EHRM, 18-12-2018, nr. 36658/05

The applicant was born in 1983 and currently lives in Paris.

In September 2003 she arrived in Moscow from Chechnya and started working at an insurance company. In October 2003 she went to a mosque where she met V. and Ku., two young Russian women who had converted to Islam.

The applicant complained under Article 6 §§ 1 and 3 (b) of the Convention that the overall fairness of the criminal proceedings against her had been undermined because she had not been able to see effectively a secret surveillance videotape shown in the courtroom. Article 6 of the Convention in the relevant part reads as follows:

The Government contested that argument.

The applicant complained under Article 6 §§ 1 and 3 (d) of the Convention that the overall fairness of the criminal proceedings against her had been undermined because she had not been able to call and examine witness A. at the trial. The relevant part of Article 6 of the Convention reads as follows:

The Government contested that argument.

The applicant complained under Article 6 §§ 1 and 3 (d) of the Convention that the overall fairness of the criminal proceedings against her had been undermined because she had not been able to call and examine at the trial the two attesting witnesses B. and K.

The Government contested that argument.

The right to call witnesses on the defendant's behalf is one of the core elements of the fair trial requirements enshrined by Article 6 as a whole. In its case-law, our Court has emphasised that the right to a fair trial holds so prominent a place in a democratic society that there can be no justification for interpreting Article 6 § 1 of the Convention restrictively.1. This is even truer in respect of the rights guaranteed by Article 6 § 3, since they are expressly qualified as ‘minimum rights’. In addition to these rights, the Court has read into Article 6 § 1 several other rights that are considered to be fundamental to the concept of a fair trial, notably the right to present one's case to the court and to participate effectively at the hearing,2. equality of arms and the privilege against self-incrimination3.. Consequently, the right under Article 6 § 3 (d) should neither be interpreted in a restrictive way nor detached from the other guarantees of a fair trial.

The Government argue that the defence waived their right to have witness A. questioned at the trial. It is true that the applicant and one of her lawyers stated that they had no objection to the reading out of A.'s testimony, and the second defence lawyer agreed to this. However, they did so after having been informed by the presiding judge that A. was on a work-related mission and could not therefore appear in court. No source for this information, provided to the trial court, has been adduced before this Court. On the day that the trial court decided not to hear witness A. in person, the court adjourned the hearing to allow the parties to prepare their closing submissions. Consequently, the need to ensure the expeditiousness of the proceedings did not serve as a justification for refusing to allow witness A. to be examined at the trial. In these circumstances and in the absence of any direct reference to a renouncement of the right to examine witness A., the manner in which the defence expressed their position could not be construed as an unequivocal waiver of the right under Article 6 § 3 (d) of the Convention.27.

Indeed, the applicant's arguments on appeal that the trial judge had not decided on the motion to examine A. and had provided no evidence of A.'s absence on a work-related mission clearly indicate that the defence had not in fact waived their right to examine A.28. Nor did the Supreme Court, sitting as an appeal court, explicitly consider the right to have been waived. The Supreme Court referred in its judgment not only to the defence's consent to the reading out of A.'s statements, but also to the reasons for his absence from the trial.29.

Furthermore, the majority do not consider that the case raised any questions of public interest which would have prevented the relevant procedural guarantees from being waived.30. I observe that one of the major points of contention between the parties is whether the applicant had advanced an entrapment plea in the domestic proceedings. This point weighs heavily in the Court's examination of the applicant's complaint concerning the failure to summon and hear witness A. In fact, the applicant's lawyer U. clearly stated during the exchange of closing arguments that ‘this whole criminal case is a set-up against Murtazaliyeva by law-enforcement agents’.31. This statement must be construed as a properly raised entrapment plea when viewed against the background of all the available case material, which demonstrates that the applicant's strategy and line of defence before the domestic courts was essentially to protest her innocence and to insist in particular on her claim that the police had planted the plastic explosives in her handbag. In this respect it should be emphasised that, contrary to the Government's argument,32. the denial of guilt does not logically preclude reliance on police incitement. Furthermore, according to the nemo tenetur principle, a defendant must not be forced to choose between a defence of being innocent and an entrapment defence. An accused may simultaneously argue that he or she did not commit the imputed offence (because the offence was never consummated or because it was committed by another person) and that he or she was entrapped into committing it.33.

Since a plea of entrapment was made and there was certain prima facie evidence of entrapment, the judicial authorities should have examined the related facts and take the necessary steps to uncover the truth, in order to determine whether there was any incitement. Taking into account the clear indicia of police entrapment and the pivotal role of witness A. in the police strategy, the trial court was under an obligation to question him. In view of the above, I cannot follow the majority in finding the Government's objection founded.

The Grand Chamber majority essentially rely on an allegedly two-tiered test derived from Perna34., requiring, firstly, a substantiated request by the defendant indicating the relevance of the testimony for ‘the establishment of the truth’ and, secondly, evidence that the refusal to call the given witness undermined the overall fairness of the proceedings. As a result, the legal standards developed by the Court in Al-Khawaja and Tahery [GC]35. and Schatschaschwili [GC]36. for the non-examination of prosecution witnesses during a trial are not applicable to the case at hand.37.

Although Article 6 § 3 (d) sets out the right of everyone charged with a criminal offence to examine ‘witnesses on his behalf under the same conditions as witnesses against him’, the classification of an individual as a witness for the defence or the prosecution leads to a different set of legal standards in the practice of the Court. Such double standards are very problematic in view of the clear stipulation in the Convention that both kinds of witnesses are to be examined ‘under the same conditions’.38.

In addition, it is highly questionable whether these double standards can be maintained in view of the multiple sets of circumstances in which a witness may testify or be asked to testify during trial. First, the classification of a witness as being presented ‘on behalf of’ or ‘against’ a defendant is not a straightforward exercise when a witness put forward by the prosecution provides testimony favourable to the defendant and, vice versa, when a witness requested by the defendant provides testimony against him or her. Second, the attendance of a witness may be requested by both the defendant and the prosecutor, in which case neither the Al-Khawaja and Tahery and Schatschaschwili criteria, nor the Perna criteria, would strictly speaking apply. Third, a witness called upon ex officio by the court itself would not fit into this dual classification. Fourth, a witness who has been questioned by the prosecution during the pre-trial stage and whose testimony at the trial is requested by the defence may refuse or may be unable to testify at trial.39. Fifth, a forensic witness is normally an independent witness, called upon to produce testimony that is neither ‘on behalf of’ nor against the defendant.

In view of this variety of circumstances, it has been the Court's case-law that the term ‘witness’ has an autonomous meaning in the Convention system, regardless of classifications under national law,40. and that a witness is considered to be a prosecution witness if his or her depositions serve ‘to a material degree’ as the basis for a conviction.41. In any individual case, it is the question whether it was the prosecution or the defence which sought the attendance and examination of a certain witness which provides a starting point for the Court's assessment. However, this is not decisive for the Court. The Court's assessment will depend on the high probative value (‘to a material degree’) of a witness's statements in the establishment of the accused's guilt by the domestic courts. This criterion for the qualification of witnesses, which takes into account the findings on the merits of the case, raises significant problems for domestic courts, in view of its ex post facto and imprecise nature. For these courts, the test to assess whether a witness should be summoned to trial and heard and cross-examined must be a precise, ex ante test, grounded in the chronological point at which the relevant court is called to assess the request that it take such evidence.42.

The circumstances of the present case are telling in this regard. I note that, despite A. having initially been listed in the bill of indictment as one of thirteen prosecution witnesses as well as a defence witness, he was not called by the prosecution; nonetheless, his testimony served as the basis for the applicant's conviction, along with other evidence. A.'s pre-trial statements as read out at the trial and referred to in the domestic courts' judgments contained factual assertions, but the courts did not cross-reference A.'s statements with those of other witnesses. In spite of this, A. was treated as a defence witness by the domestic courts, since it was the defence which had requested that he be questioned, at both the pre-trial and trial stages of the proceedings.

In addition, witnesses B. and K. were adduced by the prosecution, but it was the defence which wished to rely on their testimony.43.

To sum up this introductory point, the Court's double standards for the treatment of witnesses on behalf of and against a defendant, and the ex post facto and imprecise concept of prosecution witness on which they are based, contradict the letter and the spirit of the Convention, create unnecessary practical difficulties for the domestic courts and cause unneeded uncertainty in a field of law where clarity and precision are of the utmost importance.

The Perna approach was based on a clear-cut distinction between defence and prosecution witnesses, made on the sole basis of which party had requested that they be examined before the trial court and irrespective of the substance of the testimony to be given by each individual witness.

The original Perna test established a strict substantive criterion for assessing a defence request to produce testimonial evidence, based on the obligation to ‘support his request by explaining why it is important for the witnesses concerned to be heard and their evidence must be necessary for the establishment of the truth’.44. Contrary to what the majority claim,45. the Perna judgment did not require an overriding assessment of ‘overall fairness’, since fairness was only assessed from the point of view of Article 6 § 1 (e) of the Convention and on account of the way the evidence was taken. It is important to recall the relevant passage of the Perna judgment:

Hence, the overall fairness of the proceedings is not, in the light of both the letter and the spirit of the Perna judgment, an overriding criterion under which the applicant has to demonstrate that the refusal to call the witness prejudiced the overall fairness of the proceedings.47. Nevertheless, the Perna test was too strict and therefore unfair for the defence. It needed to be revisited from a human-rights-friendly perspective, namely a revisiting that would bring it closer to the Convention principles on evidence-taking in criminal procedure, but this has not been done by the majority, as will be demonstrated below.

The original Perna test was formulated in strict terms: the criterion established for the admission of defence witnesses was that of evidence ‘necessary’ for the establishment of the truth. The Court's practice loosened such strictness, turning to the more liberal criterion of ‘prima facie relevance’ of the evidence, with multiple formulations being given over the years.48. Between the two diverging criteria, the majority have now chosen that of witnesses ‘capable of influencing the outcome of a trial’ or ‘who can reasonably be expected to strengthen the position of the defence’.49. At first sight, it seems that the majority have chosen the more relaxed, liberal, criterion. This is also indicated by the majority's further position to the effect that scant reasoning by the defence may be sufficient to answer the first question of the test in the affirmative.50. This choice would also be in line with the case-law of the ICTY Appeals Chamber, according to which an applicant must demonstrate a reasonable basis for his belief that there is a good chance that the prospective witness will be able to give information which will materially assist him in his case, in relation to clearly identified issues relevant to the forthcoming trial.51.

The apparently liberal majority perspective is not limited to the degree of relevance or necessity of the evidence, but also to its scope, which includes evidence that is both pertinent to the outcome of the trial and to the ‘position of the defence’ in general. This has an obvious consequence. Any defence-witness testimony related to the lawfulness, reliability and credibility of the remaining evidence, as well as to the lawfulness of the public authorities' conduct in the case, including that of the judges, prosecutors and law-enforcement authorities who acted or should have acted in the defendant's case, may strengthen the position of the defence, even if it does not ultimately influence the outcome of the trial. For example, it is mandatory to call a witness who, it is claimed, might have supported an accused's allegation that he was ill-treated in order to make him confess to an offence, even if the impugned confession did not play a crucial role in establishing the applicant's guilt.52. This is for two reasons: first, because the evidence of torture or ill-treatment may impact on the position of the defence, since an issue may arise under Article 6 § 1 in respect of evidence obtained in violation of Article 3 of the Convention, even if the admission of such evidence was not decisive or crucial in securing the conviction;53. and second, because the applicable criterion for the assessment of the request for calling defence witnesses, as well as prosecution witnesses, should be an ex ante criterion, independent of the findings of the trial court.

In effect, the majority apply the first criterion in a very illiberal way, since they criticise the defence's motion for the attendance of witnesses B. and K. on the basis that ‘it did not contain any particular factual or legal arguments and did not elaborate in concrete terms on how their testimony could reasonably be expected to strengthen the case for the defence’.54. I will return to this point later.

The issue of whether the domestic courts considered the relevance of a given individual's testimony and provided sufficient reasons for their decision not to examine a witness at trial is not new.55. In fact, it was already a constituent part of the original Perna test, since the Court in that case did assess the motivation of the domestic courts and concluded that ‘the decisions in which the national authorities refused the applicant's requests are not open to criticism under Article 6’.56. In Perna, the reason for rejection of the witness was the trial court's legal finding that, even if proven to be true, the facts were irrelevant from a legal perspective.

In the present judgment, the majority add that the reasoning of the domestic courts must be ‘commensurate’, i.e. adequate in terms of scope and level of detail, with the reasons advanced by the defence.57. This deceptively simple idea does not suffice in a criminal procedure subject to the principle of legality and in a State governed by the rule of law.

As a matter of principle, the trial court cannot refuse a defence witness without good or sufficient reason,58. which is also the minimum criterion for refusal to call or to cross-examine prosecution witnesses, as set out in Al-Khawaja and Tahery and Schatschaschwili.59. Defence witnesses cannot be treated differently from prosecution witnesses, as established by the paramount principle of equality of arms.60. A good or sufficient reason must exist for the non-attendance of defence witness.61.

Hence, the dismissal of a request without giving reasons or silence by the domestic courts in respect of a sufficiently reasoned and relevant request to call a defence witness necessarily leads to a finding of a violation of Article 6.62. The reasons should relate to the specific circumstances of each witness's situation.63. The fact that a defence witness is incapable of proving the accused's innocence is not a good reason to refuse his or her examination64., for the simple reason that the defendant does not have to prove his or her innocence. It suffices that the defendant argues that the witness is in a position to raise doubts about the facts of the accusation.65. When requesting the attendance of a defence witness, the defence may have other interests than acquittal, such as obtaining a conviction for a lighter offence than that with which the defendant was initially charged, establishing mitigating circumstances, casting doubt on the credibility of the prosecution case, etc.

Most importantly, the trial court must not only give good or sufficient reasons for its decision to refuse a defence witness, but it is not allowed to anticipate the result of the examination of the evidence under question (Verbot der Beweisantizipation) — except in favour of the defendant. When deciding on whether or not to obtain the evidence requested by the defence, the court has to presume that the examination of this evidence will turn out in the defendant's favour. From this perspective, the court may decide not to call the defence witness if

In the latter case the trial court is doubly limited. A trial court cannot invoke other evidence to argue that the attendance of a defence witness is not significant for its determination of the case. A reference by the trial court to other evidence, to the effect that a witness could not have supplied new, important or significant information, is never a good reason for not calling him or her.67. This would be a gross violation of the prohibition on anticipation of the outcome of examination of evidence. Furthermore, after refusing to allow a defence witness to attend and be examined on the basis that the fact that would be proven by such evidence is insignificant for determination of the case, the trial court must subsequently stick to this undertaking in its decision on the merits and may not base any argument on this particular fact. 68.

In sum, the second Murtazaliyeva criterion is still very far from being complete, and offers a standard of protection lower than that accorded to the right to present prosecution witnesses. Indeed, the majority wrongly insist that only in exceptional circumstances will it conclude that a refusal to hear a defence witness is in contravention of Article 6.69. There is no reason for such disparate treatment. Article 6 §§ 1 and 3 (d) of the Convention contain a prohibition on the use of hearsay evidence against a defendant in criminal proceedings, but exclusion of the use of hearsay evidence is also justified when that evidence may be considered to assist the defence.70. In other words, Article 6 §§ 1 and 3 (d) not only enshrine the defendant's right to confront prosecution witnesses and question defence witnesses, but guarantees the principle of immediacy,71. which requires that the trial court be able to observe the witnesses' demeanour and thus form its own impression of their reliability.72.

According to the majority, the Court has to assess whether the trial court implemented safeguards to secure that, even in the event of an obvious deviation from the regulations set out by Article 6 § 3 of the Convention, overall fairness was maintained, and they suggest that such a safeguard would prevent the Perna test from becoming mechanical.73. Let me be clear: this is where the core and the novelty of the present case lie, and not in the alleged second criterion, as the majority claim.74.

If the overall fairness of the proceedings under Article 6 § 1 is accepted as a corrective to evaluation of compliance with the provisions of Article 6 § 3, then it necessarily has to be understood as an additional guarantee to the individual guarantees set out in Article 6 § 3. These are connected to and derive from the fairness principle in Article 6 § 1.75. However, the overall fairness test should not be understood as an unconnected but overriding substitute for the guarantees set out in Article 6 § 3, capable of correcting violations of the rights described in that paragraph.76.

The wishy-washiness that accompanies the majority's ambiguous approach towards an overriding test of overall fairness fully divests Article 6 § 3 (d) of its meaning. The majority's revisited Perna test is intentionally constructed to allow the Court, even if the answers to the first two questions are considered as strong indication for a violation, to rule otherwise. The clear provision of minimum rights as an expression of the fairness principle is sacrificed for the sake of preserving a margin of appreciation for domestic courts. Although the majority note the possibility of reaching conclusions that differ, in either direction, from the indications provided by the two prior steps, the undermining of a minimum right for the defence will, in consequence, more frequently lead to an undue burden for the defence.

It is true that the Court, like the European Commission of Human Rights before it, has regularly assessed the overall fairness of the procedure or the fairness of the proceedings taken as a whole when judging compliance with Article 6 § 1 of the Convention. However, until Dvorski,77. this exercise was conducted in order to assess the fairness of procedures which otherwise did comply with the minimum rights set forth in Article 6 § 3 (c) of the Convention. Under the Court's traditional approach, a procedure could be unfair overall even though it had complied with Article 6 § 3. To derive, a contrario sensu, that a procedure can be fair overall even though it does not comply with Article 6 § 3 — or, to put it otherwise, that the overall fairness of the procedure can allow Article 6 § 3 violations to be disregarded — is not only logically erroneous, it also denaturalises the very essence of Article 6 § 3. A thorough review of the Court's case-law will make this point crystal clear.

Nielsen v. Denmark was the first time in Strasbourg history that the test was used. This was a case in which the Commission dealt with a claim concerning the allegedly erroneous admission of medical evidence in a criminal trial. In it, the Commission stated:

As if this were not clear enough, the Commission considered it helpful to clarify that:

This asymmetry between Articles 6 § 1 and 6 § 3 was reiterated by the Court in Deweer v. Belgium, where the Court deemed that, while ‘the various rights of which a non-exhaustive list appears in [Article 6 § 3] are constituent elements, amongst others, of the notion of a fair trial in criminal proceedings’, since the applicant in that case ‘was totally deprived of such a trial… the question whether [Articles 6 § 2 and 6 § 3] were observed has no real significance in this regard: it is entirely absorbed by the question of whether [Article 6 § 1] was complied with’.80. Again: a process can be unfair even if it complies perfectly with Article 6 § 3, but nothing here suggests that a process can be fair if those minimum rights are disrespected.

Since then, the Court has applied the overall fairness of the proceeding for a multiplicity of purposes, but the relationship between paragraph 1 and paragraph 3 of Article 6 as expressed in Nielsen seems to have formed the bottom-line of the Court's reasoning — until Dvorski. A milestone in this jurisprudence is the oft-cited judgment in Barberà, Messegué and Jabardo,81. in which the Court concluded that an accumulation of procedural defects (none of which in itself amounted to a violation of one of the rights explicitly enumerated in Articles 6 § 2 and 6 § 3) meant that ‘the proceedings in question, taken as a whole, did not satisfy the requirements of a fair and public hearing’.82.

In this light, the overall fairness analysis gradually became ubiquitous, even when dealing with claims under Article 6 § 3. In some instances, for example, the Court analysed cases concerning Article 6 § 3 in the context of the overall assessment of the proceedings. In Goddi,83. for example, the Court stated that the guarantees of Article 6 § 3 are ‘constituent elements’ of ‘the general notion of a fair trial stated in paragraph 1’ and that therefore ‘the Court has examined separately each limb of the complaint and then made an overall assessment’.84. In other cases, the Court has stressed the need to interpret paragraph 3 of Article 6 in the light of the overall purpose of the norm — that is, to secure a fair procedure. This is revealed by a quote the Court has often used:

However, as a reading of the case in which these sentences were formulated reveals, this reasoning had quite the opposite intention to that of the majority in the present judgment. The Court in Artico v. Italy, when drafting these sentences,86. seems to have meant the exact opposite. In Artico, the Court was ensuring that the legal assistance that the Government claimed to have provided to the applicant was indeed effective for the purposes of Article 6, and not an ‘illusory’ right. Indeed, the Court concluded in that judgment that it was not effective: despite the appointment of counsel, the applicant had not received any substantive legal aid.87.

As a matter of tradition, the most prominent area in which the overall fairness test has been conducted is the realm of legal assistance. Under the guise of assessing the overall fairness of the proceedings, the Court has distanced itself from its previous defendant-friendly standards in favour of a more prosecution-leaning stance. Until Dvorski, the Court always dealt with Article 6 § 3(c) of the Convention in a manner consistent with the principles described above. It is true that in certain cases the Court engaged in a sort of balancing exercise between the reasons that justified certain restrictions to freely chosen legal assistance and their implications for the overall fairness of the procedure viewed in its entirety, but the Court did so only after acknowledging that such reasons existed. A milestone in this line of jurisprudence is John Murray,88. in which the Court stated:

A central difference comes immediately to mind: since the right to legal counsel can only be restricted ‘for good cause’, the Court's holistic assessment in John Murray comes only after such reasons have been demonstrated, since ‘even a lawfully exercised power of restriction is capable of depriving an accused, in certain circumstances, of a fair procedure’90.. The Court, moreover, proved extremely cautious in its assessment of the overall fairness in this case, ultimately finding a violation: ‘it is not for the Court to speculate on what the applicant's reaction, or his lawyer's advice, would have been had access not been denied during this initial period’.91.

The Court refined and developed its stance on the invalidity of unmotivated restrictions to the right to a legal counsel in Salduz [GC].92. The applicant in this memorable Grand Chamber case had been denied legal assistance during police custody, during which time he made statements when interrogated which he subsequently retracted before the investigating judge and the prosecutor. The Court reiterated its reasoning from John Murray, stating:

It is undeniable from the underlined clause that the overall fairness assessment can only proceed once the restriction has been found to be justified. The Court itself has made clear, over and over, that any restriction to the right of legal counsel must be justified in order for it to analyse its impact on the overall fairness of the proceedings:

Consistent with this reasoning, when applying the general principles to the case at hand, the Court observed:

John Murray and Salduz, therefore, make it clear that States cannot legitimately restrict access to legal assistance without valid reasons. As the above quotations show, it is indeed difficult to think of any other interpretation of the ratio decidendi of those cases. As late as 2010, the Court applied this reasoning, concluding that the absence of valid reasons for restricting an applicant's access to a lawyer was enough to find an Article 6 violation:

An equally straightforward statement was made by the Court in Pishchalnikov v. Russia, a case concerning confessions made during pre-trial proceedings after assistance of counsel had been denied.97. In this case, the Court analysed first whether ‘the applicant's access to counsel was restricted’ and whether this ‘restriction of defence rights was justified’. Immediately after conducting this dual assessment, the Court noted:

Dvorski v. Croatia concerned a ‘denial of choice’ of legal counsel, rather than an outright denial of access to it. Without citing any Court precedent in support of its finding, the majority stated:

This was the first time that the Court reversed the reasoning of John Murray and Salduz. However, two qualifications are in order. First, unlike the preceding cases, which dealt with ‘denial of access’ to a lawyer, Dvorski dealt with a ‘denial of choice’, which the Court considered a ‘less serious’ matter.100. Second, and more importantly, the assessment of the overall fairness of the procedure seems to have been conducted on the basis of a serious presumption that the applicant's initial confession, made in the absence of a freely chosen lawyer, had a ‘significant likely impact’ on the development of the criminal proceedings.101. The nature of this ‘likelihood’ is not clear, but, without doing any violence to the text, it could be read as an acknowledgement that every denial of choice of a lawyer at the critical stages of criminal proceedings could have an impact that jeopardises the fairness of the proceedings as a whole.102. This reading of Dvorski would be consistent with Salduz, although uncomfortably so.

It is only in Ibrahim and Others v. the United Kingdom that, while claiming to be ‘clarifying’ Salduz,103. the Grand Chamber in fact bluntly departs from the previous case-law on Article 6 § 3 (c) of the Convention. The Court states that the ‘absence of compelling reasons (for restricting an accused's right to counsel) does not … lead in itself to a finding of a violation of Article 6’.104. Indeed, as in John Murray and Salduz, when ‘compelling reasons are found to have been established, a holistic assessment of the entirety of the proceedings must be conducted to determine whether they were ‘fair’ for the purposes of Article 6 § 1’.105. However, when ‘there are no compelling reasons for restricting access to legal advice, the Court must apply a very strict scrutiny to its fairness assessment’.106. Interestingly, the Court does not present this as an innovation, but as a correct reading of Salduz.107. As the separate opinion of Judges Sajó and Laffranque in Ibrahim and Others108. abundantly demonstrates, this is hard to maintain.

The preceding paragraphs show that, contrary to the Court's assertions, the reasoning in Ibrahim and Others contradicts a well-established case-law that ended in Dvorski. Although claiming that Ibrahim and Others was ‘clarifying’ or ‘fleshing out’ the Salduz criteria, it was in fact abandoning that case-law and, with it, the rationale held by the Court throughout the five decades that separate Nielsen v. Denmark and Ibrahim and Others. Worse still, this U-turn in the Court's case-law was not limited to Article 6 § 3 (c), but also extended to (d).

In the case-law concerning the admission of evidence obtained through Article 3 violations, the Court has consistently applied a bright red-line rule, invalidating procedures that rely on such evidence.109. Less clear is the Court's stance regarding the admission of evidence in violation of Article 8 of the Convention.110. In cases concerning examination of witnesses under Article 6 § 3 (d), however, the Court has sometimes applied an overall fairness assessment.111. Admittedly, in order to save the Criminal Justice Act 2003 and satisfy the ‘provocative’ criticism from the Supreme Court of the United Kingdom in Horncastle,112.Al-Khawaja and Tahery departed from the ‘sole or decisive’ rule, according to which a trial was unfair if a conviction was based solely or to a decisive extent on evidence provided by a witness whom the defendant had been unable to question at any stage of the proceedings. In fact, the Grand Chamber was contradicting a solid case-law on the absolute nature of the ‘sole and decisive’ rule, in place since the judgment in Doorson.113. Such an ‘inflexible’ approach involving the non-attendance of prosecution witnesses at trial and the subsequent admission of the absent witness's untested statement as evidence would, the Court argued, run counter to its traditional stance on the need to assess the fairness of the proceedings as a whole.114. Since everything is now negotiable, the Court is not establishing a clear threshold beyond which a conviction becomes impermissibly based on non-tested pre-trial testimonial evidence. Even evidence banned in the Middle Ages, such as the statement of a dying person, can still be used.115.

Less than four years after the delivery of the Grand Chamber judgment in the case of Al-Khawaja and Tahery, it had to intervene again in the case of Schatschaschwili v. Germany, since it remained unclear whether a trial would be considered unfair if there were not good reasons for a prosecution witness's non-attendance alone, even if the untested evidence was neither sole nor decisive. The cure was worse than the disease. In a profoundly divided judgment, the Grand Chamber decided that the absence of good reasons for the non-attendance of a prosecution witness could not in itself be conclusive for the unfairness of a trial.116. As Judges Spielmann, Karakaş, Sajó and Keller denounced, such application of the three-step examination would imply its redundancy so long as the overall fairness test was fulfilled, which would not only fail to provide any guidance to the national authorities as to the appropriate application of the Al-Khawaja and Tahery test, but indeed would give them too much leeway.117. To complicate matters, the then majority added that sufficient counterbalancing factors were still needed if the untested pre-trial testimonial evidence carried significant weight.118. The present judgment closes the circle, in so far as it takes an overall approach to the examination of the fairness of the proceedings in the event of non-attendance of defence witnesses.

Even if, for the sake of argument, one were to agree with the majority regarding the need for flexibility in the Court's case-law on Article 6 of the Convention, the whole enterprise risks logical inconsistency. The fairness of a procedure can only be a procedural assessment: a procedure is not fair because the guilty were convicted or the innocent acquitted, for the simple reason that there is no way to ascertain legally whether the guilty were guilty apart from the procedure itself. Criminal justice is not about ontological justice and judges in criminal trials are not Gods. A procedure is fair when fundamental procedural rules are respected. In this sense, it is difficult to know what the assessment of whether a procedure is fair overall may mean in practice. It could mean that the Court is supposed to examine whether or not the restriction to the right at hand impaired the establishment of the material truth, but this would necessarily imply that the Court has an extra-procedural account of the material truth, which is legally impossible.

A more modest account of overall fairness fails in a similar way. One might say that the violation is ‘cured’ when it did not change the final result of the procedure — when the applicant would have been convicted and sentenced in the same manner had the procedure been flawless. But this is logically impossible, even if one concedes that this assessment can be conducted without an illegitimate extra-procedural account of the material truth. In performing this test, the Court is asked to examine the fairness of a procedure that should not have existed the way it did: for example, had the defendant had counsel of his or her choice, or had the defendant represented him or herself, then the procedure would probably have taken a different path, the defendant would have made different statements, different items of evidence would have been admitted… Ultimately, as the United States Supreme Court has put it, ‘[h]armless-error [or, in the Court's jargon, ‘overall fairness’] analysis in such a context would be a speculative inquiry into what might have occurred in an alternate universe’.119.

Indeed, the Court itself has long recognised the logical obstacles to this kind of assessment. As early as 1980, the Court was examining the case of a convicted person who had received no effective assistance from the lawyer provided by the State. The Government in the case argued that the lack of legal assistance was irrelevant, since it would not have changed the applicant's fate. As argued by the respondent State, ‘for there to be a violation of Article 6 § 3 (c) … the lack of assistance must have actually prejudiced the person charged with a criminal offence’.120. The Court expressed its disagreement in principled terms: ‘nothing in Article 6 paragraph 3 (c) indicate[s] that such proof [that effective assistance from a lawyer would have benefitted the applicant] is necessary; an interpretation that introduced this requirement into the subparagraph would deprive it in large measure of its substance. More generally, the existence of a violation is conceivable even in the absence of prejudice’.121. Even more explicitly, in John Murray, the Court replied to a similar argument by the Government, making clear that ‘it is not for the Court to speculate on what the applicant's reaction, or his lawyer's advice, would have been had access not been denied during this initial period’.122. The Grand Chamber replicated this reasoning almost literally in Salduz.123.

Even ignoring these fatal logical flaws in the majority's reasoning in the present judgment, the overall fairness test is inherently subjective and therefore extremely malleable. This feature renders it a very ill-suited standard for the Court to use in an even minimally predictable fashion. For the same reason, this standard is especially ill-suited to guiding domestic courts when assessing compliance with the Convention: telling a domestic court that procedures should be conducted in an ‘overall fair’ manner does not add much to the goodwill of domestic judges. Reasonable disagreements may exist among the domestic courts, among those which conduct the procedure and those which revise it, and between these courts and the Court, as to what ‘fair’ means in any given context, or in general. Disqualifying domestic assessments of ‘fairness’ seems less compatible with the subsidiarity principle than simply ascertaining discrete violations of specific procedural rights as these are set out in the Convention that Member States ratified. Allowing States to breach strict norms such as those of Article 6 § 3 in favour of such a fuzzy concept invites, rather than prevents, violations to Article 6.

If any evidence of the malleability of the ‘overall fairness’ concept is needed, one may look at Ibrahim and Others, where the Court fleshed out a ‘non-exhaustive list of factors’124. relevant for the fairness analysis. These factors include, among ‘[o]ther relevant procedural safeguards afforded by domestic law and practice’, specific considerations (such as defendants' age and mental capacity) but also other ‘factors’ which are as broad as the fairness assessment itself. The Court states, for example, that it will have regard to ‘the legal framework governing the pre-trial proceedings’ and ‘the quality of the evidence… taking into account the degree and nature of any compulsion’. This is a map of the size of the territory: apparently accurate, but in reality useless in terms of providing any guidance. This is shown by the fact that not even the Court itself reviewed all these elements when conducting the fairness assessment a year later in Simeonovi v. Bulgaria.125. The lack of legal certitude is compounded in the case of a refusal by defence witnesses to attend, because the Ibrahim and Others list of the ten relevant factors for the assessment of the overall fairness of the proceedings126. was totally ignored in the present case, and no effort was made to replace it by an alternative list.

If the lack of clarity already makes the overall fairness test a dangerous blank cheque for domestic courts, there is, however, one particularly dangerous factor that deserves especial attention: ‘[t]he weight of the public interest in the investigation and punishment of the particular offence in issue’.127. One need not take an extremely deontological stand to understand why this kind of reasoning should be prohibited in a court of law. The gravity and nature of offences may prompt States to adopt particular procedural provisions, and, specifically, nothing in the Convention should be understood to prevent countries from fighting terrorism, organised crime and other serious forms of crime. However, listing this factor among those relevant to assessing the fairness of proceedings is very troubling. I note that the Ibrahim and Others judgment did not refer to the public interest in the investigation and punishment of ‘the kind of offence in issue’, but of ‘the particular offence in issue’. This could be read as implying that defendants are entitled to different degrees of fairness in the procedures against them not only because of the kind of offence they are indicted for, but also because of the ‘public interest’ their case happened to arouse. This would be the ultimate paradigm of ‘street justice’, more akin to vigilante punishment than to a lawful trial. Even if I think that the Court could not possibly have intended such a meaning, it is telling that in a paragraph which is envisioned as pedagogical this kind of ambiguity can be found. However, it is unclear what else the Court could have meant. It could have meant that States are granted some discretion in imposing more restrictive procedural rules to crimes that involve especially sensitive issues. If that were the case, however, this issue should have been assessed in the previous stage, that is, when evaluating the reasons the Government provided for the restriction. Taking the public interest into consideration twice entails a double-counting — or, in other words, ‘half-counting’ the importance of the fairness of the trial.

Relatedly, assessing the overall fairness of a procedure implies, in the Ibrahim and Others logic, that the Court should analyse the materials in the criminal file,128. notably the evidence used for conviction, and make a twofold assessment. First, whether the evidence in the file would have been there had the applicant's right to a lawyer not been infringed. And second, whether that evidence would reasonably have led to conviction, even if the defendant had had legal assistance of his or her own choice. These hypothetical assessments cannot be conducted without meaningful engagement with the domestic procedural law, such as that governing standards of proof and admissibility of evidence. This seems a task that the Court is not best-equipped to carry out while simultaneously respecting Member States' authority in designing their own procedural systems.

Lastly, if the Court's decisions are to provide any guidance to domestic authorities, it is unhelpful to ask domestic courts to conduct proceedings on the basis of overall fairness. This is not only because of the fuzziness of the concept but, more fundamentally, because courts cannot know at any given point of time how a given set of proceedings will continue in the future, since different bodies are charged with different tasks at different moments. For example, a judge or public prosecutor who does not allow the defendant to have access to a lawyer of his or her choice while conducting a preliminary investigation cannot possibly know in advance whether the rest of the proceedings, carried out in part by different bodies, will be conducted ‘sufficiently fairly’ to compensate for a putative violation of the Convention. The impossibility of assessing the overall fairness of a procedure from inside the procedure itself should be a strong indicator that this is a poor standard for the Court to uphold.

All of the above applies to the rights laid down in Article 6 § 3 (d).129. Systematically, these rights are manifestations of the fairness principle, not just mere recommendations. They are also rooted in a European consensus on the principle of the immediacy of the evidence (Unmittelbarkeitsgrundsatz), from which the rule against hearsay is derived. In the light of the principle of immediacy, the close connection between the testimonial evidence and the court is philosophically perceived to provide the best possible approximation towards the truth. These principles are accepted in Europe and beyond, throughout different types of criminal-law systems, irrespective of whether and to what extent they are adversarial or inquisitorial. Although the manner in which the principle of immediacy is implemented in each national system differs, the acceptance of the principle as such is not denied.130. In the light of this consensus and the Court's long standing case-law on the subject, the principle of immediacy imposes the following guidelines for trial courts, giving them not only a retrospective standard of review, but also a standard for the admissibility of prospective evidence at trial:

Guideline 1: The principle of immediacy requires that testimonial evidence be produced in the proceedings before the competent trial court, and therefore all viable measures, and if necessary, forceful measures, should be taken to bring absent witnesses to the trial hearing.131.

Guideline 2: The test for assessing any request for the attendance or cross-examination of witnesses, which is identical for the defence and the prosecution, relies on the individual witness's prima facie relevance to the outcome of the trial and to the position of the requesting party. This is an exclusively ex ante test. Hence, in assessing any request for attendance or cross-examination of witnesses, the trial court is not allowed to anticipate the result of the examination of the evidence — except in favour of the defence.

Guideline 3: Only exceptionally may testimonial evidence produced at the pre-trial stage in criminal proceedings be read out in court and used as a ground in the judgment. The principle of legality in criminal procedure requires an exhaustive legal catalogue of good reasons for not calling or cross-examining a witness during the trial and for reading-out an absent witness's testimony at the trial. This catalogue of good reasons includes two categories of witnesses: unavailable witnesses, and witnesses in need of protection. The category of unavailable witnesses has the following sub-categories:

The category of witnesses in need of protection has the following two sub-categories:

The catalogue of these grounds must distinguish between testimonial evidence produced before a judge, public prosecutor or the police, since the evidence produced before the police or a prosecutor at the pre-trial stage cannot be equated, in its reliability and accuracy, with the evidence produced before the judge at the pre-trial stage. The catalogue of good reasons for reading out an absent witness's testimony at the trial must be more expansive when a judge has collected evidence at the pre-trial stage, and less expansive when it has been collected by a public prosecutor or the police.

Guideline 4: When there is a good reason for not calling or cross-examining a witness and for the pre-trial witness's testimony being read out in court, the trial court must ensure that sufficient counterbalancing measures were or are taken to compensate for the handicaps imposed on the defence, such as a special pre-trial hearing at which the witness's evidence was taken under adversarial conditions.141. Neither a direction to the jury that the untested statement should be given less weight,142. nor the possibility that the defendant challenge or rebut the testimony of the absent witness, by giving evidence himself or herself or by examining another witness, suffices.143.

Guideline 5: Even when sufficient counterbalancing measures were taken to compensate for the handicaps imposed on the defendant, the trial court must not base a judgment, alone or to a decisive extent, on the pre-trial testimony of an absent witness.144. Such reasoning would negate the essence of the principle of immediacy.

Guideline 6: When the trial court is faced with a contradiction between the pre-trial and the trial statements of a witness, the principle of immediacy requires that the latter should be given greater weight.

Guideline 7: These are bright red lines that cannot be crossed by a trial court, under pain of leading to a violation of the principle of immediacy and consequently the unfairness of the trial. No evidence can be taken into account by the trial court in breach of the above guidelines, unless domestic law provides for the possibility of an agreement between the prosecution and the defence regarding the reading out of pre-trial testimonial evidence, and both parties agree to that reading.

Witness A. was an officer of the Organised Crime Department and always acted on instructions from his superiors. The Government themselves described his role as follows: ‘Therefore, Mr A.'s role in the events in question was limited to his technical facilitation of operative-search activity in respect of the applicant.’145. In fact, there was a covert operation against the applicant, and witness A. acted as a covert agent whose activities led to the collection of the entire body of evidence against her, mostly through video recordings made in the dormitory room she shared with V. and Ku.

The categorisation of witness A as a defence witness is problematic. Although it had been on the applicant's behalf that A. was questioned as a witness at the pre-trial stage, it was the public prosecutor who moved to have his testimony read out at the trial and the trial court relied on this written testimony to convict the applicant. This problematic categorisation is not excluded by the fact that the national courts did not explore the probative meaning of the testimony given by A. The particularity of the situation in the present case lies in the fact that the conviction relied on the assumption that no entrapment or other kind of police-induced circumstance had occurred which could have resulted in the acquittal or exoneration of the applicant.

The applicant herself reiterated during the trial that she had insisted on a confrontation with witness A., but to no avail.146. No steps were taken to establish the whereabouts of witness A., who had taken part in the covert operation against the applicant, and to have him attend the trial.147. The trial court merely concluded that all the evidence against the applicant was admissible, and the Russian Supreme Court upheld that finding on appeal without any further elaboration. With regard to the applicant's submission on entrapment, the Supreme Court merely stated that the appeal bench was ‘unable to agree’ with the defendant, without adding any further considerations.148.

Neither the applicant nor her lawyers were present at the questioning of witness A. during the pre-trial phase. The defence was not given an opportunity to confront A. at the pre-trial stage or during the trial. Since there was no waiver of the examination of witness A. or valid consent to reading out the absent witness's testimony under Article 281 § 1 of the Russian Code of Criminal Procedure, section two of the same Code applied. It was for the trial court to verify whether the reasons for witness A.'s non-attendance were valid and, if they were not, to ensure that witness A. appeared before it, for example by issuing orders compelling him to appear.

It is plain that no good reason can be given to justify the failure to call witness A. to testify at the hearing as required by section two of Article 281. The Russian Code of Criminal Procedure explicitly provides examples of good reasons, as well as an omnibus clause referring to ‘other exceptional circumstances’. These come close to the criteria that were established by the Court in Al-Khawaja and Tahery. Besides, like many other provisions in European codes of criminal procedure, the provision in the Russian code does not differentiate between witnesses on behalf of the defence and those appearing for the prosecution.

In this context the Court has held ‘that where an accused asserts that he was incited to commit an offence, the criminal courts must carry out a careful examination of the material in the file, since for the trial to be fair within the meaning of Article 6 § 1 of the Convention, all evidence obtained as a result of police incitement must be excluded.’149. Applying these principles to the present case results in a special duty on the national court to have the evidence on whether police incitement took place closely examined, in order then to be able to exclude the respective part of the evidence. This duty did not arise from the applicant's defence concerning her perception of ‘provocation’ by law-enforcement agents alone, but also from A.'s deep involvement in the case, as became clear from his own pre-trial statement as well as the videotapes collected from a hidden surveillance camera in the applicant's flat. Another important hint which should have made clear the obligation to investigate the possibility of witness A.'s involvement in the offences was the testimony from witness V., who stated that some of the books on ‘the way of jihad’ had been given to the applicant by her acquaintance A.150.

Finally, witness V. gave contradictory statements during the pre-trial and the trial questioning151. and witness Ku. retracted the statements made by her at the pre-trial stage, instead stating that the applicant had not incited her to become a suicide bomber.152. In sum, witness V. stated during the trial that the applicant had been preparing her and Ku. to become suicide bombers and witness Ku. denied V.'s statement, stating that the applicant had not incited her to become a suicide bomber. In view of these contradictions, the pre-trial statements of both witnesses were read out during the trial. Ultimately, the trial court made use of the pre-trial statements by witnesses V. and Ku. and of the statements made at trial by V.153.

The majority conclude that the applicant was able to conduct various other procedural measures, without referring to the important question of what the cross-examination of witness A. would — at best — have meant for the applicant's defence. The procedural aim of confrontation could neither be reached by commenting on his pre-trial statements nor, for example, by questioning other witnesses about the events in which A. had been involved, as the majority claim.154. The fact that other procedural rights of the defendant were upheld by the court should be considered as a matter of course, not as an argument for lowering the threshold for assessing whether the court violated another of the defendant's rights. These rights are not to be confused with procedural safeguards with respect to evidence that could not be challenged by the defendant. With this argument the majority — despite their conviction that the national court is better placed than the Court to decide about the necessity of evidence — contravene the prohibition on anticipating evidence (Verbot der Beweisantizipation). By using the argument that the applicant could have used these other kinds of evidence, they anticipate that the result of witness A.'s cross-examination would not have changed the content of his testimony in relation to the pre-trial stage.

B. and K. were attesting witnesses. As a matter of principle, the status of an attesting witness is problematic in the light of the Convention. According to Article 60 of Criminal Procedure Code of the Russian Federation, an attesting witness is ‘a person, disinterested in the outcome of the criminal case’; he or she is entitled to take part in an investigative action and make comments on it, and these comments are to be entered into the record. An attesting witness is also entitled to familiarise himself or herself with that record.155. Since there are no legal requirements as to the legal qualification of the attesting witness, he or she may not be in a position to comment on, let alone verify, the legality of the procedure adopted by the investigator. This evidently empties the role of the attesting witness of much of its interest. Even where he or she does possess a legal qualification, the attesting witness evidently cannot replace the guarantee of legal assistance to the accused or to suspects who have not been charged, for the obvious reason that he or she is chosen by the investigator. Furthermore, there is a serious risk of misuse of this witness by the prosecutorial side, since it can be easily anticipated that the investigator will choose individuals who have his or her trust and who will not create problems. Moreover, when brought and questioned before the trial court, an attesting witness may be led to reproduce what he or she heard from the material witnesses or the accused or the suspect during the investigative action. Such questioning would correspond to misuse of the attesting witness as a hearsay witness at the trial. In order to comply with the prohibition on hearsay, the questioning of attesting witnesses would have to be limited to the circumstances surrounding the investigative action and its conduct. Finally, the Court itself has already held that the absence of attesting witnesses in court does not infringe Article 6 §§ 1 and 3 (d) of the Convention, since ‘it appears that the contribution by attesting witnesses to the proceedings was limited to statements on the manner of conducting the investigative measures’ and ‘their depositions did not serve to a material degree as a basis for the applicants' convictions’.156.

In the case at hand, attesting witnesses B. and K. were invited to attend the search of the applicant's bag. The criteria for inviting these specific persons are unknown, as are their qualifications for that purpose and whether they were sworn in. They were not questioned during the preliminary investigation and no confrontation took place between them and the applicant. In due course the defence lawyers requested their attendance at the trial, in order to determine the exact circumstances of the search of the applicant's bag. The applicant did not insist that these witnesses be questioned, but acceded to the motion of her lawyers, who argued that the questioning was necessary.157. Without any justification, the trial court rejected the defence motion. In its judgment, the trial court did not mention witnesses B. and K, but only referred to the testimony of the patrol officers and officers at the police station, who denied the allegation that the explosives had been planted in the applicant's bag, and to the official report, according to which the personal search had preceded the taking of fingerprints.158. The Supreme Court did not consider the defence request unsubstantiated or unreasoned, but argued that the attendance of witnesses B. and K. was not necessary, because the applicant claimed that the explosives had been planted in her bag before their arrival.159. Like the Government,160. the Russian courts prejudged the requested witnesses' testimony, adducing that they would in any case be unable to provide any evidence beyond that already available to the courts.161.

Contrary to the majority's claim,162. it is not true that the defence request to examine witnesses B. and K. during the trial ‘did not contain any particular factual or legal arguments’: it contained the factual subject matter of the questioning (the exact circumstances of the search) and its legal purpose (establishing whether the explosives had been planted in the applicant's bag)163.. Furthermore, when requesting the attendance of defence witnesses, the defence does not have to ‘elaborate in concrete terms’ on how their testimony would assist the case for the defence.164. Such elaboration would equate to asking the defence to surrender its trial strategy and particularly its questioning strategy. More troubling still, no such elaboration is demanded from the prosecution.165. It is patent that the majority demand from the defence much more than they require the prosecution to do. Ultimately, the majority contradict themselves, because they admit that, on the basis of the defence request alone, the attesting witnesses' testimony ‘would have ranged beyond the mere modalities of the search and the information subsequently entered in the police records’,166. and therefore acknowledge their relevance as defence witnesses; at the same time, however, they affirm that ‘further reasons for the examination of these witnesses would have been required’.167.

The majority's uneven approach becomes even more questionable when, in order to justify the non-attendance of witnesses B. and K., they embark on an open-ended criticism of the defence strategy, pointing twice to the defence's ‘general passivity’168. during the examination of the police officers. Such virulent criticism, which goes far beyond the prima facie pertinence test, is unacceptable. When assessing the request for the attendance of defence witnesses, it is absolutely illegitimate to extrapolate conclusions that are negative for the defence from its strategy in the cross-examination of prosecution witnesses. The majority here mix apples and oranges. In so doing, they assume that ‘the significance of the attesting witnesses' possible testimony … from the perspective of the trial court, was only remotely relevant to the subject matter of the accusation.’169. In other words, they contravene the prohibition on anticipation of the evidence. Even worse, in so doing, the majority verify the allegation of planted evidence only by giving credence to the very police officers who could possibly be responsible for such planting of evidence. This is simply not fair. Such blatantly unfair treatment of the defence is not compensated by the promised ‘overall fairness’ assessment.

In fact, in the eyes of the majority, the overall fairness test is ultimately applied, not to the proceedings, but to the merits of the trial court's judgment. In condoning the decision not to examine witnesses B. and K., the majority invoke the fact that the conviction was based ‘on a considerable body of evidence’.170. The alleged fairness of the outcome of the proceedings justifies the shortcomings in the procedure. The end justifies the means. The Court is going back hundreds of years in time, ignoring the fact that any short-term gains associated with more attainable convictions in some high-profile cases, such as the present one, must be assessed against longer-term systemic losses.

In this context, the subsequent allegation that the applicant, ‘assisted by professional lawyers, was able to conduct the defence effectively, confront and examine witnesses testifying against her, comment without hindrance on the incriminating evidence, adduce evidence she considered relevant and to present her account of the events to the domestic courts’ means little, but the little it means is dangerous. It is dangerous that respect for certain other procedural rules can be used as an excuse to justify a failure to comply with Article 6 § 1 (d). It is dangerous because it assumes that compliance with Article 6 § 1 (d) would have not have changed the outcome. As if the majority had divinatory power to perceive material truth beyond the limits of the procedure and to determine, in all their inexplicable, transcendental discretion, which rule of procedure should apply in each individual case. Human justice playing God's justice is justice at its worst.

Alarm bells should be ringing among lawyers as a result of this judgment. It is particularly striking that the majority state that the applicant was able to ‘adduce evidence she considered relevant’,171. as if this had been the most impeccable of trials. The true novelty of this judgment is twofold: the enhanced obligation for the defence to ‘elaborate in concrete terms’ on how the defence witness's testimony would assist the case for the defence, and the corrosive expansion of the overall fairness test to the assessment of alleged violations of the right to examine defence witnesses. This test is nothing more than a blank cheque for the domestic courts to do whatever they want with Article 6 § 3 rights and for the Court to confirm the outcome of the proceedings. After the weakening of the right to a confrontation with prosecution witness in the Al-Khawaja and Tahery and Schatschaschwili judgments and the eroding of the right to legal representation in Ibrahim and Others, the present judgment on the non-attendance of defence witnesses closes the circle, by revisiting the Perna test on the basis of the overall fairness test. With the unfortunate triangle Schatschaschwili/Ibrahim and Others/Murtazaliyeva, the Court erroneously pursues the double path of persistently watering down defence rights and ‘deferential[ly]’172. abandoning its supervisory powers to the domestic courts. The silent assault on the rights of the defence in criminal procedure has been incontrovertibly stepped up, to such an extent that one wonders where it will stop.