However, as detailed submissions were advanced to the House on the basis that the appellant could rely on sections 6 , 7 and 22 (4) of the 1998 Act, I propose to state my opinion on those submissions.

Section 5 of the 1971 Act provides:

The essence of the argument advanced on behalf of the appellant is contained in two propositions. The first proposition is that the decision of this House in B v DPP [2000] 2 AC 428 established that mens rea was an essential element of every criminal offence unless Parliament expressly or by implication provided to the contrary. Knowledge on the part of a defendant that the article or substance which he was carrying or had under his control was a controlled drug was an essential ingredient of the offence of possession of a controlled drug with intent to supply contrary to section 5 (3) of the 1971 Act. The second proposition is that subsections 28(2) and (3) transfer the onus of proof to a defendant by requiring him to prove that he neither knew nor suspected nor had reason to suspect that the article or substance which he was carrying was a controlled drug, and therefore the subsections were in breach of Article 6(2) as they reversed the burden of proof in respect of an essential ingredient of the offence.

The argument advanced on behalf of the Crown is that the essential ingredients which the prosecution has to prove to establish the offence under section 5 (3) in a case such as that against the appellant is possession of the bag and its contents, which required proof that the defendant knew that he had control over the bag and whatever object or substance was inside it, and also proof that the contents of the bag were, in fact, a controlled drug. Subsections 28(2) and (3) provide a defence which arises only after the prosecution has proved the essential ingredients of the offence, and therefore there is no transfer of the burden of proof which infringes the presumption of innocence set out in Article 6(2).

Before considering these opposing submissions it is necessary to have regard to the authorities in relation to the offences of possession of controlled drugs under section 5 (2) and 5 (3). In R v Boyesen [1982] AC 768, 773G Lord Scarman stated:

In my respectful opinion the law in relation to section 5 (3) and section 28(2) and (3) was correctly stated by Lord Lane, save that the illuminating analysis of section 28 by the Lord Justice General, Lord Rodger, in Salmon v HM Advocate demonstrates that where a defendant advances the defence that he did not know that the bag or other container which he was carrying contained a controlled drug and believed it contained a different type of article such as a video film, this defence arises under section 28(2) and not under section 28(3).

It is also relevant to observe that subsections 28(2) and (3) expressly impose not merely an evidential burden on the defendant but a persuasive (or legal) burden. Therefore the appellant derives no assistance from B v Director of Public Prosecutions as it is clear that Parliament has expressly provided that the prosecution need not prove the mental element of knowledge. A persuasive burden is one where the matter in question must be taken as proved against the defendant unless he satisfies the jury on the balance of probabilities to the contrary. An evidential burden is one where the matter must be taken as proved against the defendant unless there is sufficient evidence to raise an issue on the matter but, if there is sufficient evidence, then the burden rests on the prosecution to satisfy the jury as to the matter beyond reasonable doubt.

As I have observed, the submission advanced by the Crown is that (a) Parliament has chosen by section 5 (3) to create an offence in which knowledge by the defendant that he has custody or control of a controlled drug is not an essential element, and that the issue of lack of knowledge by the defendant only arises as a defence under subsections 28(2) or (3), and (b) accordingly under European jurisprudence there is no violation of the presumption of innocence given by Article 6(2). Viewed in the context of English law alone I consider that proposition (a) is correct, but I am unable to accept proposition (b) because I think that a similar submission on behalf of the Government of France was not wholly accepted by the European Court of Human Rights in Salabiaku v France [1988] 13 EHRR 379. In that case a presumption of criminal liability was laid down in Article 392(1) of the French Customs Code for every person who was found in possession of prohibited goods. Mr Salabiaku complained of a violation of Article 6(2). The argument advanced on behalf of the French Government was very similar to the argument advanced on behalf of the Crown in the present case. In paragraph 27 of its judgment the Court recorded the argument on behalf of the French Government as follows:

The Court held that this argument could not be accepted in its entirety and stated in paragraph 28:

Therefore, following the jurisprudence of the Court, I consider that the Crown cannot rebut an argument based on a violation of Article 6(2) by simply contending that the Government of the United Kingdom is entitled "to define the constituent elements of the . . . offence", and that a violation of Article 6(2) is avoided because the 1971 Act makes absence of knowledge of being in possession of a controlled drug a defence rather than making knowledge an ingredient of the offence which the prosecution has to prove.

Whether or not there has been a violation of Article 6(2) depends, therefore, upon whether the presumption created by section 28(2) and (3) is confined within reasonable limits which take into account the importance of what is at stake and maintains the rights of the defence.

In X v United Kingdom (19 July 1972) the applicant had been convicted of living on the immoral earnings of a prostitute. He complained of a violation of Article 6(2) on the ground that the statutory provision under which he was convicted provided that " . . . a man who lives with or is habitually in the company of a prostitute . . . shall be presumed to be knowingly living on the earnings of prostitution unless he proves the contrary." The Commission held that his complaint was manifestly ill-founded and stated in its decision:

In AG v Malta (10th December 1991) the applicant complained that he had been deprived of the presumption of innocence guaranteed by Article 6(2) by a statutory provision which provided that a director of a company is presumed guilty of an offence committed by the company unless he proves that the offence was committed without his knowledge and that he exercised all due diligence to prevent the commission of the offence. The applicant brought an application before the Constitutional Court of Malta claiming that the provision was in conflict with the article of the Constitution of Malta which guaranteed the presumption of innocence. His application was dismissed by the Constitutional Court and the European Commission also rejected his application and stated in its decision:

In a number of judgments the European Court has given guidance as to the extent to which a Convention right may be qualified by national legislation and these judgments were considered by the Privy Council in Brown v Stott [2001] 2 WLR 817 . Lord Bingham of Cornhill stated at p 836B:

Therefore in considering whether a rebuttable presumption of

It is clear from the decisions of the European Commission in X v United Kingdom and AG v Malta that the difficulty of proving knowledge on the part of the defendant is one of the factors which can justify the creation of a presumption against a defendant, where the presumption is neither irrebuttable nor unreasonable.

I am, with respect, unable to agree with the view that the problem of obtaining a conviction against a guilty person can be surmounted by imposing a evidential burden on the defendant. All that a defendant would have to do to discharge such a burden would be to adduce some evidence to raise the issue that he did not know that the article in the bag or the tablets on the table were a controlled drug, and the prosecution would then have to destroy that defence in such a manner as to leave in the jury's mind no reasonable doubt that the defendant knew that it was a controlled drug in the bag or on the table. Blackstone's Criminal Practice (2001) states at para F3.7, page 1996:

Support for the appellant's argument is to be found in the opinion of the Criminal Law Revision Committee in paragraph 140 of its Eleventh Report that: "both on principle and for the sake of clarity and convenience in practice, burdens on the defence should be evidential only." But in my opinion the threat of drugs to the well being of the community and the peculiar difficulty of proving knowledge in such cases justifies an exception to the general principle advocated by the Criminal Law Revision Committee, and this was clearly the view taken by Parliament in enacting sections 5 and 28 of the 1991 Act. Moreover the transfer of the burden of proof as to knowledge in drugs cases has the powerful support of Lord Reid and Lord Pearce in Warner v Metropolitan Police Commissioner [1969] 2 AC 256 to which I later refer.

The argument advanced against the imposition of a persuasive burden is that it creates the position that where a defendant fails to satisfy the jury on the balance of probabilities that he did not know he was carrying drugs, the jury will or may convict him notwithstanding that they have a reasonable doubt as to whether or not he had that knowledge. In theory there is force in this argument, but in my respectful opinion there is greater force in the common sense view of Lord Pearce in Sweet v Parsley [1970] AC 132 where the defendant was charged under section 5(b) of the Dangerous Drugs Act 1965 with being concerned in the management of premises used for the purpose of smoking cannabis resin. Lord Pearce said at p 157D:

Clayton and Tomlinson in The Law of Human Rights state at para 6.37:

In the course of their speeches the members of the House considered at length the issue which arises when such an defence is advanced in a drugs case. In the course of his speech Lord Pearce said at p 307B:

My Lords, when judges of such eminence considered that transferring the burden of proof in relation to knowledge would not result in an unfair trial to the defendant, I consider that thirty years later when the problem has not changed there is no reason for this House to take a different view. Section 2 of the 1998 Act now requires the House in determining a question which has arisen in connection with a Convention right to take into account judgments of the European Court and decisions of the European Commission, but in my opinion the judgments and decisions to which I have referred provide no basis for the view that under the jurisprudence of the European Court the transfer of the onus of proof as to knowledge in drugs cases would constitute a violation of Article 6(2).

Therefore my conclusion is that the difficulty in some cases of convicting those guilty of the crime of possession of a controlled drug with intent to supply, if the burden of proving knowledge beyond a reasonable doubt rests on the prosecution, is not resolved by placing an evidential burden on the defendant, and that it is necessary to impose a persuasive burden as subsections 28(2) and (3) do. I further consider that the transfer of the onus satisfies the test that it has a legitimate aim in the public interest and that there is a reasonable relationship of proportionality between the means employed and the aim sought to be realised. Accordingly I am of the opinion that subsections 28(2) and (3) do not violate Article 6(2) and I am in full agreement with the Court of Appeal on this issue.

However if, contrary to the view which I have taken, the imposition of the persuasive burden on the defendant in relation to knowledge constitutes an infringement of the presumption of innocence, a further issue arises for consideration. The case against the appellant was an extremely strong one and I agree with my noble and learned friend Lord Steyn that even if the judge had directed the jury on the basis that only an evidential burden rested on the appellant, the jury would have been bound to convict. Therefore the question arises whether, notwithstanding that the jury would have certainly convicted him on such a direction, it would be right for an appellate court to hold that there had been a violation of Article 6(2) resulting in an unfair trial and that consequently the conviction was unsafe and must be quashed. One answer would be to hold that there would be a breach of Article 6(2) if, on ordinary principles of construction, sections 5 and 28(2)(3) have the effect of imposing a persuasive burden, and that accordingly under section 3 of the 1998 Act those sections should be read and given effect in a way which is compatible with Article 6(2), so that they impose only an evidential burden, and then to hold that a jury would have been bound to convict if they had been directed on the basis that only an evidential burden rested on the defendant with the result that the conviction was not unsafe.

However I would take a different view and I would accept the second of the written propositions submitted by counsel for the appellant:

In Hoang v. France 1992 16 EHRR 53, 78-80 where the Court held that there was no violation of article 6, it stated at paragraph 33:

There will be cases where a conviction cannot stand and must be quashed irrespective of the strength of the evidence against the defendant because the trial as a whole is judged to be unfair. But in my opinion in this case where the unfairness is claimed to arise from the transfer of the onus and it was open to the appellant to seek to rebut the presumption, and where there can be no doubt that the jury would have convicted if only an evidential burden had rested on the appellant, then the imposition of a persuasive burden as to knowledge resulted in no injustice and accordingly, in my opinion no breach of Article 6(2).

For the reasons which I have given I would answer the first certified question in accordance with paragraphs 21-23 of this speech, the second certified question "No", and the third certified question "No" in relation to an alleged breach of his Convention rights by the trial court and I would dismiss the appeal.