Justice For All Australians

May the Special Immigration Appeals Commission ("SIAC"), a superior court of record established by statute, when hearing an appeal under section 25 of the Anti-terrorism, Crime and Security Act 2001 by a person certified and detained under sections 21 and 23 of that Act, receive evidence which has or may have been procured by torture inflicted, in order to obtain evidence, by officials of a foreign state without the complicity of the British authorities? That is the central question which the House must answer in these appeals. The appellants, relying on the common law of England, on the European Convention on Human Rights and on principles of public international law, submit that the question must be answered with an emphatic negative. The Secretary of State agrees that this answer would be appropriate in any case where the torture had been inflicted by or with the complicity of the British authorities. He further states that it is not his intention to rely on, or present to SIAC or to the Administrative Court in relation to control orders, evidence which he knows or believes to have been obtained by a third country by torture. This intention is, however, based on policy and not on any acknowledged legal obligation. Like any other policy it may be altered, by a successor in office or if circumstances change. The admission of such evidence by SIAC is not, he submits, precluded by law. Thus he contends for an affirmative answer to the central question stated above. The appellants' case is supported by written and oral submissions made on behalf of 17 well-known bodies dedicated to the protection of human rights, the suppression of torture and maintenance of the rule of law.

The appeals now before the House are a later stage of the proceedings in which the House gave judgment in December 2004: A and others v Secretary of State for the Home Department, X and another v Secretary of State for the Home Department [2004] UKHL 56, [2005] 2 AC 68. In their opinions given then, members of the House recited the relevant legislative provisions and recounted the relevant history of the individual appellants up to that time. To avoid wearisome repetition, I shall treat that material as incorporated by reference into this opinion, and make only such specific reference to it as is necessary for resolving these appeals.

The 2001 Act was this country's legislative response to the grave and inexcusable crimes committed in New York, Washington DC and Pennsylvania on 11 September 2001, and manifested the government's determination to protect the public against the dangers of international terrorism. Part 4 of the Act accordingly established a new regime, applicable to persons who were not British citizens, whose presence in the United Kingdom the Secretary of State reasonably believed to be a risk to national security and whom the Secretary of State reasonably suspected of being terrorists as defined in the legislation. By section 21 of the Act he was authorised to issue a certificate in respect of any such person, and to revoke such a certificate. Any action of the Secretary of State taken wholly or partly in reliance on such a certificate might be questioned in legal proceedings only in a prescribed manner.

Sections 22 and 23 of the Act recognised that it might not, for legal or practical reasons, be possible to deport or remove from the United Kingdom a suspected international terrorist certified under section 21, and power was given by section 23 to detain such a person, whether temporarily or indefinitely. This provision was thought to call for derogation from the provisions of article 5(1)(f) of the European Convention, which it was sought to effect by a Derogation Order, the validity of which was one of the issues in the earlier stages of the proceedings.

Section 25 of the Act enables a person certified under section 21 to appeal to SIAC against his certification. On such an appeal SIAC must cancel the certificate if "(a) it considers that there are no reasonable grounds for a belief or suspicion of the kind referred to in section 21(1)(a) or (b), or (b) it considers that for some other reason the certificate should not have been issued". If the certificate is cancelled it is to be treated as never having been issued, but if SIAC determines not to cancel a certificate it must dismiss the appeal. Section 26 provides that certifications shall be the subject of periodic review by SIAC.

SIAC was established by the Special Immigration Appeals Commission Act 1997, which sought to reconcile the competing demands of procedural fairness and national security in the case of foreign nationals whom it was proposed to deport on the grounds of their danger to the public. Thus by section 1 (as amended by section 35 of the 2001 Act) SIAC was to be a superior court of record, now (since amendment in 2002) including among its members persons holding or having held high judicial office, persons who are or have been appointed as chief adjudicators under the Nationality, Immigration and Asylum Act 2002, persons who are or have been qualified to be members of the Immigration Appeal Tribunal and experienced lay members. All are appointed by the Lord Chancellor, who is authorised by section 5 of the Act to make rules governing SIAC's procedure. Such rules, which must be laid before and approved by resolution of each House of Parliament, have been duly made. Such rules may, by the express terms of sections 5 and 6, provide for the proceedings to be heard without the appellant being given full particulars of the reason for the decision under appeal, for proceedings to be held in the absence of the appellant and his legal representative, for the appellant to be given a summary of the evidence taken in his absence and for appointment by the relevant law officer of a legally qualified special advocate to represent the interests of an appellant in proceedings before SIAC from which the appellant and his legal representative are excluded, such person having no responsibility towards the person whose interests he is appointed to represent.

The rules applicable to these appeals are the Special Immigration Appeals Commission (Procedure) Rules 2003 (SI 2003/1034). Part 3 of the Rules governs appeals under section 25 of the 2001 Act. In response to a notice of appeal, the Secretary of State, if he intends to oppose the appeal, must file a statement of the evidence on which he relies, but he may object to this being disclosed to the appellant or his lawyer (rule 16): if he objects, a special advocate is appointed, to whom this "closed material" is disclosed (rule 37). SIAC may overrule the Secretary of State's objection and order him to serve this material on the appellant, but in this event the Secretary of State may choose not to rely on the material in the proceedings (rule 38). A special advocate may make submissions to SIAC and cross-examine witnesses when an appellant is excluded and make written submissions (rule 35), but may not without the directions of SIAC communicate with an appellant or his lawyer or anyone else once the closed material has been disclosed to him (rule 36). Rule 44(3) provides that SIAC "may receive evidence that would not be admissible in a court of law". The general rule excluding evidence of intercepted communications, now found in section 17(1) of the Regulation of Investigatory Powers Act 2000, is expressly disapplied by section 18(1)(e) in proceedings before SIAC. SIAC must give written reasons for its decision, but insofar as it cannot do so without disclosing information which it would be contrary to the public interest to disclose, it must issue a separate decision which will be served only on the Secretary of State and the special advocate (rule 47).

Of the 10 appellants now before the House, all save 2 were certified and detained in December 2001. The two exceptions are B and H, certified and detained in February and April 2002 respectively. Each of them appealed against his certification under section 25. Ajouaou and F voluntarily left the United Kingdom, for Morocco and France respectively, in December 2001 and March 2002, and their certificates were revoked following their departure. C's certificate was revoked on 31 January 2005 and D's on 20 September 2004. Abu Rideh was transferred to Broadmoor Hospital under sections 48 and 49 of the Mental Health Act 1983 in July 2002. Conditions for his release on bail were set by SIAC on 11 March 2005, and on the following day his certificate was revoked and a control order (currently the subject of an application for judicial review) was made under the Prevention of Terrorism Act 2005, enacted to replace Part 4 of the 2001 Act. Events followed a similar pattern in the cases of E, A and H, save that none was transferred to Broadmoor and notice of intention to deport (currently the subject of challenge) was given to A and H in August 2005, since which date they have been detained. The control orders made in their cases were discharged. B's case followed a similar course to A's, save that he was transferred to Broadmoor under sections 48 and 49 of the 1983 Act in September 2005. In the case of G, bail conditions were set by SIAC in April 2004 and revised on 10 March 2005. His certificate was revoked and a control order made under the 2005 Act on 12 March 2005. He was given notice of intention to deport (which he is challenging) on 11 August 2005, and he has since been detained. His control order was discharged.

The appellants' appeals to SIAC under section 25 of the 2001 Act were heard in groups between May and July 2003. During these hearings argument and evidence were directed both to general issues relevant to all or most of the appeals and to specific issues relevant to individual cases. SIAC heard open evidence when the appellants and their legal representatives were present and closed evidence when they were excluded but special advocates were present. On 29 October 2003 judgments were given dismissing all the appeals. There were open judgments on the general and the specific issues, and there were also closed judgments. On the question central to these appeals to the House, raised in its present form when the proceedings before it were well advanced, SIAC gave an affirmative answer: the fact that evidence had, or might have been, procured by torture inflicted by foreign officials without the complicity of the British authorities was relevant to the weight of the evidence but did not render it legally inadmissible. In lengthy judgments given on 11 August 2004, a majority of the Court of Appeal (Pill and Laws LJJ, Neuberger LJ in part dissenting) upheld this decision: [2004] EWCA Civ 1123, [2005] 1 WLR 414. Despite the repeal of Part 4 of the 2001 Act by the 2005 Act, the appellants' right of appeal to the House against the Court of Appeal's decision under section 7 of the 1997 Act is preserved by section 16(4) of the Prevention of Terrorism Act 2005, and no question now arises as to the competency of any of these appeals.

The appellants submit that the common law forbids the admission of evidence obtained by the infliction of torture, and does so whether the product is a confession by a suspect or a defendant and irrespective of where, by whom or on whose authority the torture was inflicted.

It is, I think, clear that from its very earliest days the common law of England set its face firmly against the use of torture. Its rejection of this practice was indeed hailed as a distinguishing feature of the common law, the subject of proud claims by English jurists such as Sir John Fortescue (De Laudibus Legum Angliae, c. 1460-1470, ed S.B. Chrimes, (1942), Chap 22, pp 47-53), Sir Thomas Smith (De Republica Anglorum, ed L Alston, 1906, book 2, chap 24, pp 104-107), Sir Edward Coke (Institutes of the Laws of England (1644), Part III, Chap 2, pp 34-36). Sir William Blackstone (Commentaries on the Laws of England, (1769) vol IV, chap 25, pp 320-321), and Sir James Stephen (A History of the Criminal Law of England, 1883, vol 1, p 222). That reliance was placed on sources of doubtful validity, such as chapter 39 of Magna Carta 1215 and Felton's Case as reported by Rushworth (Rushworth's Collections, vol (i), p 638) (see D. Jardine, A Reading on the Use of Torture in the Criminal Law of England Previously to the Commonwealth, 1837, pp 10-12, 60-62) did not weaken the strength of received opinion. The English rejection of torture was also the subject of admiring comment by foreign authorities such as Beccaria (An Essay on Crimes and Punishments, 1764, Chap XVI) and Voltaire (Commentary on Beccaria's Crimes and Punishments, 1766, Chap XII). This rejection was contrasted with the practice prevalent in the states of continental Europe who, seeking to discharge the strict standards of proof required by the Roman-canon models they had adopted, came routinely to rely on confessions procured by the infliction of torture: see A L Lowell, "The Judicial Use of Torture" (1897) 11 Harvard L Rev 220-233, 290-300; J Langbein, Torture and the Law of Proof: Europe and England in the Ancien Regime (1977); D. Hope, "Torture" [2004] 53 ICLQ 807 at pp 810-811. In rejecting the use of torture, whether applied to potential defendants or potential witnesses, the common law was moved by the cruelty of the practice as applied to those not convicted of crime, by the inherent unreliability of confessions or evidence so procured and by the belief that it degraded all those who lent themselves to the practice.

Despite this common law prohibition, it is clear from the historical record that torture was practised in England in the 16th and early 17th centuries. But this took place pursuant to warrants issued by the Council or the Crown, largely (but not exclusively) in relation to alleged offences against the state, in exercise of the Royal prerogative: see Jardine, op cit.; Lowell, op cit., pp 290-300). Thus the exercise of this royal prerogative power came to be an important issue in the struggle between the Crown and the parliamentary common lawyers which preceded and culminated in the English civil war. By the common lawyers torture was regarded as (in Jardine's words: op cit, pp 6 and 12) "totally repugnant to the fundamental principles of English law" and "repugnant to reason, justice, and humanity." One of the first acts of the Long Parliament in 1640 was, accordingly, to abolish the Court of Star Chamber, where torture evidence had been received, and in that year the last torture warrant in our history was issued. Half a century later, Scotland followed the English example, and in 1708, in one of the earliest enactments of the Westminster Parliament after the Act of Union in 1707, torture in Scotland was formally prohibited. The history is well summarised by Sir William Holdsworth (A History of English Law, vol 5, 3rd ed (1945), pp 194-195, footnotes omitted):

As Jardine put in (op. cit., p 13):

This condemnation is more aptly categorised as a constitutional principle than as a rule of evidence.

Since there has been no lawfully sanctioned torture in England since 1640, and the rule that unsworn statements made out of court are inadmissible in court was well-established by at latest the beginning of the 19th century (Cross & Tapper on Evidence, 10th edn (2004), p 582), there is an unsurprising paucity of English judicial authority on this subject. In Pearse v Pearse (1846) 1 De G & Sm 12, 28-29, 63 ER 950, 957, Knight Bruce V-C observed:

That was not a case involving any allegation of torture. Such an allegation was however made in R (Saifi) v Governor of Brixton Prison [2001] 1 WLR 1134 where the applicant for habeas corpus resisted extradition to India on the ground, among others, that the prosecution relied on a statement obtained by torture and since retracted. The Queen's Bench Divisional Court (Rose LJ and Newman J) accepted the magistrate's judgment that fairness did not call for exclusion of the statement, but was clear (para 60 of the judgment) that the common law and domestic statute law (section 78 of the Police and Criminal Evidence Act 1984) gave effect to the intent of article 15 of the International Convention against Torture and other Cruel, Inhuman or Degrading Treatment or Punishment 1984 (1990, Cm 1775), "the Torture Convention", to which more detailed reference is made below.

The appellants relied, by way of partial analogy, on the familiar principle that evidence may not be given by a prosecutor in English criminal proceedings of a confession made by a defendant, if it is challenged, unless the prosecution proves beyond reasonable doubt that the confession had not been obtained by oppression of the person who made it or in consequence of anything said or done which was likely, in the circumstances existing at the time, to render unreliable any confession which might be made by him in consequence thereof. This rule is now found in section 76 of the Police and Criminal Evidence Act 1984, but enacts a rule established at common law and expressed in such decisions as Ibrahim v The King [1914] AC 599, 609-610, R v Harz and Power [1967] AC 760, 817, and Lam Chi-ming v The Queen [1991] 2 AC 212, 220.

Plainly this rule provides an inexact analogy with evidence obtained by torture. It applies only to confessions by defendants, and it provides for exclusion on grounds very much wider than torture, or even inhuman or degrading treatment. But it is in my opinion of significance that the common law (despite suggestions to that effect by Parke B and Lord Campbell CJ in R v Baldry (1852) 2 Den 430, 445, 446-447, 169 ER 568, 574, 575, and by the Privy Council, in judgments delivered by Lord Sumner, in Ibrahim v The King [1914] AC 599, 610 and Lord Hailsham of St Marylebone in Director of Public Prosecutions v Ping Lin [1976] AC 574, 599-600) has refused to accept that oppression or inducement should go to the weight rather than the admissibility of the confession. The common law has insisted on an exclusionary rule. See, for a clear affirmation of the rule, Wong Kam-ming v The Queen [1980] AC 247.

In R v Warickshall (1783) 1 Leach 263, 168 ER 234, this rule was justified on the ground that involuntary statements are inherently unreliable. That justification is, however, inconsistent with the principle which the case established, that while an involuntary statement is inadmissible real evidence which comes to light as a result of such a statement is not. Two points are noteworthy. First, there can ordinarily be no surer proof of the reliability of an involuntary statement than the finding of real evidence as a direct result of it, as was so in Warickshall's case itself, but that has never been treated as undermining the rule. Secondly, there is an obvious anomaly in treating an involuntary statement as inadmissible while treating as admissible evidence which would never have come to light but for the involuntary statement. But this is an anomaly which the English common law has accepted, no doubt regarding it as a pragmatic compromise between the rejection of the involuntary statement and the practical desirability of relying on probative evidence which can be adduced without the need to rely on the involuntary statement.

Later decisions make clear that while the inherent unreliability of involuntary statements is one of the reasons for holding them to be inadmissible there are other compelling reasons also. In Lam Chi-ming v The Queen [1991] 2 AC 212, 220, in a judgment delivered by Lord Griffiths, the Privy Council summarised the rationale of the exclusionary rule:

Lord Griffiths described the inadmissibility of a confession not proved to be voluntary as perhaps the most fundamental rule of the English criminal law. The rationale explained by Lord Griffiths was recently endorsed by the House in R v Mushtaq [2005] UKHL 25, [2005] 1 WLR 1513, paras 1, 7, 27, 45-46, 71. It is of course true, as counsel for the Secretary of State points out, that in cases such as these the attention of the court was directed to the behaviour of the police in the jurisdiction where the defendant was questioned and the trial was held. This was almost inevitably so. But it is noteworthy that in jurisdictions where the law is in general harmony with the English common law reliability has not been treated as the sole test of admissibility in this context. In Rochin v California 342 US 165 (1952) Frankfurter J, giving the opinion of the United States Supreme Court, held that a conviction had been obtained by "conduct that shocks the conscience" (p 172) and referred to a "general principle" that "States in their prosecutions respect certain decencies of civilized conduct" (p 173). He had earlier (p 169) referred to authority on the due process clause of the United States constitution which called for judgment whether proceedings "offend those canons of decency and fairness which express the notions of justice of English-speaking peoples even toward those charged with the most heinous offenses." In The People (Attorney General) v O'Brien [1965] IR 142, 150, the Supreme Court of Ireland held, per Kingsmill Moore J, that "to countenance the use of evidence extracted or discovered by gross personal violence would, in my opinion, involve the State in moral defilement." The High Court of Australia, speaking of a discretion to exclude evidence, observed (per Barwick CJ in R v Ireland (1970) 126 CLR 321, 335), that "Convictions obtained by the aid of unlawful or unfair acts may be obtained at too high a price." In R v Oickle [2000] 2 SCR 3, a large majority of the Supreme Court of Canada cited with approval (para 66) an observation of Lamer J that "What should be repressed vigorously is conduct on [the authorities'] part that shocks the community" and considered (para 69) that while the doctrines of oppression and inducements were primarily concerned with reliability, the confessions rule also extended to protect a broader concept of voluntariness that focused on the protection of the accused's rights and fairness in the criminal process.

The appellants submit, in reliance on common law principles, that the obtaining of evidence by the infliction of torture is so grave a breach of international law, human rights and the rule of law that any court degrades itself and the administration of justice by admitting it. If, therefore, it appears that a confession or evidence may have been procured by torture, the court must exercise its discretion to reject such evidence as an abuse of its process.

In support of this contention the appellants rely on four recent English authorities. The first of these is R v Horseferry Road Magistrates' Court, Ex p Bennett [1994] 1 AC 42. This case was decided on the factual premise that the applicant had been abducted from South Africa and brought to this country in gross breach of his rights and the law of South Africa, at the behest of the British authorities, to stand trial here, and on the legal premise that a fair trial could be held. The issue, accordingly, was whether the unlawful abduction of the applicant was an abuse of the court's process to which it should respond by staying the prosecution. The House held, by a majority, that it was. The principle laid down most clearly appears in the opinion of Lord Griffiths at pp 61-62:

Counsel for the Secretary of State points out that the members of the majority attached particular significance to the involvement of the British authorities in the unlawful conduct complained of, and this is certainly so: see the opinion of Lord Griffiths at p 62F, Lord Bridge of Harwich at pp 64G and 67G and Lord Lowry at pp 73G, 76F and 77D. But the appellants point to the germ of a wider principle. Thus Lord Lowry (p 74G) understood the court's discretion to stay proceedings as an abuse of process to be exercisable where either a fair trial is impossible or "it offends the court's sense of justice and propriety to be asked to try the accused in the circumstances of a particular case." He opined (p 76C):

Lord Lowry's opinion did not earn the concurrence of any other member of the House, but the appellants contend that this wider principle is applicable in the extreme case of evidence procured by torture. In United States v Toscanino 500 F 2d 267 (1974) the US Court of Appeals reached a decision very similar to Bennett.

In R v Latif [1996] 1 WLR 104 the executive misconduct complained of was much less gross than in Bennett, and the outcome was different. Speaking for the House, Lord Steyn (at pp 112-113) acknowledged a judicial discretion to stay proceedings as an abuse if they would "amount to an affront to the public conscience" and where "it would be contrary to the public interest in the integrity of the criminal justice system that a trial should take place." In that case the conduct complained of was not so unworthy or shameful that it was an affront to the public conscience to allow the prosecution to proceed.

The premises of the Court of Appeal's decision in R v Mullen [2000] QB 520 were similar to those in Bennett, save that a fair trial had already taken place and Mullen had already been convicted of very serious terrorist offences, and sentenced to 30 years' imprisonment, before he was alerted to the misconduct surrounding his abduction from Zimbabwe. Despite the fairness of the trial, his conviction was quashed. Giving the reserved judgment of the court, Rose LJ said (at pp 535-536):

The fourth authority relied on for its statements of principle was R v Looseley, Attorney General's Reference (No 3 of 2000) [2001] UKHL 53, [2001] 1 WLR 2060, which concerned cases of alleged entrapment. At the outset of his opinion (para 1) my noble and learned friend Lord Nicholls of Birkenhead declared that:

A stay is granted in a case of entrapment not to discipline the police (para 17) but because it is improper for there to be a prosecution at all for the relevant offence, having regard to the state's involvement in the circumstances in which it was committed. To prosecute in a case where the state has procured the commission of the crime is (para 19) "unacceptable and improper" and "an affront to the public conscience." Such a prosecution would not be fair in the broad sense of the word. My noble and learned friend Lord Hoffmann, having referred to Canadian authority and to Bennett, accepted Lord Griffiths' description of the power to stay in the case of behaviour which threatened basic human rights or the rule of law as (para 40) "a jurisdiction to prevent abuse of executive power".

If, contrary to their submission (and to the opinion of the Divisional Court in R (Saifi) v Governor of Brixton Prison: see para 13 above) the common law and section 78 of the 1984 Act are not, without more, enough to require rejection of evidence which has or may have been procured by torture, whether or not with the complicity of the British authorities, the appellants submit that the European Convention compels that conclusion.

It is plain that SIAC (and, for that matter, the Secretary of State) is a public authority within the meaning of section 6 of the Human Rights Act 1998 and so forbidden to act incompatibly with a Convention right. One such right, guaranteed by article 3, is not to be subjected to torture or to inhuman or degrading treatment. This absolute, non-derogable prohibition has been said (Soering v United Kingdom (1989) 11 EHRR 439, para 88) to enshrine "one of the fundamental values of the democratic societies making up the Council of Europe". The European Court has used such language on many occasions (Aydin v Turkey (1997) 25 EHRR 251, para 81).

Article 6 of the Convention guarantees the right to a fair trial. Different views have in the past been expressed on whether, for purposes of article 6, the proceedings before SIAC are to be regarded as civil or criminal. Rather than pursue this debate the parties are agreed that the appellants' challenge to their detention pursuant to the Secretary of State's certification in any event falls within article 5(4). That provision entitles anyone deprived of his liberty by arrest or detention to take proceedings by which the lawfulness of his detention shall be decided speedily by a court and his release ordered if the detention is not lawful. It is well-established that such proceedings must satisfy the basic requirements of a fair trial: Garcia Alva v Germany (2001) 37 EHRR 335; R (West) v Parole Board, R (Smith) v Parole Board (No 2) [2005] UKHL 1, [2005] 1 WLR 350. Sensibly, therefore, the parties are agreed that the applicability of article 6 should be left open and the issue resolved on the premise that article 5(4) applies.

The Secretary of State submits that under the Convention the admissibility of evidence is a matter left to be decided under national law; that under the relevant national law, namely, the 2001 Act and the Rules, the evidence which the Secretary of State seeks to adduce is admissible before SIAC; and that accordingly the admission of this evidence cannot be said to undermine the fairness of the proceedings. I shall consider the effect of the statutory scheme in more detail below. The first of these propositions is, however, only half true. It is correct that the European Court of Human Rights has consistently declined to articulate evidential rules to be applied in all member states and has preferred to leave such rules to be governed by national law: see, for example, Schenk v Switzerland (1988) 13 EHRR 242, para 46; Ferrantelli and Santangelo v Italy (1996) 23 EHRR 288, para 48; Khan v United Kingdom (2000) 31 EHRR 1016, para 34. It has done so even where, as in Khan, evidence was acknowledged to have been obtained unlawfully and in breach of another article of the Convention. But in these cases and others the court has also insisted on its responsibility to ensure that the proceedings, viewed overall on the particular facts, have been fair, and it has recognised that the way in which evidence has been obtained or used may be such as to render the proceedings unfair. Such was its conclusion in Saunders v United Kingdom (1996) 23 EHRR 313, a case of compulsory questioning, and in Teixeira de Castro v Portugal (1998) 28 EHRR 101, para 39, a case of entrapment. A similar view would have been taken by the Commission in the much earlier case of Austria v Italy (1963) 6 YB 740, 784, had it concluded that the victims whom Austria represented had been subjected to maltreatment with the aim of extracting confessions. But the Commission observed that article 6(2) could only be regarded as being violated if the court subsequently accepted as evidence any admissions extorted in this manner. This was a point made by my noble and learned friend Lord Hoffmann in the much more recent devolution case of Montgomery v H M Advocate, Coulter v H M Advocate [2003] 1 AC 641, 649, when he observed:

Lord Hoffmann, in R v Governor of Brixton Prison, Ex p Levin [1997] AC 741, 748, did not exclude the possibility (he did not have to decide) that evidence might be rejected in extradition proceedings if, though technically admissible, it had been obtained in a way which outraged civilised values. Such was said to be the case in R (Ramda) v Secretary of State for the Home Department [2002] EWHC 1278 (Admin), unreported, 27 June 2002, where the applicant resisted extradition to France on the ground that the evidence which would be relied on against him at trial had been obtained by torture and that he would be unable to resist its admission. The Queen's Bench Divisional Court concluded (para 22) that if these points were made out, his trial would not be fair and the Secretary of State would be effectively bound to refuse to extradite him. In the very recent case of Mamatkulov and Askarov v Turkey (App Nos 46827/99 and 46951/99, unreported, 4 February 2005) Judges Bratza, Bonello and Hedigan delivered a joint partly dissenting opinion, in the course of which they held in paras 15-17:

The approach of these judges is consistent with the even more recent decision of the Court in Harutyunyan v Armenia (App No 36549/03, unreported, 5 July 2005) where in paras 2(b) and (f) the Court ruled:

Had the Court found that the complaints of coercion and torture appeared to be substantiated, a finding that article 6(1) had been violated would, in my opinion, have been inevitable. As it was, the Court did not rule that these complaints were inadmissible. Nor did it dismiss them. It adjourned examination of the applicant's complaints concerning the alleged violation of his right to silence and the admission in court of evidence obtained under torture.

The appellants' submission has a further, more international, dimension. They accept, as they must, that a treaty, even if ratified by the United Kingdom, has no binding force in the domestic law of this country unless it is given effect by statute or expresses principles of customary international law: J H Rayner (Mincing Lane) Ltd v Department of Trade and Industry [1990] 2 AC 418; R v Secretary of State for the Home Department, Ex p Brind [1991] 1 AC 696; R v Lyons [2002] UKHL 44, [2003] 1 AC 976. But they rely on the well-established principle that the words of a United Kingdom statute, passed after the date of a treaty and dealing with the same subject matter, are to be construed, if they are reasonably capable of bearing such a meaning, as intended to carry out the treaty obligation and not to be inconsistent with it: Garland v British Rail Engineering Ltd [1983] 2 AC 751, 771. The courts are obliged under section 2 of the 1998 Act to take Strasbourg jurisprudence into account in connection with a Convention right, their obligation under section 3 is to interpret and give effect to primary and subordinate legislation in a way which is compatible with Convention rights so far as possible to do so and it is their duty under section 6 not to act incompatibly with a Convention right. If, and to the extent that, development of the common law is called for, such development should ordinarily be in harmony with the United Kingdom's international obligations and not antithetical to them. I do not understand these principles to be contentious.

The appellants' argument may, I think, be fairly summarised as involving the following steps:

(1) The European Convention is not to be interpreted in a vacuum, but taking account of other international obligations to which member states are subject, as the European Court has in practice done.

(2) The prohibition of torture enjoys the highest normative force recognised by international law.

(3) The international prohibition of torture requires states not merely to refrain from authorising or conniving at torture but also to suppress and discourage the practice of torture and not to condone it.

(4) Article 15 of the Torture Convention requires the exclusion of statements made as a result of torture as evidence in any proceedings.

(5) Court decisions in many countries have given effect directly or indirectly to article 15 of the Torture Convention.

(6) The rationale of the exclusionary rule in article 15 is found not only in the general unreliability of evidence procured by torture but also in its offensiveness to civilised values and its degrading effect on the administration of justice.

(7) Measures directed to counter the grave dangers of international terrorism may not be permitted to undermine the international prohibition of torture.

It is necessary to examine these propositions in a little detail.

(1) Interpretation of the Convention in a wider international context.

Article 31 of the Vienna Convention on the Law of Treaties, reflecting principles of customary international law, provides in article 31(3)(c) that in interpreting a treaty there shall be taken into account, together with the context, any relevant rules of international law applicable in the relations between the parties. The European Court has recognised this principle (Golder v United Kingdom (1975) 1 EHRR 524, para 29, HN v Poland (Application No 77710/01, 13 September 2005, unreported, para 75)), and in Al-Adsani v United Kingdom (2001) 34 EHRR 273, para 55, it said (footnotes omitted):

The Court has in its decisions invoked a wide range of international instruments, including the United Nations Convention on the Rights of the Child 1989 and the Beijing Rules (V v United Kingdom (1999) 30 EHRR 121, paras 76-77), the Council of Europe Standard Minimum Rules for the Treatment of Prisoners (S v Switzerland (1991) 14 EHRR 670, para 48) and the 1975 Declaration referred to in para 31 below (Ireland v United Kingdom (1978) 2 EHRR 25, para 167). More pertinently to these appeals, the Court has repeatedly invoked the provisions of the Torture Convention: see, for example, Aydin v Turkey (1997) 25 EHRR 251, para 103; Selmouni v France (1999) 29 EHRR 403, para 97. In Soering v United Kingdom (1989) 11 EHRR 439, para 88, the Court said (footnotes omitted):

(2) The international prohibition of torture.

The preamble to the United Nations Charter (1945) recorded the determination of member states to reaffirm their faith in fundamental human rights and the dignity and worth of the human person and to establish conditions under which justice and respect for the obligations arising from treaties and other sources of international law can be maintained. The Charter was succeeded by the Universal Declaration of Human Rights 1948, the European Convention 1950 and the International Covenant on Civil and Political Rights 1966, all of which (in articles 5, 3 and 7 respectively, in very similar language) provided that no one should be subjected to torture or inhuman or degrading treatment.

On 9 December 1975 the General Assembly of the United Nations, without a vote, adopted Resolution 3452 (XXX), a Declaration on the Protection of All Persons from Being Subjected to Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment. This included (in article 1) a definition of torture as follows:

Articles 2-4 provided as follows:

Action was then taken to prepare a convention. This action culminated in the Torture Convention, which came into force on 26 June 1987. All member states of the Council of Europe are members with the exception of Moldova, Andorra and San Marino, the last two of which have been signed but not yet ratified.

The Torture Convention contained, in article 1, a definition of torture:

1. For the purposes of this Convention, 'torture' means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in or incidental to lawful sanctions.

2. This article is without prejudice to any international instrument or national legislation which does or may contain provisions of wider application."

It is noteworthy that the torture must be inflicted by or with the complicity of an official, must be intentional, and covers treatment inflicted for the purpose of obtaining information or a confession. Articles 2, 3 and 4 provide:

1. Each State Party shall take effective legislative, administrative, judicial or other measures to prevent acts of torture in any territory under its jurisdiction.

2. No exceptional circumstances whatsoever, whether a state of war or a threat of war, internal political instability or any other public emergency, may be invoked as a justification of torture.

3. An order from a superior officer or a public authority may not be invoked as a justification of torture.

1. No State Party shall expel, return ('refouler') or extradite a person to another State where there are substantial grounds for believing that he would be in danger of being subjected to torture.

2. For the purpose of determining whether there are such grounds, the competent authorities shall take into account all relevant considerations including, where applicable, the existence in the State concerned of a consistent pattern of gross, flagrant or mass violations of human rights.

1. Each State Party shall ensure that all acts of torture are offences under its criminal law. The same shall apply to an attempt to commit torture and to an act by any person which constitutes complicity or participation in torture.

2. Each State Party shall make these offences punishable by appropriate penalties which take into account their grave nature."

It is common ground in these proceedings that the international prohibition of the use of torture enjoys the enhanced status of a jus cogens or peremptory norm of general international law. For purposes of the Vienna Convention, a peremptory norm of general international law is defined in article 53 to mean "a norm accepted and recognized by the international community of States as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character". In R v Bow Street Metropolitan Stipendiary Magistrate, Ex p Pinochet Ugarte (No 3) [2000] 1 AC 147, 197-199, the jus cogens nature of the international crime of torture, the subject of universal jurisdiction, was recognised. The implications of this finding were fully and authoritatively explained by the International Criminal Tribunal for the Former Yugoslavia in Prosecutor v Furundzija [1998] ICTY 3, 10 December 1998 in a passage which, despite its length, calls for citation (footnotes omitted):

There can be few issues on which international legal opinion is more clear than on the condemnation of torture. Offenders have been recognised as the "common enemies of mankind" (Demjanjuk v Petrovsky 612 F Supp 544 (1985), 566, Lord Cooke of Thorndon has described the right not to be subjected to inhuman treatment as a "right inherent in the concept of civilisation" (Higgs v Minister of National Security [2000] 2 AC 228, 260), the Ninth Circuit Court of Appeals has described the right to be free from torture as "fundamental and universal" (Siderman de Blake v Argentina 965 F 2d 699 (1992), 717) and the UN Special Rapporteur on Torture (Mr Peter Koojimans) has said that "If ever a phenomenon was outlawed unreservedly and unequivocally it is torture" (Report of the Special Rapporteur on Torture, E/CN.4/1986/15, para 3).

(3) The duty of states in relation to torture.

As appears from the passage just cited, the jus cogens erga omnes nature of the prohibition of torture requires member states to do more than eschew the practice of torture. In Kuwait Airways Corporation v Iraqi Airways Co (Nos 4 and 5) [2002] UKHL 19, [2002] 2 AC 883, paras 29, 117, the House refused recognition to conduct which represented a serious breach of international law. This was, as I respectfully think, a proper response to the requirements of international law. In General Comment 20 (1992) on article 7 of the ICCPR, the UN Human Rights Committee said, in para 8:

Article 41 of the International Law Commission's draft articles on Responsibility of States for internationally wrongful acts (November 2001) requires states to cooperate to bring to an end through lawful means any serious breach of an obligation under a peremptory norm of general international law. An advisory opinion of the International Court of Justice on the Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (9 July 2004, General List No 131), para 159 explained the consequences of the breach found in that case:

There is reason to regard it as a duty of states, save perhaps in limited and exceptional circumstances, as where immediately necessary to protect a person from unlawful violence or property from destruction, to reject the fruits of torture inflicted in breach of international law. As McNally JA put it in S v Nkomo 1989 (3) ZLR 117, 131:

(4) Article 15 of the Torture Convention.

Article 12 of the 1975 Declaration provided:

Article 15 of the Torture Convention repeats the substance of this provision, subject to a qualification:

The additional qualification makes plain the blanket nature of this exclusionary rule. It cannot possibly be read, as counsel for the Secretary of State submits, as intended to apply only in criminal proceedings. Nor can it be understood to differentiate between confessions and accusatory statements, or to apply only where the state in whose jurisdiction the proceedings are held has inflicted or been complicit in the torture. It would indeed be remarkable if national courts, exercising universal jurisdiction, could try a foreign torturer for acts of torture committed abroad, but could nonetheless receive evidence obtained by such torture. The matter was succinctly put in the Report by Mr Alvaro Gil-Robles, the Council of Europe Commissioner for Human Rights, in his Report on his visit to the United Kingdom in November 2004 (8 June 2005, Comm DH (2005)6):

(5) State practice.

A Committee against Torture was established under article 17 of the Torture Convention to monitor compliance by member states. The Committee has recognised a duty of states, if allegations of torture are made, to investigate them: PE v France, 19 December 2002, CAT/C/29/D/193/2001, paras 5.3, 6.3; GK v Switzerland, 12 May 2003, CAT/C/30/D/219/2002), para 6.10. The clear implication is that the evidence should have been excluded had the complaint been verified.

In Canada, article 15 of the Torture Convention has been embodied in the criminal code: see India v Singh 108 CCC (3d) 274 (1996), para 20. In France, article 15 has legal effect (French Republic v Haramboure, Cour de Cassation, Chambre Criminelle, 24 January 1995, No. de pourvoi 94-81254), and extradition to Spain was refused where allegations that a witness statement had been procured by torture in Spain was judged not to have been adequately answered (Le Ministère Public v Irastorza Dorronsoro, Cour d'Appel de Pau, No 238/2003, 16 May 2003). In the Netherlands, it was held by the Supreme Court to follow from article 3 of the European Convention and article 7 of the ICCPR that if witness statements had been obtained by torture they could not be used as evidence: Pereira, 1 October 1996, nr 103.094, para 6.2. In Germany, as in France, article 15 has legal effect: El Motassadeq, decision of the Higher Regional Court of Hamburg, 14 June 2005, para 2.

In the United States, torture was recognised to be prohibited by the law of nations even before the Torture Convention was made: Filartiga v Peña-Irala 630 F 2d 876 (1980). Earlier still, it had been said to be

(6) The rationale of the exclusionary rule.

In their work on The United Nations Convention against Torture (1988), p 148, Burgers and Danelius suggest that article 15 of the Torture Convention is based on two principles:

It seems indeed very likely that the unreliability of a statement or confession procured by torture and a desire to discourage torture by devaluing its product are two strong reasons why the rule was adopted. But it also seems likely that the article reflects the wider principle expressed in article 69(7) of the Rome Statute of the International Criminal Court, which has its counterpart in the Rules of Procedure and Evidence of the International Criminal Tribunals for the Former Yugoslavia and Rwanda:

The appellants contend that admission as evidence against a party to legal proceedings of a confession or an accusatory statement obtained by inflicting treatment of the severity necessary to fall within article 1 of the Torture Convention will "shock the community", infringe that party's rights and the fairness of the proceedings (R v Oickle: see para 17 above), shock the judicial conscience (United States v Hensel 509 F Supp 1364 (1981), p 1372), abuse or degrade the proceedings (United States v Toscanino 500 F 2d 267 (1974), p 276), and involve the state in moral defilement (The People (Attorney General) v O'Brien: see para 17 above).

(7) The impact of terrorism

The European Court has emphasised that article 3 of the European Convention is an absolute prohibition, not derogable in any circumstances. In Chahal v United Kingdom (1996) 23 EHRR 413, para 79, it ruled:

That the Torture Convention, including article 15, enjoys the same absolute quality is plain from the text of article 2, quoted in para 32 above.

It is true, as the Secretary of State submits, that States Members of the United Nations and the Council of Europe have been strongly urged since 11 September 2001 to cooperate and share information in order to counter the cruel and destructive evil of terrorism. But these calls have been coupled with reminders that human rights, and international and humanitarian law, must not be infringed or compromised. Thus, while the Council of Europe's Parliamentary Assembly recommendation 1534 of 26 September 2001 refers to co-operation "on the basis of the Council of Europe's values and legal instruments", it also refers to Parliamentary Assembly Resolution 1258, para 7 of which states:

The Council of Europe Convention on the Prevention of Terrorism of 16 May 2005, recalling in its preamble

Other similar examples could be given.

The United Nations pronouncements are to the same effect. Thus Security Council resolution 1373 of 28 September 2001 called for co-operation and exchange of information to prevent terrorist acts, but also reaffirmed resolution 1269 of 19 October 1999 which called for observance of the principles of the UN Charter and the norms of international law, including international humanitarian law. By Security Council resolution 1566 of 8 October 2004 states were reminded

Again, other similar examples could be given. The General Assembly has repeatedly made the same point: see, for example, resolution 49/60 of 9 December 1994; resolution 51/210 of 17 December 1996; and resolution 59/290 of 13 April 2005. The Secretary General of the UN echoed the same theme in statements of 4 October 2002, 6 March 2003 and 10 March 2005.

The events of 11 September prompted the Committee against Torture to issue a statement on 22 November 2001 (CAT/C/XXVII/Misc 7) in which it said:

A statement to similar effect was made by the Committee against Torture, the Special Rapporteur on Torture, the Chairperson of the 22nd session of the Board of Trustees of the United Nations Voluntary Fund for Victims of Torture and the Acting United Nations Commissioner for Human Rights on 26 June 2004 (CAT Report to the General Assembly, A/59/44 (2004), para 17). In its Conclusions and Recommendations on the United Kingdom dated 10 December 2004 (CAT/C/CR/33/3), having received the United Kingdom's fourth periodic report, the Committee welcomed the Secretary of State's indication that he did not intend to rely upon or present evidence where there is a knowledge or belief that torture has taken place but recommended that this be appropriately reflected in formal fashion, such as legislative incorporation or undertaking to Parliament, and that means be provided whereby an individual could challenge the legality of any evidence plausibly suspected of having been obtained by torture in any proceeding.

This recommendation followed the judgment of the Court of Appeal in these appeals. Concern at the effect of that judgment was also expressed by the International Commission of Jurists on 28 August 2004, which declared that "Evidence obtained by torture, or other means which constitute a serious violation of human rights against a defendant or third party, is never admissible and cannot be relied on in any proceedings," and by the Council of Europe Commissioner for Human Rights, Mr Gil-Robles in his Report cited in para 35 above. In a Report of 9 June 2005 on a visit made to the United Kingdom in March 2004, the Council of Europe's Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (CPT/Inf (2005) 10), para 31, observed:

In Resolution 1433, adopted on 26 April 2005, on the Lawfulness of Detentions by the United States in Guantanamo Bay, the Parliamentary Assembly of the Council of Europe called on the United States to cease the practice of rendition and called on member states to respect their obligation under article 15 of the Torture Convention.

The House has not been referred to any decision, resolution, agreement or advisory opinion suggesting that a confession or statement obtained by torture is admissible in legal proceedings if the torture was inflicted without the participation of the state in whose jurisdiction the proceedings are held, or that such evidence is admissible in proceedings related to terrorism.

While counsel for the Secretary of State questions the effect and applicability of some of the material on which the appellants rely, he founds his case above all on the statutory scheme established by Part 4 of the 2001 Act. He builds on the appellants' acceptance that the Secretary of State may, when forming the reasonable belief and suspicion required for certification under section 21, and when acting on that belief to arrest, search and detain a suspect, act on information which has or may have been obtained by torture inflicted in a foreign country without British complicity. That acceptance, he submits, supports the important and practical need for the security services and the Secretary of State to obtain intelligence and evidence from foreign official sources, some of which (in the less progressive countries) might dry up if their means of obtaining intelligence and evidence were the subject of intrusive enquiry. But it would create a mismatch which Parliament could not have intended if the Secretary of State were able to rely on material at the certification stage which SIAC could not later receive. It would, moreover, emasculate the statutory scheme, which is specifically designed to enable SIAC, constituted as it is, to see all relevant material, even such ordinarily inadmissible material as may be obtained on warranted intercepts. This is reflected in rule 44(3) of the applicable Rules, which dispenses with all rules of evidence, including any that might otherwise preclude admission of evidence obtained by torture in the circumstances postulated. This is not a negligible argument, and a majority of the Court of Appeal broadly accepted it. There are, however, in my opinion, a number of reasons why it must be rejected.

I am prepared to accept (although I understand the interveners represented by Mr Starmer QC not to do so) that the Secretary of State does not act unlawfully if he certifies, arrests, searches and detains on the strength of what I shall for convenience call foreign torture evidence. But by the same token it is, in my view, questionable whether he would act unlawfully if he based similar action on intelligence obtained by officially-authorised British torture. If under such torture a man revealed the whereabouts of a bomb in the Houses of Parliament, the authorities could remove the bomb and, if possible, arrest the terrorist who planted it. There would be a flagrant breach of article 3 for which the United Kingdom would be answerable, but no breach of article 5(4) or 6. Yet the Secretary of State accepts that such evidence would be inadmissible before SIAC. This suggests that there is no correspondence between the material on which the Secretary of State may act and that which is admissible in legal proceedings.

This is not an unusual position. It arises whenever the Secretary of State (or any other public official) relies on information which the rules of public interest immunity prevent him adducing in evidence: Makanjuola v Commissioner of Police of the Metropolis [1992] 3 All ER 617, 623 e to j; R v Chief Constable of West Midlands Police, Ex p Wiley [1995] 1 AC 274, 295F-297C. It is a situation which arises where action is based on a warranted interception and there is no dispensation which permits evidence to be given. This may be seen as an anomaly, but (like the anomaly to which the rule in R v Warickshall gives rise) it springs from the tension between practical common sense and the need to protect the individual against unfair incrimination. The common law is not intolerant of anomaly.

There would be a much greater anomaly if the duty of SIAC, hearing an appeal under section 25, were to decide whether the Secretary of State had entertained a reasonable belief and suspicion at the time of certification. But, as noted above in para 5, SIAC's duty is to cancel the certificate if it considers that there "are" no reasonable grounds for a belief or suspicion of the kind referred to. This plainly refers to the date of the hearing. The material may by then be different from that on which the Secretary of State relied. He may have gathered new and better information; or some of the material on which he had relied may have been discredited; or he may have withdrawn material which he was ordered but was unwilling to disclose. SIAC must act on the information lawfully before it to decide whether there are reasonable grounds at the time of its decision.

I am not impressed by the argument based on the practical undesirability of upsetting foreign regimes which may resort to torture. On the approach of the Court of Appeal majority, third party torture evidence, although legally admissible, must be assessed by SIAC in order to decide what, if any, weight should be given to it. This is an exercise which could scarcely be carried out without investigating whether the evidence had been obtained by torture, and, if so, when, by whom, in what circumstances and for what purpose. Such an investigation would almost inevitably call for an approach to the regime which is said to have carried out the torture.

The Secretary of State is right to submit that SIAC is a body designed to enable it to receive and assess a wide range of material, including material which would not be disclosed to a body lacking its special characteristics. And it would of course be within the power of a sovereign Parliament (in breach of international law) to confer power on SIAC to receive third party torture evidence. But the English common law has regarded torture and its fruits with abhorrence for over 500 years, and that abhorrence is now shared by over 140 countries which have acceded to the Torture Convention. I am startled, even a little dismayed, at the suggestion (and the acceptance by the Court of Appeal majority) that this deeply-rooted tradition and an international obligation solemnly and explicitly undertaken can be overridden by a statute and a procedural rule which make no mention of torture at all. Counsel for the Secretary of State acknowledges that during the discussions on Part 4 the subject of torture was never the subject of any thought or any allusion. The matter is governed by the principle of legality very clearly explained by my noble and learned friend Lord Hoffmann in R v Secretary of State for the Home Department, Ex p Simms [2000] 2 AC 115, 131:

It trivialises the issue before the House to treat it as an argument about the law of evidence. The issue is one of constitutional principle, whether evidence obtained by torturing another human being may lawfully be admitted against a party to proceedings in a British court, irrespective of where, or by whom, or on whose authority the torture was inflicted. To that question I would give a very clear negative answer.

I accept the broad thrust of the appellants' argument on the common law. The principles of the common law, standing alone, in my opinion compel the exclusion of third party torture evidence as unreliable, unfair, offensive to ordinary standards of humanity and decency and incompatible with the principles which should animate a tribunal seeking to administer justice. But the principles of the common law do not stand alone. Effect must be given to the European Convention, which itself takes account of the all but universal consensus embodied in the Torture Convention. The answer to the central question posed at the outset of this opinion is to be found not in a governmental policy, which may change, but in law.

The appellants broaden their argument to contend that all the principles on which they rely apply to inhuman and degrading treatment, if inflicted by an official with the requisite intention and effect, as to torture within the Torture Convention definition. It is, of course, true that article 3 of the European Convention (and the comparable articles of other human rights instruments) lump torture and inhuman or degrading treatment together, drawing no distinction between them. The European Court did, however, draw a distinction between them in Ireland v United Kingdom (1978) 2 EHRR 25, holding that the conduct complained of was inhuman or degrading but fell short of torture, and article 16 of the Torture Convention draws this distinction very expressly:

Ill-treatment falling short of torture may invite exclusion of evidence as adversely affecting the fairness of a proceeding under section 78 of the 1984 Act, where that section applies. But I do not think the authorities on the Torture Convention justify the assimilation of these two kinds of abusive conduct. Special rules have always been thought to apply to torture, and for the present at least must continue to do so. It would, on the other hand, be wrong to regard as immutable the standard of what amounts to torture. This is a point made by the European Court in Selmouni v France (1999) 29 EHRR 403, paras 99-101 (footnotes omitted):

It may well be that the conduct complained of in Ireland v United Kingdom, or some of the Category II or III techniques detailed in a J2 memorandum dated 11 October 2002 addressed to the Commander, Joint Task Force 170 at Guantanamo Bay, Cuba, (see The Torture Papers: The Road to Abu Ghraib, ed K Greenberg and J Dratel, (2005), pp 227-228), would now be held to fall within the definition in article 1 of the Torture Convention.

The appellants contend that it is for a party seeking to adduce evidence to establish its admissibility if this is challenged. The Secretary of State submits that it is for a party seeking to challenge the admissibility of evidence to make good the factual grounds on which he bases his challenge. He supports this approach in the present context by pointing to the reference in article 15 of the Torture Convention to a statement "which is established to have been made as a result of torture." There is accordingly said to be a burden on the appellant in the SIAC proceedings to prove the truth of his assertion.

I do not for my part think that a conventional approach to the burden of proof is appropriate in a proceeding where the appellant may not know the name or identity of the author of an adverse statement relied on against him, may not see the statement or know what the statement says, may not be able to discuss the adverse evidence with the special advocate appointed (without responsibility) to represent his interests, and may have no means of knowing what witness he should call to rebut assertions of which he is unaware. It would, on the other hand, render section 25 appeals all but unmanageable if a generalised and unsubstantiated allegation of torture were in all cases to impose a duty on the Secretary of State to prove the absence of torture. It is necessary, in this very unusual forensic setting, to devise a procedure which affords some protection to an appellant without imposing on either party a burden which he cannot ordinarily discharge.

The appellant must ordinarily, by himself or his special advocate, advance some plausible reason why evidence may have been procured by torture. This will often be done by showing that evidence has, or is likely to have, come from one of those countries widely known or believed to practise torture (although they may well be parties to the Torture Convention and will, no doubt, disavow the practice publicly). Where such a plausible reason is given, or where SIAC with its knowledge and expertise in this field knows or suspects that evidence may have come from such a country, it is for SIAC to initiate or direct such inquiry as is necessary to enable it to form a fair judgment whether the evidence has, or whether there is a real risk that it may have been, obtained by torture or not. All will depend on the facts and circumstances of a particular case. If SIAC is unable to conclude that there is not a real risk that the evidence has been obtained by torture, it should refuse to admit the evidence. Otherwise it should admit it. It should throughout be guided by recognition of the important obligations laid down in articles 3 and 5(4) of the European Convention and, through them, article 15 of the Torture Convention, and also by recognition of the procedural handicaps to which an appellant is necessarily subject in proceedings from which he and his legal representatives are excluded.

Since a majority of my noble and learned friends do not agree with the view I have expressed on this point, and since it is of practical importance, I should explain why I do not share their opinion.

I agree, of course, that the reference in article 15 to "any statement which is established to have been made as a result of torture" would ordinarily be taken to mean that the truth of such an allegation should be proved. That is what "established" ordinarily means. I would also accept that in any ordinary context the truth of the allegation should be proved by the party who makes it. But the procedural regime with which the House is concerned in this case, described in paragraphs 6-7 and 55 above, is very far from ordinary. A detainee may face the prospect of indefinite years of detention without charge or trial, and without knowing what is said against him or by whom. Lord Woolf CJ was not guilty of overstatement in describing an appellant to SIAC, if denied access to the evidence, as "undoubtedly under a grave disadvantage" (M v Secretary of State for the Home Department [2004] EWCA Civ 324, [2004] 2 All ER 863, para 13). The special advocates themselves have publicly explained the difficulties under which they labour in seeking to serve the interests of those they are appointed to represent (Constitutional Affairs Committee of the House of Commons, The operation of the Special Immigration Appeals Commission (SIAC) and the use of Special Advocates, Seventh Report of Session 2004-05, vol II, HC 323-II, Ev 1-12, 53-61).

My noble and learned friend Lord Hope proposes, in paragraph 121 of his opinion, the following test: is it established, by means of such diligent enquiries into the sources that it is practicable to carry out and on a balance of probabilities, that the information relied on by the Secretary of State was obtained under torture? This is a test which, in the real world, can never be satisfied. The foreign torturer does not boast of his trade. The security services, as the Secretary of State has made clear, do not wish to imperil their relations with regimes where torture is practised. The special advocates have no means or resources to investigate. The detainee is in the dark. It is inconsistent with the most rudimentary notions of fairness to blindfold a man and then impose a standard which only the sighted could hope to meet. The result will be that, despite the universal abhorrence expressed for torture and its fruits, evidence procured by torture will be laid before SIAC because its source will not have been "established".

The authorities relied on by my noble and learned friends Lord Hope of Craighead and Lord Rodger of Earlsferry to support their conclusion are of questionable value at most. In El Motassadeq, a decision of the Higher Regional Court of Hamburg of 14 June 2005, the United States Department of Justice supplied the German court, for purposes of a terrorist trial proceeding in Germany with reference to the events of 11 September 2001, with summaries of statements made by three Arab men. There was material suggesting that the statements had been obtained by torture, and the German court sought information on the whereabouts of the witnesses and the circumstances of their examination. The whereabouts of two of the witnesses had been kept secret for several years, but it was believed the American authorities had access to them. The American authorities supplied no information, and said they were not in a position to give any indications as to the circumstances of the examination of these persons. Two American witnesses who attended to give evidence took the same position. One might have supposed that the summaries would, without more, have been excluded. But the German court, although noting that it was the United States, whose agents were accused of torture, which was denying information to the court, proceeded to examine the summaries and found it possible to infer from internal evidence that torture had not been used. This is not a precedent which I would wish to follow. But at least the defendant knew what the evidence was.

In Mamatkulov and Askarov v Turkey (Application Nos 46827/99 and 46951/99, unreported, 4 February 2005) the applicants had resisted an application by the Republic of Uzbekistan to extradite them from Turkey to stand trial on very serious charges in Uzbekistan. They resisted extradition on the ground, among others, that if returned to Uzbekistan they would be tortured. There was material to show that that was not a fanciful fear. On application made by them to the European Court of Human Rights, it indicated to Turkey under rule 39 of its procedural rules that the extradition should not take place until it had had an opportunity to examine the validity of the applicants' fears. But in breach of this measure, and in violation of article 34 of the Convention, Turkey surrendered the applicants. The Chamber found, in effect, that no findings of fact could be made since the applicants had been denied an opportunity to have inquiries made to obtain evidence in support of their allegations: paragraph 57 of the judgment. The approach of the Grand Chamber appears from paragraphs 68 and 69 of its judgment:

Despite a compelling dissent, from which I have quoted in paragraph 26 above, the Grand Chamber concluded that Turkey had not violated article 3 of the Convention in surrendering the applicants. It did so in reliance on assurances received by Turkey from the Uzbek Government and the Uzbek Public Prosecutor before and after the surrender, and medical reports by doctors at the Uzbek prison where the applicants were being held. These matters were not sufficient to allay the concerns of the minority, and understandably, since Turkey's unlawful conduct prevented the European Court examining the case as it would have wished. But the applicants were able to participate fully in the proceedings in Turkey and were not denied knowledge of the case against them.

I regret that the House should lend its authority to a test which will undermine the practical efficacy of the Torture Convention and deny detainees the standard of fairness to which they are entitled under article 5(4) or 6(1) of the European Convention. The matter could not be more clearly put than by my noble and learned friend Lord Nicholls of Birkenhead in the closing paragraph of his opinion.

The Court of Appeal were unable to conclude that there was no plausible suspicion of torture in these cases. I would accordingly allow the appeals, set aside the orders made by SIAC and the Court of Appeal, and remit all the cases to SIAC for reconsideration in the light of the opinions of the House.

LORD NICHOLLS OF BIRKENHEAD

My Lords,

Torture is not acceptable. This is a bedrock moral principle in this country. For centuries the common law has set its face against torture. In early times this did not prevent the use of torture under warrants issued by the King or his Council. But by the middle of the 17th century this practice had ceased. In 1628 John Felton assassinated the Duke of Buckingham. He was pressed to reveal the names of his accomplices. The King's Council debated whether 'by the Law of the Land they could justify the putting him to the Rack'. The King, Charles I, said that before this was done 'let the Advice of the Judges be had therein, whether it be Legal or no'. The King said that if it might not be done by law 'he would not use his Prerogative in this Point'. So the judges were consulted. They assembled at Serjeants' Inn in Fleet Street and agreed unanimously that Felton 'ought not by the Law to be tortured by the Rack, for no such Punishment is known or allowed by our Law': Rushworth, Historical Collections (1721) vol 1, pages 638-639.

Doubt has been cast on the historical accuracy of this account: Jardine, 'Use of Torture in the Criminal Law of England', (1837), pages 61-62. The precise detail does not matter. What matters is that never again did the Privy Council issue a torture warrant. Nor, after 1640, did the king issue a warrant under his own signet: see Professor Langbein, 'Torture and the Law of Proof', pages 134-135. In Scotland prohibition of torture came later, after the union of the two kingdoms, under section 5 of the Treason Act 1708.

It is against the background of this long established principle and practice that your Lordships' House must now decide whether an English court can admit as evidence in court proceedings information extracted by torture administered overseas. If an official or agent of the United Kingdom were to use torture, or connive at its use, in order to obtain information this information would not be admissible in court proceedings in this country. That is not in doubt. It would be an abuse of the process of the United Kingdom court for the United Kingdom government to seek to adduce in evidence information so obtained. The court would not for one moment countenance such conduct by the state. But what if agents of other countries extract information by use of torture? Is this information admissible in court proceedings in this country?

Torture attracts universal condemnation, as amply demonstrated by my noble and learned friend Lord Bingham of Cornhill. No civilised society condones its use. Unhappily, condemnatory words are not always matched by conduct. Information derived from sources where torture is still practised gives rise to the present problem. The context is cross-border terrorism. Countering international terrorism calls for a flow of information between the security services of many countries. Fragments of information, acquired from various sources, can be pieced together to form a valuable picture, enabling governments of threatened countries to take preventative steps. What should the security services and the police and other executive agencies of this country do if they know or suspect information received by them from overseas is the product of torture? Should they discard this information as 'tainted', and decline to use it lest its use by them be regarded as condoning the horrific means by which the information was obtained?

The intuitive response to these questions is that if use of such information might save lives it would be absurd to reject it. If the police were to learn of the whereabouts of a ticking bomb it would be ludicrous for them to disregard this information if it had been procured by torture. No one suggests the police should act in this way. Similarly, if tainted information points a finger of suspicion at a particular individual: depending on the circumstances, this information is a matter the police may properly take into account when considering, for example, whether to make an arrest.

In both these instances the executive arm of the state is open to the charge that it is condoning the use of torture. So, in a sense, it is. The government is using information obtained by torture. But in cases such as these the government cannot be expected to close its eyes to this information at the price of endangering the lives of its own citizens. Moral repugnance to torture does not require this.

The next step is to consider whether the position is the same regarding the use of this information in legal proceedings and, if not, why not. In my view the position is not the same. The executive and the judiciary have different functions and different responsibilities. It is one thing for tainted information to be used by the executive when making operational decisions or by the police when exercising their investigatory powers, including powers of arrest. These steps do not impinge upon the liberty of individuals or, when they do, they are of an essentially short-term interim character. Often there is an urgent need for action. It is an altogether different matter for the judicial arm of the state to admit such information as evidence when adjudicating definitively upon the guilt or innocence of a person charged with a criminal offence. In the latter case repugnance to torture demands that proof of facts should be found in more acceptable sources than information extracted by torture.

Difficulties arise at the interface between the different approaches permitted to the executive on the one hand and demanded of the courts on the other hand. Problems occur where the lawfulness of executive decisions is challenged in court and there is an apparent 'mismatch', as the Secretary of State described it, between the material lawfully available to the executive and the evidence a court will admit in its proceedings. Suppose a case where the police take into account information obtained by torture abroad when arresting a person, and that person subsequently challenges the lawfulness of his arrest. Can the police give evidence of this information in court when seeking to justify the arrest?

In my view they can. It would be remarkable if the police could not. That would create a bizarre situation. It would mean the police may rely on this evidence when making an arrest, but not if the lawfulness of the arrest is challenged. That would be a curious application of a moral principle. That would be to treat a moral principle as giving with one hand and taking away with the other. That makes no sense. Either the police may rely on such information when carrying out their duties, or they may not. If they can properly have regard to such information despite its tainted source, and in the particular case do so, they should not be precluded from referring to this information in court when giving evidence seeking to justify their decisions and actions. Repugnance to the use in court of information procured by torture does not require the police to give an incomplete account of the matters they took into account when making their decisions. (Different considerations apply where, in the interests of national security, there are statutory or other restrictions on the use of certain matters in legal proceedings, such as the contents of intercepted communications or information attracting public interest immunity. In these cases the 'mismatch' arises from a perceived need to preserve confidentiality, not from the application of a broad moral principle.)

So far I have noted the distinction between executive decisions of an essentially operational or short-term character and judicial decisions on criminal charges. Tainted information may be taken into account in the former case but not the latter. I have also noted that when reviewing the lawfulness of such executive decisions a court may have regard to all the matters the decision-maker properly took into account.

But this categorisation by no means covers the whole ground. Many cases do not conform to this simple division of functions. Executive decisions, such as deportation, may have serious long-term consequences for an individual. And judicial supervision of an executive decision may take different forms. The Anti-terrorism, Crime and Security Act 2001 is a recent instance. Certification of a person as a 'suspected international terrorist' is the responsibility of the Secretary of State. The issue of this certificate authorises the minister to exercise extensive powers, including power under section 23 to detain the certified person indefinitely in certain circumstances. This power of detention, in its adverse impact on an individual, goes far beyond the adverse impact of executive acts such as search and arrest. Detention by order of the executive under the 2001 Act is not a preliminary step leading to a criminal charge.

Despite this difference, in the case of this Act the rationale underlying the distinction between the executive's ability to take into account information procured by torture and the court's refusal to admit such evidence holds good. It holds good because the Special Immigration Appeals Commission, or SIAC in short, is required to review every certificate, by way of appeal or otherwise, and form its own view on whether reasonable grounds currently exist for believing a person's presence is a risk to national security and for suspecting he is a terrorist: sections 25 and 26. If SIAC considers these grounds do not exist the certificate must be cancelled. Thus the certificate issued by the Secretary of State will lead nowhere if SIAC considers reasonable grounds do not exist. The certificate, although a prerequisite to exercise of the Secretary of State's powers under the Act, will be comparatively short-lived in its effect if SIAC considers the necessary reasonable grounds do not exist. In other words, the certificate is in the nature of an essential preliminary step.

For its part, in forming its own view on whether reasonable grounds exist SIAC is discharging a judicial function which calls for proof of facts by evidence. The ethical ground on which information obtained by torture is not admissible in court proceedings as proof of facts is applicable in these cases as much as in other judicial proceedings. That is the present case.

Similar problems are bound to arise with other counter-terrorism legislation. One instance concerns decisions by the Secretary of State to deport on the ground that deportation is conducive to the public good as being in the interests of national security. An appeal lies to SIAC, which must allow an appeal if the decision involved the exercise of discretion by the minister and SIAC considers the discretion should have been exercised differently: section 2 of the Special Immigration Appeals Commission Act 1997, as substituted by the Nationality, Immigration and Asylum Act 2002. Another instance concerns non-derogating control orders made by the Secretary of State under section 2 of the Prevention of Terrorism Act 2005. Here the role of the court is expressed to be of a different and more limited character than under the 2001 Act. Under the 2005 Act the supervisory role of the court regarding non-derogating control orders is essentially limited to considering whether the relevant decision of the Secretary of State is 'flawed'. In deciding this issue the court must apply the 'principles applicable on an application for judicial review': section 3(11).

Whether the Secretary of State may take tainted information into account when making decisions under statutory provisions such as these, and whether SIAC's function requires or permits evidence to be given of all the matters taken into account by the Secretary of State, are questions for another day. They do not call for decision on these appeals, and they were not the subject of submissions. It would not be right therefore to express any view on these issues.

For these reasons, and those stated by my noble and learned friends, I would allow these appeals.

In doing so I associate myself with the observations of Lord Bingham of Cornhill on the burden of proof where the admissibility of evidence is challenged before SIAC on the ground it may have been procured by torture. The contrary approach would place on the detainee a burden of proof which, for reasons beyond his control, he can seldom discharge. In practice that would largely nullify the principle, vigorously supported on all sides, that courts will not admit evidence procured by torture. That would be to pay lip-service to the principle. That is not good enough.

My Lords,

On 23 August 1628 George Villiers, Duke of Buckingham and Lord High Admiral of England, was stabbed to death by John Felton, a naval officer, in a house in Portsmouth. The 35-year-old Duke had been the favourite of King James I and was the intimate friend of the new King Charles I, who asked the judges whether Felton could be put to the rack to discover his accomplices. All the judges met in Serjeants' Inn. Many years later Blackstone recorded their historic decision:

That word honour, the deep note which Blackstone strikes twice in one sentence, is what underlies the legal technicalities of this appeal. The use of torture is dishonourable. It corrupts and degrades the state which uses it and the legal system which accepts it. When judicial torture was routine all over Europe, its rejection by the common law was a source of national pride and the admiration of enlightened foreign writers such as Voltaire and Beccaria. In our own century, many people in the United States, heirs to that common law tradition, have felt their country dishonoured by its use of torture outside the jurisdiction and its practice of extra-legal "rendition" of suspects to countries where they would be tortured: see Jeremy Waldron, Torture and Positive Law: Jurisprudence for the White House 105 Columbia Law Review 1681-1750 (October, 2005)

Just as the writ of habeas corpus is not only a special (and nowadays infrequent) remedy for challenging unlawful detention but also carries a symbolic significance as a touchstone of English liberty which influences the rest of our law, so the rejection of torture by the common law has a special iconic importance as the touchstone of a humane and civilised legal system. Not only that: the abolition of torture, which was used by the state in Elizabethan and Jacobean times to obtain evidence admitted in trials before the court of Star Chamber, was achieved as part of the great constitutional struggle and civil war which made the government subject to the law. Its rejection has a constitutional resonance for the English people which cannot be overestimated.

During the last century the idea of torture as a state instrument of special horror came to be accepted all over the world, as is witnessed by the international law materials collected by my noble and learned friend Lord Bingham of Cornhill. Among the many unlawful practices of state officials, torture and genocide are regarded with particular revulsion: crimes against international law which every state is obliged to punish wherever they may have been committed.

It is against that background that one must examine the Secretary of State's submission that statements obtained abroad by torture are admissible in appeals to the Special Immigration Appeals Commission ("SIAC") under section 25 of the Anti-terrorism, Crime and Security Act 2001. First, he says that there is no authority to the contrary. He accepts that the common law has long held that confessions obtained by torture are inadmissible against an accused person. Indeed, the common law went a good deal further and by the end of the eighteenth century was refusing to admit confessions which had been obtained by threats or promises of any kind. But nothing was said about statements obtained from third parties. The general rule is that any relevant evidence is admissible. As Lord Goddard said in Kuruma v The Queen [1955] AC 197, 203, "the court is not concerned with how the evidence was obtained". He referred to a remark of Crompton J in R v Leathem (1861) 8 Cox CC 498, 501, overruling an objection to production of a letter which had been discovered in consequence of an inadmissible statement made by the accused: "It matters not how you get it; if you steal it even, it would be admissible."

It is true that there are no cases in which statements from third parties have been held inadmissible on the ground that they had been obtained by torture. But the reason is not because such statements have been admitted in an ordinary English court. That has never happened. It is because ever since the late 17th century, any statements made by persons not testifying before the court have been excluded, whatever the circumstances in which they were made. There was no need to consider whether they had been obtained by torture. They were simply rejected as hearsay. One must therefore try to imagine what the judges would have said if there had been no hearsay rule. Is it credible that, while rejecting a confession obtained by torture from the accused, they would have admitted a confession incriminating the accused which had been obtained by torturing an accomplice? Such a proceeding was precisely what had been held to be unlawful in the case of Felton. It is absurd to suppose that the judges would have said that the torture was illegal but that a statement so obtained would nevertheless be admissible.

As is shown by cases like Kuruma, not all evidence unlawfully obtained is inadmissible. Still less is evidence inadmissible only because it was discovered in consequence of statements which would not themselves be admissible, as in Leathem and the leading case of R v Warickshall (1783) 1 Leach 263, in which evidence that stolen goods were found under the bed of the accused was admitted notwithstanding that the discovery was made in consequence of her inadmissible confession. But the illegalities with which the courts were concerned in Kuruma and Leathem were fairly technical. Lord Goddard was not considering torture. In any case, since Kuruma the law has moved on. English law has developed a principle, illustrated by cases like R v Horseferry Road Magistrates' Court, Ex p Bennett [1994] 1 AC 42, that the courts will not shut their eyes to the way the accused was brought before the court or the evidence of his guilt was obtained. Those methods may be such that it would compromise the integrity of the judicial process, dishonour the administration of justice, if the proceedings were to be entertained or the evidence admitted. In such a case the proceedings may be stayed or the evidence rejected on the ground that there would otherwise be an abuse of the processes of the court.

As for the rule that we do not necessarily exclude the "fruit of the poisoned tree", but admit relevant evidence discovered in consequence of inadmissible confessions, this is the way we strike a necessary balance between preserving the integrity of the judicial process and the public interest in convicting the guilty. And even when the evidence has been obtained by torture - the accomplice's statement has led to the bomb being found under the bed of the accused - that evidence may be so compelling and so independent that it does not carry enough of the smell of the torture chamber to require its exclusion. But that is not the question in this case. We are concerned with the admissibility of the raw product of interrogation under torture.

The curious feature of this case is that although the Secretary of State advances these arguments based on the limited scope of the confession rule and the general principle that all relevant evidence is admissible, he does not contend for what would be the logical consequence if he was right, namely, that evidence obtained from third parties by torture in the United Kingdom would also be admissible. He accepts that it would not. But he submits that the exclusionary rule is confined to cases in which the torture has been used by or with the connivance of agents of the United Kingdom. So the issue is a narrow one: not whether an exclusionary rule exists, but whether it should extend to torture inflicted by foreigners without the assistance or connivance of anyone for whom the United Kingdom is responsible.

Furthermore, the Secretary of State has attempted to fend off concern by the International Committee Against Torture over whether his position was in accordance with our obligations under article 15 of the UN Convention Against Torture ("Each State Party shall ensure that any statement which is established to have been made as a result of torture shall not be invoked as evidence in any proceedings") by saying that he does not intend to "rely upon or present evidence where there is a knowledge or belief that torture has taken place". No doubt he thought that in addition to being an international obligation, that was the least that decency required. But the Secretary of State insists that this is a matter of policy which he is free to change or depart from. So the question remains over whether such evidence is admissible as a matter of English law.

The answer to that question depends upon the purpose of the rule excluding evidence obtained by torture, which, as we have seen, the Secretary of State largely admits to exist. Is it to discipline the executive agents of the state by demonstrating that no advantage will come from torturing witnesses, or is it to preserve the integrity of the judicial process and the honour of English law? If it is the former, then of course we cannot aspire to discipline the agents of foreign governments. Their torturers would probably accept with indifference the possibility that the work of their hands might be rejected by an English court. If it is the latter, then the rule must exclude statements obtained by torture anywhere, since the stain attaching to such evidence will defile an English court whatever the nationality of the torturer. I have no doubt that the purpose of the rule is not to discipline the executive, although this may be an incidental consequence. It is to uphold the integrity of the administration of justice.

The Secretary of State's second argument is that while there may be a general rule which excludes all evidence obtained by torture in an ordinary criminal trial, proceedings before SIAC are different. The function of SIAC under section 25 of the 2001 Act is not to convict anyone of an offence but to decide whether there are reasonable grounds for belief or suspicion that a person's presence in the United Kingdom is a risk to national security or that he is a terrorist: subsection (2)(a). There is no restriction upon the information which the Secretary of State may consider in forming such a belief or suspicion. In the exercise of his functions, he may rely upon statements from any source and in some cases it may be foolish of him not to do so. If the Security Services receive apparently credible information from a foreign government that bombs are being made at an address in south London, it would be irresponsible of the Secretary of State not to instigate a search of the premises because he has a strong suspicion that the statement has been obtained by torture. So, it is said, the exclusionary rule would produce a "mismatch" between the evidence upon which the Secretary of State could rely and the evidence upon which SIAC could rely in the exercise of its supervisory jurisdiction over the Secretary of State under the Act. Furthermore, rule 44(3) of the Special Immigration Appeals Commission (Procedure) Rules 2003 (SI 2003/1034) specifically provides that the Commission "may receive evidence that would not be admissible in a court of law". The purpose of that rule, it is argued, is to allow SIAC to consider any evidence which could have been considered by the Secretary of State.

In my opinion the "mismatch" to which counsel for the Secretary of State refers is almost inevitable in any case of judicial supervision of executive action. It is not the function of the courts to place limits upon the information available to the Secretary of State, particularly when he is concerned with national security. Provided that he acts lawfully, he may read whatever he likes. In his dealings with foreign governments, the type of information that he is willing to receive and the questions that he asks or refrains from asking are his own affair. As I have said, there may be cases in which he is required to act urgently and cannot afford to be too nice in judging the methods by which the information has been obtained, although I suspect that such cases are less common in practice than in seminars on moral philosophy.

But the 2001 Act makes the exercise by the Secretary of State of his extraordinary powers subject to judicial supervision. The function of SIAC under section 25 is not to decide whether the Secretary of State at some particular time, perhaps at a moment of emergency, acted reasonably in forming some suspicion or belief. It is to form its own opinion, after calm judicial process, as to whether it considers that there are reasonable grounds for such suspicion or belief. It is exercising a judicial, not an executive function. Indeed, the fact that the exercise of the draconian powers conferred by the Act was subject to review by the judiciary was obviously an important reason why Parliament was willing to confer such powers on the Secretary of State.

In my opinion Parliament, in setting up a court to review the question of whether reasonable grounds exist for suspicion or belief, was expecting the court to behave like a court. In the absence of clear express provision to the contrary, that would include the application of the standards of justice which have traditionally characterised the proceedings of English courts. It excludes the use of evidence obtained by torture, whatever might be its source.

Rule 44(3) is in my opinion far too general in its terms to justify a departure from such a fundamental principle. It plainly disapplies technical rules of evidence like the hearsay rule. But I cannot for a moment imagine that anyone in Parliament who considered the statutory power to make rules of procedure for SIAC could have thought that it was authorising a rule which allowed the use of evidence obtained by torture or that the Secretary of State who made the regulations thought he was doing so. Such a provision, touching upon the honour of our courts and our country, would have to be expressly provided in primary legislation so that it could be debated in Parliament.

In my opinion therefore, there is a general rule that evidence obtained by torture is inadmissible in judicial proceedings. That leaves the question of what counts as evidence obtained by torture. What is torture and who has the burden of proving that it has been used? In Ireland v United Kingdom (1978) 2 EHRR 25 the European Court delicately refrained from characterising various interrogation techniques used by the British authorities in Northern Ireland as torture but nevertheless held them to be "inhuman treatment". The distinction did not matter because in either case there was a breach of article 3 of the Convention. For my part, I would be content for the common law to accept the definition of torture which Parliament adopted in section 134 of the Criminal Justice Act 1988, namely, the infliction of severe pain or suffering on someone by a public official in the performance or purported performance of his official duties. That would in my opinion include the kind of treatment characterised as inhuman by the European Court of Human Rights in Ireland v United Kingdom but would not include all treatment which that court has held to contravene article 3.

That leaves the question of the burden of proof, on which I am in agreement with my noble and learned friend Lord Bingham of Cornhill. In proceedings in which the appellant to SIAC may have no knowledge of the evidence against him, it would be absurd to require him to prove that it had been obtained by torture. Article 15 of the Torture Convention, which speaks of the use of torture being "established", could never have contemplated a procedure in which the person against whom the statement was being used had no idea of what it was or who had made it. It must be for SIAC, if there are reasonable grounds for suspecting that to have been the case (for example, because of evidence of the general practices of the authorities in the country concerned) to make its own inquiries and not to admit the evidence unless it is satisfied that such suspicions have been rebutted. One of the difficulties about the Secretary of State's carefully worded statement that it would not be his policy to rely upon evidence "where there is a knowledge or belief that torture has taken place" is that it leaves open the question of how much inquiry the Secretary of State is willing to make. It appears to be the practice of the Security Services, in their dealings with those countries in which torture is most likely to have been used, to refrain, as a matter of diplomatic tact or a preference for not learning the truth, from inquiring into whether this was the case. It may be that in such a case the Secretary of State can say that he has no knowledge or belief that torture has taken place. But a court of law would not regard this as sufficient to rebut real suspicion and in my opinion SIAC should not do so.

In view of the great importance of this case for the reputation of English law, I have thought it right to express my opinion in my own words. But I have had the advantage of reading in draft the speech of my noble and learned friend Lord Bingham of Cornhill and there is nothing in it with which I would wish to disagree.

My Lords,

I have had the advantage of reading in draft the speech of my noble and learned friend Lord Bingham of Cornhill. His account of the background to this case is so complete that I hesitate to say anything that might detract from it. But it is one thing to condemn torture, as we all do. It is another to find a solution to the question that this case raises which occupies the moral high ground but at the same time serves the public interest and is practicable. Condemnation is easy. Finding a solution to the question is much more difficult. It requires much more thought. So it is on that aspect of the case in particular, after looking at the history, that I should like to concentrate.

Torture, one of most evil practices known to man, is resorted to for a variety of purposes and it may help to identify them to put this case into its historical context. The lesson of history is that, when the law is not there to keep watch over it, the practice is always at risk of being resorted to in one form or another by the executive branch of government. The temptation to use it in times of emergency will be controlled by the law wherever the rule of law is allowed to operate. But where the rule of law is absent, or is reduced to a mere form of words to which those in authority pay no more than lip service, the temptation to use torture is unrestrained. The probability of its use will rise or fall according the scale of the perceived emergency.

In the first place, torture may be used on a large scale as an instrument of blatant repression by totalitarian governments. That is what was alleged in R v Bow Street Metropolitan Stipendiary Magistrate, Ex p Pinochet Ugarte (No 3) [2000] 1 AC 147, where the picture presented by the draft charges against Senator Pinochet which had been prepared by the Spanish judicial authorities was of a conspiracy. It was a conspiracy of the most evil kind - to commit widespread and systematic torture and murder to obtain control of the government and, having done so, to maintain control of government by those means for so long as might be necessary. Or it may be used in totalitarian states as a means of extracting confessions from individuals whom the authorities wish to put on trial so that they can be used against them in evidence.

The examples I have just mentioned are of torture as an instrument of power. But the use of torture to obtain confessions was also sanctioned by the judiciary in many civil law jurisdictions, and it remained part of their criminal procedure until the latter part of the 17th century. This was never part of English criminal procedure and, as there was no need for it, its use for this purpose was prohibited by the common law. But warrants for the use of torture were issued from time to time by the Privy Council against prisoners in the Tower under the Royal Prerogative. Four hundred years ago, on 4 November 1605, Guy Fawkes was arrested when he was preparing to blow up the Parliament which was to be opened the next day, together with the King and all the others assembled there. Two days later James I sent orders to the Tower authorising torture to be used to persuade Fawkes to confess and reveal the names of his co-conspirators. His letter stated that "the gentler tortours" were first to be used on him, and that his torturers were then to proceed to the worst until the information was extracted out of him. On 9 November 1605 he signed his confession with a signature that was barely legible and gave the names of his fellow conspirators. On 27 January 1606 he and seven others were tried before a special commission in Westminster Hall. Signed statements in which they had each confessed to treason were shown to them at the trial, acknowledged by them to be their own and then read to the jury: Carswell, Trial of Guy Fawkes (1934), pp 90-92.

This practice came to an end in 1640 when the Act of 16 Charles I, c 10, abolished the Star Chamber. The jurisdiction of the Privy Council in all matters affecting the liberty of the subject was transferred to the ordinary courts, which until then in matters of State the executive could by-pass. Torture continued to be used in Scotland on the authority of the Privy Council until the end of the 17th century, but the practice was brought to an end there after the Union by section 5 of the Treason Act 1708. That section, which remains in force subject only to one minor amendment (see Statute Law (Repeals) Act 1977, Sch I, Part IV) and applies to England as well as Scotland, declares that no person accused of any crime can be put to torture.

We are not concerned in this case with the use of torture for either of the purposes that I have mentioned so far. But they do not exhaust the uses for which torture may be sanctioned by governments. The use with which this case is concerned is the extraction of information from those who are thought to have something that may be of use to them by the security services. Information - the gathering of intelligence - is a crucial weapon in the battle by democracies against international terrorism. Experience has shown from the beginning of time that those who are hostile to the state are reluctant to part with information that might disrupt or inhibit their activities. They usually have to be persuaded to release it. Handled responsibly, the methods that are used fall well short of what could reasonably be described as torture. But in unscrupulous hands the means of persuasion are likely to be violent and intended to inflict severe physical or mental pain or suffering. In the hands of the most unscrupulous the only check on the level of violence is likely to be the need to keep the person alive so that, if he has any information that may be useful, he can communicate it to his interrogators.

It was not unknown during the 17th century, while torture was still being practised here, for statements extracted by this means to be used as evidence in criminal proceedings to obtain the conviction of third parties. J H Langbein, Torture and the Law of Proof: Europe and England in the Ancien Regime (University of Chicago Press, 1977), p 94 has shown that a warrant was issued by the Privy Council in 1551 for the torture of persons committed to the Tower on suspicion of being involved in the alleged treason of the Duke of Somerset. The confession obtained from William Crane was read, in Crane's absence, at the Duke's trial: Heath, Torture and English Law: An Administrative and Legal History from the Plantagenets to the Stuarts (1982), p 75.

When the jurisdiction of the Star Chamber was abolished in England prisoners were transferred to Scotland so that they could be forced by the Scots Privy Council which still used torture to provide information to the authorities. This is illustrated by the case of Robert Baillie of Jerviswood whose trial took place in Edinburgh in December 1684. A detailed description of the events of that trial can be found in Fountainhall's Decisions of the Lords of Council and Session, vol I, 324-326: for a summary, see Torture [2004] 53 ICLQ 807, 818-820. Robert Baillie had been named by William Spence, who was suspected of being involved in plotting a rebellion against the government of Charles II, as one of his co-conspirators. Spence gave this information having been arrested in London and taken to Edinburgh, where he was tortured. Baillie in his turn was arrested in England and taken to Scotland, where he was put on trial before a jury in the High Court of Justiciary in Edinburgh. All objections having been repelled by the trial judge, the statement which Spence had given under torture was read to the jury. Baillie was convicted the next day, and the sentence of death that was passed on him was executed that afternoon. There is a warning here for us. "Extraordinary rendition", as it is known today, is not new. It was being practised in England in the 17th century.

Baron Hume, Commentaries on the Law of Scotland respecting Crimes (Edinburgh, 1844), vol ii, p 324, described the use of torture for the purpose of discovering transgressors as a barbarous engine. So it was. It had increasingly come to be recognised that there was a level beyond which, however great the threat and however imminent its realisation, resort to this means of extracting information was unacceptable. The need of the authorities to resort to extreme measures for their own protection had, of course, disappeared with the arrival of the period of stability that came with the ending of the Stuart dynasty. But one can detect in Hume's language a revulsion against its use which would have certainly been voiced by the judges of his time, had it been necessary for them to do so.

The threat of rebellion and revolution having disappeared, the developing common law did not find it necessary to grapple with the question whether statements obtained by the use of torture should continue to be admissible against third parties in any proceedings as evidence. There is no doubt that they would be caught today by the rule that evidence of the facts referred to in a statement made by a third party, however that statement was obtained, is hearsay: Teper v The Queen [1952] AC 480, 486, per Lord Normand. Alison, Principles and Practice of the Criminal Law of Scotland (1833), vol ii, 510-11 states that hearsay is in general inadmissible evidence. He bases this proposition on the best evidence rule, and declares that the rule is "firmly established both in the Scotch and English law". But we cannot be absolutely confident that judges in the latter part of the 19th century would have been prepared to rely on the hearsay rule to exclude such evidence. In R v Birmingham Overseers (1861) 1 B & S 763, 767, Cockburn CJ said:

If, as this passage indicates, the hearsay objection went only to the weight of the evidence, the judges would have had to face up to the more fundamental question whether at common law it was an abuse of the judicial process to rely on it.

I think that it is plain that the barbarity of the practice, as Hume describes it, would have led inevitably to the conclusion that the use against third parties of statements obtained in this way as evidence in any proceedings was unacceptable. This would have been a modest but logical extension of the rule already enshrined in statute by section 5 of the Treason Act 1708, that no person accused of a crime could be put to torture. The effect of that section was to render confession evidence obtained by this means inadmissible. It would have been a small but certain step to apply the same rule to statements obtained in the same way from third parties.

This is the background to the ratification by the United Kingdom of the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment which was adopted by the United Nations General Assembly on 10 December 1984 and entered into force on 26 June 1987. The Convention was designed to provide an international system which denied a safe haven to the official torturer. But long before it was entered into state torture was an international crime in the highest sense, as Lord Browne-Wilkinson pointed out in R v Bow Street Metropolitan Stipendiary Magistrate, Ex p Pinochet Ugarte (No 3) [2000] 1 AC 147, p 198G. The rule set out in article 15 of the Convention about the use of statements obtained by the use of torture must be seen in this light. Article 15 provides:

This provision has not been incorporated into our domestic law, unlike the declaration that the use of torture is a crime wherever it was committed which was made part of our law by section 134 of the Criminal Justice Act 1988. But I would hold that the formal incorporation of the evidential rule into domestic law was unnecessary, as the same result is reached by an application of common law principles. The rule laid down by article 15 was accepted by the United Kingdom because it was entirely compatible with our own law. The use of such evidence is excluded not on grounds of its unreliability - if that was the only objection to it, it would go to its weight, not to its admissibility - but on grounds of its barbarism, its illegality and its inhumanity. The law will not lend its support to the use of torture for any purpose whatever. It has no place in the defence of freedom and democracy, whose very existence depends on the denial of the use of such methods to the executive.

Once torture has become acclimatised in a legal system it spreads like an infectious disease, hardening and brutalising those who have become accustomed to its use: Holdsworth, A History of English Law, vol v, p 194. As Jackson J in his dissenting opinion in Korematsu v United States, 323 US 214 (1944), 246 declared, once judicial approval is given to such conduct, it lies about like a loaded weapon ready for the hand of any authority that can bring forward a plausible claim of an urgent need. A single instance, if approved to meet the threat of international terrorism, would establish a principle with the power to grow and expand so that everything that falls within it would be regarded as acceptable. Without hesitation I would hold that, subject to the single exception referred to in article 15, the admission of any statements obtained by this means against third parties is absolutely precluded in any proceedings as evidence. I would apply this rule irrespective of where, or by whom, the torture was administered.

Rule 44(3) of the Special Immigration Appeals Commission (Procedure) Rules 2003 (2003/1034) provides that the Commission may receive evidence that would not be admissible in a court of law. But I consider, in agreement with all your Lordships, that this rule is incompatible with the fundamental nature of the objection to the admission of statements obtained by the use of torture, wherever it was administered, and that it does not extend to them. That being the nature of the objection, the question whether it can be overridden and, if so, in what circumstances must be left to the legislature. This is not a matter that can be left to implication. Nothing short of an express provision will do, to which Parliament has unequivocally committed itself.

There are ample grounds for suspecting that the use of torture on detainees suspected of involvement in international terrorism is widespread in countries with whom the security services of the United Kingdom are in contact. The Secretary of State's position is that he does not rely on information that he knows has been obtained by torture, as a matter of principle. But he is willing to accept and act upon information whose origin is obscure and undetectable, in the knowledge that it may have come from countries that use torture. He says that it is for the party who objects to its use on the ground that torture was used to make good his objection. What then is the approach that SIAC should take to this issue?

I agree that a conventional approach to the burden of proof is inappropriate in this context. It would be wholly unrealistic to expect the detainee to prove anything, as he is denied access to so much of the information that is to be used against him. He cannot be expected to identify from where the evidence comes, let alone the persons who have provided it. All he can reasonably be expected to do is to raise the issue by asking that the point be considered by SIAC. There is, of course, so much material in the public domain alleging the use of torture around the world that it will be easy for the detainee to satisfy that simple test. All he needs to do is point to the fact that the information which is to be used against him may have come from one of the many countries around the world that are alleged to practise torture, bearing in mind that even those who say that they do not use torture apply different standards from those that we find acceptable. Once the issue has been raised in this general way the onus will pass to SIAC. It has access to the information and is in a position to look at the facts in detail. It must decide whether there are reasonable grounds to suspect that torture has been used in the individual case that is under scrutiny. If it has such a suspicion, there is then something that it must investigate as it addresses its mind to the information that is put before it which has been obtained from the security services.

Guidance needs to be given on this point too. Do the facts need to be established beyond a reasonable doubt or do they need to be established only on a balance of probabilities? To answer this question we must know what it is that has to be established. It is at the point of defining what SIAC must inquire into that, with the greatest of respect, I begin to differ from Lord Bingham. He says that it is for SIAC to initiate or direct such inquiry as is necessary to enable it to form a fair judgment whether the evidence has, or whether there is a real risk that it may have been, obtained by torture or not. But it is one thing if what SIAC is to be required to do is to form a fair judgment as to whether the evidence has, or may have been, obtained by torture. It is another if what it is to be required to do is to form a fair judgment as to whether it has not, or may not, have been obtained by torture.

Lord Bingham then says that SIAC should refuse to admit the evidence if it is unable to conclude that there is not a real risk that the evidence has been obtained by torture. My own position, for reasons that I shall explain more fully in the following paragraphs, is that SIAC should refuse to admit the evidence if it concludes that the evidence was obtained by torture. I am also firmly of the view that, if it approaches the issue in this way, it should apply the lower standard of proof. The liberty of the subject dictates this. So SIAC should not admit the evidence if it concludes on a balance of probabilities that it was obtained by torture. In other words, if SIAC is left in doubt as to whether the evidence was obtained in this way, it should admit it. But it must bear its doubt in mind when it is evaluating the evidence. Lord Bingham's position, as I understand it, is that if it is left in doubt SIAC should exclude the evidence. That, in short, is the only difference between us.

I must now explain why I believe that the question which SIAC must address should be put positively rather than negatively. The effect of rule 44(3) of the Procedure Rules is that sources of all kinds may be relied upon, far removed from what a court of law would regard as the best evidence. SIAC may be required to look at information coming to the attention of the security services at third or fourth hand and from various sources, the significance of which cannot be determined except by looking at the whole picture which it presents. The circumstances in which the information was first obtained may be incapable of being detected at all or at least of being determined without a long and difficult inquiry which would not be practicable. So it would be unrealistic to expect SIAC to demand that each piece of information be traced back to its ultimate source and the circumstances in which it was obtained investigated so that it could be proved piece by piece, that it was not obtained under torture. The threshold cannot be put that high. Too often we have seen how the lives of innocent victims and their families are torn apart by terrorist outrages. Our revulsion against torture, and the wish which we all share to be seen to abhor it, must not be allowed to create an insuperable barrier for those who are doing their honest best to protect us. A balance must be struck between what we would like to achieve and what can actually be achieved in the real world in which we all live. Articles 5(4) and 6(1) of the European Convention, to which Lord Bingham refers in para 62, must be balanced against the right to life that is enshrined in article 2 of the Convention.

I would take as the best guide to what is practicable the approach that article 15 of the Torture Convention takes to this issue. The United Nations has adopted it, and it has the support of all the signatories to the Convention. So it deserves to be respected as the best guide that international law has to offer on this issue. First, the exclusionary rule that it lays down applies to statements obtained under torture, not to information that may have been discovered as a result of them. Logic might suggest that the fruits of the poisoned tree should be discarded too. But the law permits evidence to be led however it was obtained, if the evidence is in itself admissible: Kuruma v The Queen [1955] AC 197. Secondly, the exclusionary rule applies to "any proceedings". Mr Burnett QC for the Secretary of State suggested that this phrase should be read as extending to criminal proceedings only, but I would not so read it. The word "any" is all-embracing and it is perfectly capable of applying to the proceedings conducted by SIAC.

Thirdly, and crucially, the exclusionary rule extends to any statement that "is established" to have been made under torture. The rule does not require it to be shown that the statement was not made under torture. It does not say that the statement must be excluded if there is a suspicion of torture and the suspicion has not been rebutted. Nor does it say that it must be excluded if there is a real risk that it was obtained by torture. An evaluation of risk is appropriate if the question at issue relates to the future: see Mamatkulov and Askarov v Turkey (Application Nos 46827/99 and 46951/99) 4 February 2005, para 71. The question in that case was whether there was a real risk for the purposes of article 3 of the European Convention at the time of their extradition that the applicants would be tortured. The rule that article 15 lays down looks at what has happened in the past. It applies to a statement that is established to have been made under torture. In my opinion the test that it lays down is the test that should be applied by SIAC. It too must direct its inquiry to what has happened in the past. Is it established, by means of such diligent inquiries into the sources that it is practicable to carry out and on a balance of probabilities, that the information relied on by the Secretary of State was obtained under torture? If that is the position, article 15 requires that the information must be left out of account in the overall assessment of the question whether there were no reasonable grounds for a belief or suspicion of the kind referred to in section 21(1) (a) or (b) of the Anti-terrorism, Crime and Security Act 2001. The same rule must be followed in any other judicial process where information of this kind would otherwise be admissible.

Support for this approach is to be found in a decision in the case of El Motassadeq of the Hanseatisches Oberlandesgericht (the Hanseatic Court of Appeals, Criminal Division), Hamburg of 14 June 2005, NJW 2005, 2326. El Motassadeq had been charged with conspiracy to cause the attacks of 11 September 2001 on the United States of America and with membership of an illegal organisation. The court had been provided by the US Department of Justice with summaries of statements of three witnesses which, subject to certain safeguards, were admissible under its Code of Criminal Procedure as equivalent to written records of statements by these witnesses. The court was, of course, aware from press articles and other reports that there were indications that suspected Al Qaeda members had been subjected to torture within the meaning of article 1 of the Convention, and it was contended that these statements should be excluded under article 15. Repeated requests to the competent US authorities for information about the circumstances of the examination of these witnesses met with no response, and attempts to obtain this information through the German authorities were blocked on the ground that the information had been given to them for intelligence purposes only and that a breach of the limitations of use would jeopardise the security interests of the Federal Republic of Germany. In this situation the court had no option but to base its assessment of the question whether torture had been used on available, publicly accessible sources. On the one hand the White House denied that it used or condoned torture. On the other hand it had admitted that it did not view Al Qaeda prisoners as coming under the protection of international human rights agreements on the treatment of prisoners of war. This was enough to raise the suspicion that torture had been used. There was a question to answer on this point.

The court's conclusions are to be found in the following paragraphs of the certified translation:

…

In a concluding paragraph the court said that it was mindful of the problems posed by the possible use of torture and would take this into consideration when assessing the information in the summaries, adding: "This does not imply legitimisation of the use of torture, even in view of the enormous scale of the attacks of 11 September 2001."

The significant points that I would draw from that case are these. The court was careful to distinguish between the generalised allegations of torture which were to be found in the press articles and other materials - sufficient, it might well be said, to raise a suspicion of torture - and the position of these three witnesses in particular. What it was looking for was evidence which established that the statements of these three witnesses in particular had been obtained under torture. The test which it was asked to apply was that laid down by the article. The evidence for assuming that torture had been used was said to be weak, and the contents of the statements tended to show that torture had not been used. The court did not go so far as to say that it was unable to conclude that there was not a real risk that the evidence had been obtained by torture. It was left in a state of doubt on this point. If it had applied the test which Lord Bingham suggests, the result would have been different because it had been denied access to information about the precise circumstances.

Article 15 of the Convention does not compel us to adopt the test which Lord Bingham suggests, and there are good reasons - as the case of El Motassadeq so clearly demonstrates - for thinking that the terms on which information is passed to the intelligence services would make it impossible for it to be met in practice. Your Lordships were provided with a statement by the Director General of the Security Service which indicates that the problems of obtaining access to the sources of information from foreign intelligence services are just as acute in this country as they appear to have been in Germany. In my opinion the public interest requires us to refrain from setting up a barrier to the use of such information which other nations do not impose on themselves and which is likely in practice to be insuperable. I do not believe that the test which I suggest is one that in the real world can never be satisfied. Nor do I believe that applying the test which the Convention itself lays down in the way I suggest would undermine the practical efficiency of the Convention. I think that we should adhere to what the Convention requires us to do, while making it clear that the issue as to whether torture has been used in the individual case is of the highest importance and that it must, of course, receive the most anxious scrutiny.

There is a fourth element in article 15 which ought to be noticed, although the issue has not been focussed by the facts of this case. The exclusionary rule that article 15 of the Torture Convention lays down extends to statements obtained by the use of torture, not to those obtained by the use of cruel, inhuman or degrading treatment or punishment. That is made clear by article 16.1 of the Convention. The borderline between torture and treatment or punishment of that character is not capable of precise definition. As John Cooper, Cruelty - an analysis of Article 3 (2003), para 1-02 points out, the European Committee for the Prevention of Torture are unwilling to produce a clear and comprehensive interpretation of these terms, their approach being that these are different types of ill-treatment, more or less closely linked. Views as to where the line is to be drawn may differ sharply from state to state. This can be seen from the list of practices authorised for use in Guantanamo Bay by the US authorities, some of which would shock the conscience if they were ever to be authorised for use in our own country. SIAC must exercise its own judgment in addressing this issue, which is ultimately one of fact. It should not be deterred from treating conduct as torture by the fact that other states do not attach the same label to it. The standard that it should apply is that which we would wish to apply in our own time to our own citizens.

For these reasons, although I take a different view from my noble and learned friend Lord Bingham as to the advice that should be given to SIAC, I too would allow the appeals and make the order that he proposes.

My Lords,

I have ultimately come to agree with your Lordships that the appeal should be allowed, but, I confess, I have found the issue far from easy. In resolving it, I have derived considerable assistance from the closely reasoned judgments in the Court of Appeal. Unfortunately, outside the courts, the decision of the majority, Pill and Laws LJJ, has been subjected to sweeping criticisms which to a large extent ignore their reasoning and the very factors which led them to their conclusion.

It should not be necessary to emphasise that the difficulties which troubled the majority in the Court of Appeal and which have troubled me do not arise from any doubt about the unacceptable nature of torture. That has long been unquestioned in this country. The history of the matter shows that torture has been rejected by English common law for many centuries. In Scotland, torture was used until the end of the seventeenth century. For the most part, when used at all, torture seems to have been employed to extract confessions from political conspirators who might be expected to be more highly motivated to resist ordinary methods of interrogation. Such confessions would often contain damning information about other members of the conspiracy. Eventually, section 5 of the Treason Act 1708 declared that no person accused of any crime can be put to torture. The provision is directed at those accused of crime, but this does not mean that Parliament would have been happy for mere witnesses to crime to be tortured. On the contrary, it is an example of the phenomenon, well known in the history of the law from ancient Rome onwards, of a legislature not bothering with what is obvious and dealing only with the immediate practical problem. By 1708, it went without saying that you did not torture witnesses: now Parliament was making it clear that you were not to torture suspects either. So the prohibition on the torture of both witnesses and suspects is deeply ingrained in our system. The corollary of the prohibition is that any statements obtained by officials torturing witnesses or suspects are inadmissible. Most of the considerations of public policy which lead courts to reject such statements are equally applicable to torture carried out abroad by foreign officials. The question for the House is whether that general approach applies to proceedings in SIAC under the Anti-terrorism, Crime and Security Act 2001 ("the 2001 Act").

Information obtained by torture may be unreliable. But all too often it will be reliable and of value to the torturer and his masters. That is why torturers ply their trade. Sadly, the Gestapo rolled up resistance networks and wiped out their members on the basis of information extracted under torture. Hence operatives sent to occupied countries were given suicide pills to prevent them from succumbing to torture and revealing valuable information about their mission and their contacts. In short, the torturer is abhorred as a hostis humani generis not because the information he produces may be unreliable but because of the barbaric means he uses to extract it.

The premise of this appeal is that, despite the United Nations Convention against Torture and any other obligations under international law, some states still practise torture. More than that, those states may supply information based on statements obtained under torture to the British security services who may find it useful in unearthing terrorist plots. Moreover, when issuing a certificate under section 21 of the 2001 Act, the Secretary of State may have to rely on material that includes such statements.

Mr Starmer QC, who appeared for Amnesty and a number of other interveners, indicated that, in their view, it would be wrong for the Home Secretary to rely on such statements since it would be tantamount to condoning the torture by which the statements were obtained. That stance has the great virtue of coherence; but the coherence is bought at too dear a price. It would mean that the Home Secretary might have to fail in one of the first duties of government, to protect people in this country from potential attack. Not surprisingly therefore, Mr Emmerson QC for the appellants was at pains to accept that, when deciding whether to issue a certificate, the Home Secretary was not obliged to check the origins of any statement and could take it into account even if he knew, or had reason to suspect, that it had been obtained by torture. But, he submitted, when SIAC came to discharge its functions under section 25 or 26 of the 2001 Act, in any case where the issue was raised, it could not take account of a statement unless the members were satisfied, beyond reasonable doubt, that it had not been obtained by torture.

On this approach there is a stark disjunction between what the Home Secretary can properly do and what SIAC can properly do. It is, of course, true that, because of public interest immunity or section 17(1) of the Regulation of Investigatory Powers Act 2000, a party to a litigation may not be able to lead evidence of a matter which it was nevertheless legitimate for him to take into account. Such analogies cast little light, however, on a situation where the disjunction arises between sections in the same Act.

Parliament gave jurisdiction in proceedings under sections 25 and 26 of the 2001 Act to SIAC, which had been established by the Special Immigration Appeals Act 1997 in order to meet the criticisms of the European Court of Human Rights in Chahal v United Kingdom (1996) 23 EHRR 413. SIAC is tailor-made to deal with sensitive cases where intelligence material has to be considered. One member of the court will have had experience in handling such material. Section 18(1)(e) of the 2000 Act disapplies section 17(1) and so allows the Commission to consider the content of intercepts. Rule 44(2) of the Special Immigration Appeals Commission (Procedure) Rules 2003 allows the Commission to receive evidence in documentary or any other form, while rule 44(3) allows it to receive evidence that would not be admissible in a court of law. By giving jurisdiction to SIAC, Parliament must have intended that the appeal or review should be considered by a body that was not bound by the ordinary rules of evidence and that was, in general, free to consider all the material that the Home Secretary had taken into account when issuing his certificate. Not surprisingly, therefore, in section 29(1) Parliament provided that any action of the Secretary of State taken wholly or partly in reliance on a section 21 certificate could be questioned only in legal proceedings under section 25 or 26 or under section 2 of the 1997 Act - proceedings in other courts would not be satisfactory since they would not be able to consider the same range of material. Of course, after the certificate was issued, material might often come to hand which strengthened, or even superseded, the material on which the Home Secretary had relied. Conversely, new evidence, or criticism of the existing evidence during the hearing, might undermine the basis for the Home Secretary's decision. SIAC can take account of all that. What is not immediately clear, to me at least, is that Parliament would have contemplated that the specialist tribunal would have to shut its eyes to statements which the Home Secretary was entitled, or perhaps even bound, to take into account. Why should the Secretary of State be entitled to use such a statement to issue a certificate under section 21 if, in default of any additional information, SIAC is then bound to cancel that certificate under section 25 because the members cannot look at the critical statement?

My noble and learned friend, Lord Nicholls of Birkenhead, seeks to resolve the dilemma on the basis that the Secretary of State's certificate is in the nature of an essential preliminary step, which will be short-lived in its effect if SIAC considers that the necessary reasonable grounds do not exist. So the definitive decision is taken by SIAC, which is subject to the ethical rule that information obtained by torture is not admissible in court proceedings as proof of facts. Potentially attractive though such an analysis is, it is rather difficult to square with the fact that, if there is no appeal, SIAC is not required to review the Home Secretary's certificate for six months after it has been issued: section 26(1). A certificate which Parliament regards as sufficient warrant for a suspect's detention for six months is not, in essence, short-lived or a mere preliminary step. And, the appellants concede, such a certificate can properly be based on a statement obtained by torture.

According to the appellants, it is an abuse of process for the Home Secretary to produce evidence of a statement obtained by torture in proceedings before SIAC. In my view it is an abuse of language to characterise the Home Secretary's action as an abuse of process. He does not instigate the process before SIAC and seeks no order from the Commission: he merely seeks to resist an appeal brought against his decision or to withstand a review of that decision. It was perfectly proper for him to rely on the statement when issuing his certificate. There is therefore no abuse of executive power in this country for SIAC to punish by rejecting the statement and it is no part of the function of British courts to attempt to discipline officials of a friendly country. Besides anything else, the idea that foreign torturers would pause for a moment because of a decision by SIAC to reject a statement which they had extracted verges on the absurd.

One therefore comes back to the centuries-old view that statements obtained by torture are unacceptable. To rely on them is inconsistent with the notion of justice as administered by our courts. The Home Secretary does not defile SIAC by introducing such a statement, but he does ask it to rely on a type of statement which British courts would, ordinarily, reject on broad grounds of public policy. SIAC is, of course, different in many ways, as the relevant legislation and regulations show. Therefore, if there were any sign that Parliament had considered the point when passing the Special Immigration Appeals Commission Act 1997 or the 2001 Act, there might be a case for holding that the necessary implication of sections 21, 25 and 26 of the 2001 Act was that SIAC should take account of statements obtained by torture in another country. But that particular issue does not arise since Parliament was never asked to consider the question, either when passing these Acts or when approving the 2003 Rules, including the permissive rule 44(3). The point does not appear to have occurred to anyone. In any event, the revulsion against torture is so deeply ingrained in our law that, in my view, a court could receive statements obtained by its use only where this was authorised by express words, or perhaps the plainest possible implication, in a statute. Here, there are no express words and the provisions actually approved by Parliament do not go so far as to show that the officious bystander who asked whether SIAC could rely on a statement obtained by torture would have been testily suppressed with an "Oh, of course!" from the legislature. I therefore hold that SIAC should not take account of statements obtained by torture.

The courts' deep-seated objection is to torture and to statements obtained by torture. The rejection of such statements is an exception to the general rule that relevant evidence is admissible even if it has been obtained unlawfully. On the other hand, the public interest does not favour SIAC rejecting statements that have not in fact been obtained by torture. More particularly, the public interest does not favour rejecting statements merely because there is a suspicion or risk that they may have been obtained in that way. Reports from various international bodies may well furnish grounds for suspicion that a country has been in the habit of using torture. That cannot be enough. To trigger the exclusion, it must be shown that the statement in question has been obtained by torture.

I draw support for that general approach from the judgment of the Grand Chamber of the European Court of Human Rights in Mamatkulov and Askarov v Turkey, 4 February 2005. The court had to consider allegations that Turkey had violated article 3 of the Convention by extraditing the applicants to Uzbekistan where political dissidents, such as the applicants, were tortured in prison. In support of their allegations, the applicants "referred to reports by 'international investigative bodies' in the human rights field denouncing both an administrative practice of torture and other forms of ill-treatment of political dissidents, and the Uzbek régime's repressive policy towards dissidents." The Grand Chamber held that, by itself, such generalised information was not sufficient even to establish that there was a real risk that the applicants would be subjected to torture in Uzbekistan. The court said this, at paras 71 - 73 (internal cross-reference omitted):

In fact, there was no further evidence to support the applicants' specific allegations. Rather, the other evidence, led on behalf of Turkey, tended to contradict them and the Grand Chamber was unable to conclude that substantial grounds had existed for believing that the applicants faced a real risk of treatment proscribed by Article 3. If generalised information about a country is not enough to establish that there is a real risk that a given individual will be tortured there in the future, it cannot be sufficient, either, to establish that a given statement has been extracted there by torture in the past.

As my noble and learned friend, Lord Hope of Craighead, has explained, the Hanseatic Oberlandesgericht in Hamburg adopted a somewhat similar approach in El Motassadeq NJW 2005, 2326. There the court was considering whether article 15 of the Convention against Torture prevented it from using summaries of certain witness statements supplied by the United States. Apparently, the witnesses were members of Al Qa'eda, and the suggestion was that the statements had been obtained by torture. The court asked the German government for information, but the relevant government departments were unable to provide any information from the competent American authorities since it had been supplied to them for intelligence purposes only. In that situation, the court could only evaluate the considerable volume of publicly available material suggesting that suspects had been subjected to torture. What the court was looking for was proof that the three witnesses in question had been tortured. The available material referred to only one of them and, while there was quite a lot of general information about the treatment of other suspected Al Qa'eda members, the court noted that none of the information was based on verifiable, named sources. Even taking account of the fact that the United States authorities had prevented the court from having access to more reliable sources, the court concluded that it had not been proved that torture had been used in the examination of the three witnesses, especially having regard to certain exculpatory elements in their statements.

The reasoning of the court, at pp 2329-2330, is instructive. It was under a duty to discover the truth and so the prohibition on the use of evidence had to remain the exception rather than being elevated into the rule. Therefore, the principle "in dubio pro reo" did not apply and the facts justifying the prohibition had to be established to the court's satisfaction. If substantial doubts remained, the possible violation had not been proved and the relevant statement could be used. The court therefore took the view that it was their duty to consider the summaries so as to investigate the facts of the case as fully as possible, but they would take the allegations into account in evaluating the evidence.

In my view the same factors as weighed with the Oberlandesgericht should weigh with the House. Once the House has held that statements obtained by torture must be excluded, the special advocates representing suspects such as the appellants are likely to raise the point whenever information appears to come from a country with a poor record on torture. Special advocates can indeed be expected to ask their clients about possible sources of information against them before they see the closed material. At the hearing the special advocates will present information provided by international organisations or derived from books and articles to paint the picture of conditions in the country concerned. But that cannot be a sufficient basis for SIAC to be satisfied that any particular statement has been obtained by torture. More is required.

Of course, the suspects themselves will not be able to assist the special advocate in finding more information during the closed hearing. But that is not so great a disadvantage as may appear at first sight, since it is in any event unlikely that they would be able to cast light on the specific circumstances in which a particular statement had been taken by the overseas authorities. So, usually at least, any investigation will have to be done by others. On behalf of the Home Secretary, Mr Burnett QC explained how those in the relevant departments who were preparing a case for a SIAC hearing would sift through the material, on the lookout for anything that might suggest that torture had been used. The Home Secretary accepted that he was under a duty to put any such material before the Commission. With the aid of the relevant intelligence services, doubtless as much as possible will be done. And SIAC itself will wish to take an active role in suggesting possible lines of investigation, just as the Hamburg court did.

In the nature of the case and with the best will in the world, there is likely to be a limit to what can be discovered about what went on during an investigation by the authorities in another country. Foreign states can be asked, but cannot be forced, to provide information. How far such requests can be pushed without causing damage to international relations must be a matter for the judgment of the Government and not for SIAC or any court.

When everything possible has been done, it may turn out that the matter is left in doubt and that, using their expertise, SIAC cannot be satisfied on the balance of probabilities that the statement in question has been obtained by torture. If so, in my view, SIAC can look at the statement but should bear its doubtful origins in mind when evaluating it. My noble and learned friend, Lord Bingham of Cornhill, proposes, however, that the statement should be excluded whenever SIAC is unable to conclude that there is not a real risk that the evidence has been obtained by torture. It respectfully appears to me that this would be to replace the true rule, that statements obtained by torture must be excluded, with a significantly different rule, that statements must be excluded unless there is not a real risk that they have been obtained by torture. In effect, the true rule would be inverted. There is no warrant for Lord Bingham's preferred rule in the common law, in article 15 of the Convention against Torture or elsewhere in international law. Moreover, it would run counter to the approach in the two decisions which I have mentioned. The real objection, however, is that, for all the reasons given by the German court, it would be unsound. If adopted, such an approach would ignore the exceptional nature of the exclusion, which requires that the relevant factual basis be established. It would mean that exclusion would be liable to become the rule rather than the exception. It would encourage objections. It would prevent SIAC from relying on statements which were in fact obtained quite properly. It would impede SIAC in its task of discovering the facts that it needs to form its judgment. I would therefore reject that approach and agree with my noble and learned friends, Lord Hope of Craighead and Lord Brown of Eaton-under-Heywood, that SIAC should ask itself whether it is established, by means of such diligent inquiries into the sources as it is practicable to carry out, and on the balance of probabilities, that the information relied on by the Secretary of State was obtained under torture.

My Lords,

The abhorrence felt by civilised nations for the use of torture is amply demonstrated by the material comprehensively set out in the opinion of my noble and learned friend Lord Bingham of Cornhill. While it is regrettably still practised by some states, the condemnation expressed in all of the international instruments to which he has referred is universal. Some of these adjure states to do their utmost to ensure that torture does not take place, while others urge them not to admit in evidence in any proceedings statements obtained by the use of torture.

The objections to the admission of evidence obtained by the use of torture are twofold, based, first, on its inherent unreliability and, secondly, on the morality of giving any countenance to the practice. The unreliability of such evidence is notorious: in most cases one cannot tell whether correct information has been wrung out of the victim of torture - which undoubtedly occurred distressingly often in Gestapo interrogations in occupied territories in the Second World War - or whether, as is frequently suspected, the victim has told the torturers what they want to hear in the hope of relieving his suffering. Reliable testimony of the latter comes from Senator John McCain of Arizona, who when tortured in Vietnam to provide the names of the members of his flight squadron, listed to his interrogators the offensive line of the Green Bay Packers football team, in his own words, "knowing that providing them false information was sufficient to suspend the abuse": Newsweek, November 21, 2005, p 50.

The moral issue arises most acutely when it is established from other evidence that the information obtained under torture appears in fact to be true. Should the legal system admit it in evidence in legal proceedings (where as a matter of law such hearsay evidence may be admitted) or should it refuse on moral grounds to allow it to be used, despite its apparent reliability? On this issue I entirely agree with your Lordships' conclusion that such evidence should not be admitted, reliable or not, even if the price is the loss of the prospect that some pieces of information relevant to the issue of the activities of the person concerned may be given to the tribunal and relied upon by it in reaching its decision.

In so holding I am very conscious of the vital importance in the present state of global terrorism of being able to muster all material information in order to prevent the perpetration of violent acts endangering the lives of our citizens. I agree with the frequently expressed view that this imperative is of extremely high importance. I should emphasise that my conclusion relates only to the process of proof before judicial tribunals such as SIAC and is not intended to affect the very necessary ability of the Secretary of State to use a wide spectrum of material in order to take action to prevent danger to life and property. In the sphere of judicial decision-making there is another imperative of extremely high importance, the duty of states not to give any countenance to the use of torture. Recognising this is in no way to be "soft on torture", a gibe too commonly levelled against those who seek to balance the opposing imperatives.

I have to conclude, in agreement with your Lordships, that the duty not to countenance the use of torture by admission of evidence so obtained in judicial proceedings must be regarded as paramount and that to allow its admission would shock the conscience, abuse or degrade the proceedings and involve the state in moral defilement (Lord Bingham's opinion, para 39). In particular, I would agree with the statement of Mr Alvaro Gil-Robles (cited, ibid, para 35) that

In following this course our state will, as Neuberger LJ observed in the Court of Appeal (para 497), retain the moral high ground which an open democratic society enjoys. It will uphold the values encapsulated in the judgment of the Supreme Court of Israel in Public Committee Against Torture in Israel v Israel (1999) 7 BHRC 31, para 39:

It then has to be considered by what means it may be possible to give effect in our law to this moral imperative. It was argued on behalf of the appellants that it may be done by accepting that the principles of the United Nations Convention Against Torture ("UNCAT") form part of our law, by resort to article 6(1) of the Convention for the Protection of Human Rights and Fundamental Freedoms ("the Convention") or by regarding it as a valid principle of the common law. I do not find it necessary to explore either of the first two avenues, which are not without their difficulties, for I am satisfied that the common law can accommodate the principles involved.

Some of your Lordships have expressed the opinion that the common law as it stands would forbid the reception in evidence of any statement obtained by the use of torture: see the opinions of my noble and learned friends Lord Bingham of Cornhill at para 52 and Lord Hope of Craighead at para 112. This view may well be justified historically, but even if it requires some extension of the common law I am of the clear opinion that the principle can be accommodated. We have long ceased to give credence to the fiction that the common law consists of a number of pre-ordained rules which merely require discovery and judicial enunciation. Two centuries ago Lord Kenyon recognised that in being formed from time to time by the wisdom of man it grew and increased from time to time with the wisdom of mankind: R v Lord Rusby (1800) Pea (2) 189 at 192. Sir Frederick Pollock referred in 1890 in his Oxford Lectures, p 111 to the "freshly growing fabric of the common law" and McCardie J spoke in Prager v Blatspiel, Stamp and Heacock Ltd [1924] 1 KB 566 at 570 of the demand of an expanding society for an expanding common law. Similarly, in the US Supreme Court 121 years ago Matthews J said in Hurtado v California (1884) 110 US 516 at 531 that

As Peter du Ponceau said of the common law (A Dissertation on the Nature and Extent of the Jurisdiction of the Courts, (1824), Preface):

I am satisfied that, whether or not it has ever been affirmatively declared that the common law declines to allow the admission of evidence obtained by the use of torture, it is quite capable now of embracing such a rule. If that is any extension of the existing common law, it is a modest one, a necessary recognition of the conclusions which should be drawn from long established principles. I accordingly agree with your Lordships that such a rule should be declared to represent the common law. It is only right that this should be done in what Tennyson described as

The issue on which I have found it most difficult to reach a satisfactory principled conclusion is that of the approach which SIAC should take to deciding when a statement should be rejected, an issue on which your Lordships have not found it possible to speak with one voice. I have been much exercised by the difficulties inherent in the acceptance of either of the views which have been expressed, but I am conscious of the importance of laying down a clearly defined and workable rule which can be applied by SIAC (or similar bodies which may have to deal with the same problem).

Several possible ways of approaching the issue were mooted in the course of argument. Counsel for the appellants advanced the proposition that once the issue has been raised that a statement may have been obtained by the use of torture the onus should rest upon the Secretary of State to prove beyond reasonable doubt that it was not so obtained. I would unhesitatingly reject this proposition as unsustainable. That is confirmed by experience of inordinately long voir dires in terrorist cases in which the admissibility of confessions has been contested. Not only would the process severely disrupt the course of work in SIAC, it would be wholly impossible for the Secretary of State to obtain the evidence of the parade of witnesses commonly called in such voir dires - gaolers, doctors, interviewers etc - to cover in minute detail the time spent in custody by the maker of the statement. The opposite extreme suggested on behalf of the Secretary of State was that the appellant should have to prove on the balance of probabilities that a challenged statement was obtained by the use of torture before it is rejected. The objections in principle and practice to the imposition of such a burden on an appellant are equally conclusive. He may not even know what material has been adduced before SIAC. The special advocate is given the material, but he has little or no means of investigation and is not permitted to disclose the information to the appellant or his solicitors, so has no one from whom to obtain sufficient instructions.

I agree with your Lordships that consideration of this question by the conventional approach to the burden of proof is both unhelpful and inappropriate. It seems to me rather to equate to the process described by Lord Bingham in R v Lichniak [2002] UKHL 47, [2003] 1 AC 903 at para 16 as "an administrative process requiring [the board] to consider all the available material and form a judgment"; cf Re McClean [2004] NICA 14, para 77, where McCollum LJ said of a similar process that it was "not the establishment of a concrete fact but rather the formulation of an opinion or impression", which was not capable of proof in the manner usually contemplated by the law of evidence. I accordingly agree with the view expressed by Lord Bingham (para 56 of his opinion) and Lord Hope (para 116) that once the appellant has raised in a general way a plausible reason why evidence adduced may have been procured by torture, the onus passes to SIAC to consider the suspicion, investigate it if necessary and so far as practicable and determine by reference to the appropriate test whether the evidence should be admitted and taken into account.

What that test should be is the issue on which your Lordships are divided. Lord Bingham is of the opinion (para 56) that if SIAC is unable to conclude that there is not a real risk that the evidence has been obtained by torture, it should refuse to admit it. Lord Hope, on the other hand, has propounded a different test, which he describes as putting the question which SIAC has to decide positively rather than negatively. It has to be established on the balance of probabilities that the particular piece of evidence was obtained by the use of torture; and unless it has in SIAC's judgment been so established, after it has completed any investigation carried out and weighed up the material before it, then it must not reject it on that ground.

I have found the choice between these tests the most difficult part of this case. Lord Bingham has cogently described the difficulties facing an appellant before SIAC and the potential injustice which he sees as the consequence if the Hope test is adopted. Lord Hope for his part places some emphasis on the severity of the practical problems which would face SIAC in negativing the use of torture to obtain any given statement, and expresses his concern that it would constitute "an insuperable barrier for those who are doing their honest best to protect us". In support of his view Lord Hope points in particular to the terms of article 15 of UNCAT, which requires states to ensure that any statement "which is established to have been made as a result of torture" shall not be invoked in any proceedings.

After initially favouring the Bingham test, I have been persuaded that the Hope test should be adopted by SIAC in determining whether statements should be admitted when it is claimed that they may have been obtained by the use of torture. Those who oppose the latter test have raised the spectre of the widespread admission of statements coming from countries where it is notorious that torture is regularly practised. This possibility must of course give concern to any civilised person. It may well be, however, that the two tests will produce a different result in only a relatively small number of cases if the members of SIAC use their considerable experience and their discernment wisely in scrutinising the provenance of statements propounded, as I am confident they will. Moreover, as my noble and learned friend Lord Brown of Eaton-under-Heywood points out in para 166 of his opinion, intelligence is commonly made up of pieces of material from a large number of sources, with the consequence that the rejection of one or some pieces will not necessarily be conclusive. While I fully appreciate the force of the considerations advanced by Lord Bingham in paras 58 and 59 of his opinion, I feel compelled to agree with Lord Hope's view in para 118 that the test which he proposes would, as well as involving fewer practical problems, strike a better balance in the way he there sets out.

On this basis I would accordingly allow the appeals and make the order proposed.

My Lords,

Torture is an unqualified evil. It can never be justified. Rather it must always be punished. So much is not in doubt. It is proclaimed by the Convention against Torture and many other international instruments and now too by section134 of the Criminal Justice Act 1988. But torture may on occasion yield up information capable of saving lives, perhaps many lives, and the question then inescapably arises: what use can be made of this information? Unswerving logic might suggest that no use whatever should be made of it: a revulsion against torture and an anxiety to discourage rather than condone it perhaps dictate that it be ignored: the ticking bomb must be allowed to tick on. But there are powerful countervailing arguments too: torture cannot be undone and the greater public good thus lies in making some use at least of the information obtained, whether to avert public danger or to bring the guilty to justice.

Several of your Lordships have remarked on the tensions in play and have noted the balances struck by the law, different balances according to whether one is focusing on the executive or the judicial arm of the state. Essentially it comes to this. Two types of information are involved: first, the actual statement extracted from the detainee under torture ("the coerced statement"); second, the further information to which the coerced statement, if followed up, may lead ("the fruit of the poisoned tree" as it is sometimes called). Generally speaking it is accepted that the executive may make use of all information it acquires: both coerced statements and whatever fruits they are found to bear. Not merely, indeed, is the executive entitled to make use of this information; to my mind it is bound to do so. It has a prime responsibility to safeguard the security of the state and would be failing in its duty if it ignores whatever it may learn or fails to follow it up. Of course it must do nothing to promote torture. It must not enlist torturers to its aid (rendition being perhaps the most extreme example of this). But nor need it sever relations even with those states whose interrogation practices are of most concern. So far as the courts are concerned, however, the position is different. Generally speaking the court will shut its face against the admission in evidence of any coerced statement (that of a third party is, of course, in any event inadmissible as hearsay); it will, however, admit in evidence the fruit of the poisoned tree. The balance struck here ("a pragmatic compromise" as my noble and learned friend Lord Bingham of Cornhill describes it at para 16 of his opinion) appears plainly from section 76 of the Police and Criminal Evidence Act 1984. There is, moreover, this too to be said: whereas coerced statements may be intrinsically unreliable, the fruits they yield will have independent evidential value.

All this is entirely understandable. As several of your Lordships have observed, the functions and responsibilities of the executive and the judiciary are entirely different, a difference reflected indeed in article 15 of the Torture Convention itself. Article 15's concern is with the use of "any statement . . . made as a result of torture . . . as evidence in any proceedings". It creates no bar to the use of coerced statements as a basis for executive action. And, of course, it says nothing whatever about the fruits of the poisoned tree.

None of this is contentious. The dispute arising on these appeals concerns only a single, comparatively narrow issue: the use of certain coerced statements on appeals before the Special Immigration Appeals Commission (SIAC) under section 25 of the Anti-terrorism, Crime and Security Act 2001 (the 2001 Act). The statements in question are those made by detainees abroad, coerced by the authorities of a foreign state without the complicity of any British official. It is the Crown's case that strictly speaking these are admissible in evidence before SIAC, a tribunal charged not with adjudicating upon the appellant's guilt but only with deciding whether reasonable grounds exist for suspecting him to be an international terrorist and for believing his presence here to be a risk to national security.

In common with the other members of this Committee and essentially for the reasons they give, I too would reject the Crown's contention. In question here is not the power of the executive but rather the integrity of the judicial process. SIAC is a court of law (indeed a superior court of record). And as was pointed out in M v Secretary of State for the Home Department [2004] 2 All ER 863, SIAC's function on an appeal under section 25 is not to review the exercise by the Secretary of State of his power of certification under section 21, but rather to decide for itself whether, at the time of the hearing, there are "reasonable grounds" for the suspicion and belief required under section 21. True it is that the statements in question are sought to be relied upon not to convict the appellant of any offence but rather to found such suspicion and belief as would justify his continued detention under section 23. It is difficult to see, however, why this consideration should strengthen rather than weaken the Crown's argument: no court will readily lend itself to indefinite detention without charge, let alone trial. (Parliament, indeed, has recently demonstrated its own unease in this area by refusing to legislate for up to 90 days detention of arrested terrorist suspects prior to charge.) At all events, for the detention to continue under the 2001 Act, Parliament required that SIAC must independently sanction this deprivation of liberty.

In short, I would hold that SIAC could never properly uphold a section 23 detention order where the sole or decisive evidence supporting it is a statement established to have been coerced by the use of torture. To hold otherwise would be, as several of your Lordships have observed, to bring British justice into disrepute. And this is so notwithstanding that the appellant was properly certified and detained by the Secretary of State in the interests of national security, notwithstanding that the legislation (now, of course, repealed) allowed the appellant's continuing detention solely on the ground of suspicion and belief, notwithstanding that the incriminating coerced statement was made not by the appellant himself but by some third party, and notwithstanding that it was made abroad and without the complicity of any British official.

To what extent, it is perhaps worth asking, does such a ruling impede the executive in its vitally important task of safeguarding the country so far as possible against terrorism? To my mind to a very limited extent indeed. In the first place it is noteworthy that the ruling will merely substitute an exclusionary rule of evidence for the Secretary of State's own publicly stated policy not in any event to rely on evidence which he knows or believes to have been obtained by torture abroad. Secondly, the intelligence case against the suspect would, we are told, ordinarily consist of material from a large number of sources—a "mosaic" or "jigsaw" of information as it has been called; it is most unlikely that the sole or decisive evidence will be a coerced statement. It follows, therefore, that the possibility of a detention order under section 23 being discharged on a section 25 appeal to SIAC because of the rejection of a coerced statement is comparatively remote. And certainly there is nothing in SIAC's open determination in relation to E's appeal (the first in which Mr Emmerson QC submitted that information extracted by torture should be excluded by rule of law rather than merely afforded less weight) to suggest the contrary:

But theoretically it could happen and in that event, it is suggested, the Secretary of State would be disadvantaged in two distinct ways. Most obviously, perhaps, he would be unable to continue to detain someone whose detention he judged necessary on grounds of national security. To the straightforward response "so be it, the rule of law so requires", I would add this. There is a certain unreality in discussing the discharge of detention orders as the legislation now stands. The power to detain suspected international terrorists under section 23 of the 2001 Act is now a matter of history. In December 2004 your Lordships in A v Secretary of State for the Home Department [2005] 2 AC 68, declared section 23 to be incompatible with articles 5 and 14 of the European Convention on Human Rights and with effect from 14 March 2005 the whole of Part 4 of the Act was repealed by section 16 of the Prevention of Terrorism Act 2005 (save only with regard to extant appeal proceedings, preserved by section 16(4) of the 2005 Act).

No doubt the effects of your Lordships' judgment will spill over into other court proceedings designed to provide a judicial check on the exercise of other executive powers to place constraints of one sort or another on terrorist suspects in the interests of national security—most notably appeals to SIAC under section 2 of the Special Immigration Appeals Commission Act 1997 against deportation orders, and statutory applications to the Administrative Court challenging control orders under the Prevention of Terrorism Act 2005. For the reasons already given, however, it seems unlikely that the exclusionary rule concerning coerced statements, even assuming that it applies equally in these related contexts (which was not the subject of specific argument before us) will affect many, if any, individual cases.

The other way in which it has been suggested that the Secretary of State may be disadvantaged by your Lordships' ruling is in the event that he has to defend himself against a civil claim, for example for false imprisonment. With regard to this possibility I find myself in strong agreement with the view expressed by Lord Nicholls of Birkenhead in para 72 of his opinion: it would make no sense to allow (indeed encourage) the Secretary of State to make use of all information available to him in deciding how to exercise his executive power in the public interest and then prohibit his reliance upon part of that information (coerced statements) when faced with a claim for false imprisonment. Rather he should be permitted to refer to such statements, not of course, in reliance upon their truth, but merely to explain his state of mind at the time he took the action impugned.

Perhaps, however, a better answer to this particular difficulty is after all to be found in section 21(9) of the 2001 Act (although no argument was in fact addressed upon it):

A comparable provision with regard to control orders is, one notes, to be found in section 11(1) of the 2005 Act.

It follows from all this that your Lordships' decision on these appeals should not be seen as a significant setback to the Secretary of State's necessary efforts to combat terrorism. Rather it confirms the right of the executive to act on whatever information it may receive from around the world, while at the same time preserving the integrity of the judicial process and vindicating the good name of British justice.

I turn finally to the burden of proof. I agree with Lord Hope of Craighead (at para 121 of his opinion) that SIAC should ask itself whether it is "established, by means of such diligent inquiries into the sources that it is practicable to carry out and on a balance of probabilities, that the information relied on by the Secretary of State was obtained under torture." Only if this is established is the statement inadmissible. If, having regard to the evidence of a particular state's general practices and its own inquiries, SIAC were to conclude that there is no more than a possibility that the statement was obtained by torture, then in my judgment this would not have been established and the statement would be admissible.

The difficulty I have with the "real risk" test espoused by certain of your Lordships, apart from the fact that classically such a test addresses future dangers (as, for example, the risk of torture or other article 3 ill-treatment which the European Court of Human Rights in Soering v United Kingdom (1989) 11 EHRR 439 understandably refused to countenance) rather than past uncertainties, is that it would require SIAC to ignore entirely (rather than merely discount to whatever extent it thought appropriate) any statement not proved to have been made voluntarily. That, at least, is how I understand the "real risk" test to apply: if SIAC were left in any substantial (ie other than minimal) doubt as to whether torture had been used, the statement would be shut out, however reliable it appeared to be and notwithstanding that SIAC concluded that it had probably been made voluntarily. That seems to me a surprising and unsatisfactory test. If I have misunderstood the proposed test and if all that it involves is SIAC shutting out a statement whenever they simply cannot decide one way or the other on the balance of probabilities whether it has been extracted by torture (a rare case one would suppose given the expertise of the tribunal) then my difficulty would be substantially lessened although I would still prefer the test favoured by Lord Hope of Craighead and Lord Rodger of Earlsferry.

It is one thing to say, as in Soering, that someone cannot be deported whilst there exists the possibility that he may be tortured—or, indeed, as the dissentient minority said in Mamatkulov and Askarov v Turkey (Application Nos 46827/99 and 46951/99, unreported, 4 February 2005), if they run a real risk of suffering a flagrant denial of justice— quite another to say that the integrity of the court's processes and the good name of British justice requires that evidence be shut out whenever it cannot be positively proved to have been given voluntarily.

For these reasons, and for the reasons given by Lord Bingham and others of my noble and learned friends, I too would allow these appeals and make the order proposed.