CONCURRING OPINION OF JUDGE KŪRIS

I

1. This is an outstanding case – owing not only and not so much to its own merits, but more to its relation to one earlier case initiated by the same applicant, Ramanauskas v. Lithuania ([GC], no. 74420/01, ECHR 2008). The present judgment sheds some new light on that ten-year-old judgment and, more generally, on the Court ’ s methodology of the assessment of proof and also, conceivably, on its predisposition to give or not to give credence to the parties ’ submissions in the agent provocateur cases.

2. Mr Ramanauskas, a professional lawyer with years of experience, applied to the Court again for a similar reason: the alleged incitement to commit a crime. In his first case he admitted that he had succumbed to what he called “undue pressure” and had taken a bribe. In the instant case he pleaded not guilty: allegedly there had been an incitement, but no criminal activity on his part. For most people it is enough to step on a rake once. Others keep trying, especially if stepping for the first time did not hurt or the bruises paid off.

3. In the first case the Court found for the applicant. Being ceremonially recognised to have fallen victim of a malicious provocation staged by the authorities, he was acquitted. His slate having been cleaned, he could even have returned to his prosecutorial position. He chose to start a new career as an advocate. In this new capacity, he did not dilly-dally about visiting (at the invitation of an officer at a correctional facility, who admitted his guilt and was convicted) a prisoner to consult the latter on the matter of how much it would cost him in bribes to be released on probation and to receive cash for illicit activities from an intermediary. Whether the applicant would have greased anyone ’ s palms in the judicial system (he mentioned names) or would have pocketed all or part of the money without having accommodated the alleged instigator with the “service” requested would be a matter of sheer speculation. What is certain is that he promised to provide the illicit “service” and accepted the money.

4. Ramanauskas v. Lithuania (cited above) is a landmark case indeed. As a Grand Chamber judgment, it has attained the status of a leading case and is repeatedly cited in just about all the subsequent agent provocateur cases, not excluding the instant one. It would be that judgment in which the relevant principles, which I do not intend to contest here in their essence , are set out.

5. But does that judgment as a whole (not confined to its doctrinal part) still maintain the quality of an authority – not in the specific power-related or judicial sense, but in the original meaning of the word ( auctoritas ) with its connotations of particular convincingness, reputation and legitimacy? Today one perhaps could doubt this – not completely gratuitously.

6. Whatever its doctrinal merits, that judgment effectively exonerated the applicant from culpability for his hapless frailty to the lure of felonious honorarium and veritably emboldened him to embark on yet another bribery exploit. Buoyed up by his recent triumph in Strasbourg, he exulted over it (with a bit of varnish) to his collocutor, the alleged instigator. When caught red-handed, the applicant (as was his habit?) protested that he had fallen victim to a provocation and provided a preposterous explanation as to the purpose for which he had accepted the money and no explanations as to the reason for which the authorities might once again have been after him.

7. It is hard to shake the impression that this prosecutor-turned-convict-turned-applicant-turned-acquitted-turned-advocate expected the Lithuanian courts to swallow any old story furnished by him, because, from the normative angle, they were under an obligation to give credence hook, line and sinker to virtually any tale. Indeed, it is well-nigh impossible to rebut with one hundred percent certainty even the most inconceivable version of events of a person claiming to have fallen victim of an entrapment. Should the domestic courts not buy the applicant ’ s story, this had to be done by ECtHR, which in its innocence imposed on itself an obligation to assess the alleged victims ’ allegations leniently, even gullibly, if Ramanauskas v. Lithuania (cited above) is read for what it says. In the Court ’ s words, which not only migrate through its subsequent case-law, but also have set themselves firmly in domestic case-law (see paragraph 41 of the judgment), “[i] t falls to the prosecution to prove that there was no incitement, provided that the defendant ’ s allegations are not wholly improbable” and “[i]n the absence of any such proof, it is the task of the judicial authorities to examine the facts of the case and to take the necessary steps to uncover the truth in order to determine whether there was any incitement” ( see Ramanauskas, § 70 , cited above ; my emphasis).

II

8. The first of the quotes provided in the preceding paragraph includes the word “wholly”. The key word. Dictionaries instruct us that “wholly” means “fully”, “totally”, “absolutely”, “perfectly”, etc. But let us start with the second quote cited, the “uncovering the truth” clause.

9. There is no need to enter into the philosophical facets of the notion of “truth” and the conundrum of the “truth ’ s” full attainability, which, as we know, is often frugal. No one will ever learn for sure whether the Grand Chamber was (or was not) purposely mindful of these entanglements, but it skilfully (or happily) escaped the subaqueous rocks of epistemology, as it phrased the “uncovering the truth” clause quite cautiously. The Court does not require that the “truth” be “uncovered” in its entirety, only that the “necessary steps” be taken to lead in the direction of its “uncovering”, and only insofar they are “necessary” to “determine whether there was any incitement”. Also, it is not made explicit that all the “steps” must be taken, only those “necessary”. The Court understandingly notes that it is “ aware of the difficulties inherent in the police ’ s task of searching for and gathering evidence for the purpose of detecting and investigating offences” (ibid., § 49) and that it is “mindful of ... the difficulties of the task of investigating offences” (ibid., § 73). One presumes that it should be no less sensitive to the difficulties encountered by the courts dealing with the cases, where the persons implicated in corrupt activities claim to have been incited to commit them. It is especially onerous for the courts to “uncover the truth”, when they have to choose between competing testimonies, where the prosecution ’ s version is only feebly supported, or even not supported, by a plethora of authorities-unrelated witnesses ’ testimonies or non-testimonial evidence, while the defendant ’ s story, though implausible in real life, could have been “not wholly improbable” under some extremely fluky, fortuitous coincidence of circumstances. The crucial factor is whether the Court grants the alleged victim ’ s version the benefit of doubt or dismisses it as “wholly improbable”. If that version is potent or, on the contrary, utterly fantastic, the Court ’ s task is relatively easy. Still, there is no sharp boundary between the core of undisputable plausibility and the surrounding province of sheer wanderings: in both these domains the certainty of getting closer to the “truth” raises no great concerns, but between them lies (if I may import this Wittgensteinian-Hartian construct) the penumbra of doubt where a greater or lesser judicial discretion is exercised. When the Court is necessitated to enter that penumbra, it faces the delicacy of walking the fine line between remaining the master of “characterisation to be given in law to the facts of the case”, as it often rightly calls itself, and mutating into being the master of such recognition or refutation of the facts presented by the parties where it effectively discovers them, thus becoming a “fourth-instance court” (or even a “first-instance court”).

These considerations impel us to have a closer look at the other clause cited in paragraph 7 above – the “not wholly improbable” clause.

10. The Court can hardly be said to have been sufficiently vigilant in wording that clause – at least, it was not as circumspect as in phrasing the “uncover the truth” clause. In Ramanauskas v. Lithuania (cited above) the “not wholly improbable” clause has become the Court ’ s translation of (and a surrogate to) the well-known and long-lived adage in dubio pro reo , which is one of the pillars of the fair criminal procedure in particular and of the rule of law in general. In dubio pro reo requires that a ny reasonable doubt must benefit the accused. The latter thus must assuredly benefit also from a doubt as to whether he or she had not been incited to commit a criminal offence which otherwise would have not been committed, but only from such a doubt which allows for the possibility (or probability), however slight, that the events evolved in some condonable way, different from the one asserted by the prosecution, but altogether not unlikely in comparison to how things normally are in life . In dubio pro reo is a pragmatic principle. It does not give a blind eye on the shared human experience .

11. In contrast to in dubio pro reo with its reliance on the perception of how things normally are in life, the “not wholly improbable” is an absolutist formula. It explicitly postulates not the “beyond the reasonable doubt” standard, but the one where there can be left no doubt at all . By employing the formalistic, arithmetic-scented adverb “wholly”, the Court substituted the complete improbability, or the probability that equals zero, for the pragmatically reasonable impossibility, or the chance that something took place being factually inconceivable from the angle of the shared human experience, or the knowledge of how things normally are in life . By confusing the improbability as an arithmetical zero and the improbability as pragmatically reasonable impossibility, the clause in question tells that the accused must benefit from virtually any doubt, however meagre, unless the latter is absolutely hollow, unnatural, i.e. “ wholly improbable”, because everything what is not unnatural is also “not wholly improbable” by definition . The accused thus should benefit also from doubts which are artificially invented and purely imaginative, but not “unnatural” in the strict sense of the word. It would be very difficult to conclusively rebut each and every fanciful version, if the probability of them having taken place does not render them unnatural and therefore does not equal zero.

12. Maybe the above is mere quibbling. Chicanery? Not really. The legal text is read, prima facie , according to its plain meaning. True, literal reading is seldom the ultimate reading. Maybe therefore a more liberal reading of the “not wholly improbable” clause is warranted? Yes and no. Yes, because, as is generally accepted, the plain meaning of the provisions must not be given (and in the Court ’ s case-law at large is not given) undue prominence. There is a range of interpretative instruments aimed at alleviating the constraints of the plain meaning rule, e.g.: the golden rule (in its narrower or broader versions); the mischief rule; the purposive approach; etc. Even the plain meaning rule itself has its softer version, not rejected even by the most ardent adherents of textualism. These instruments, in principle, allow for not applying the “not wholly improbable” clause in its strictly verbatim sense and not to take at face value any story of the alleged victim of incitement to commit a criminal act, provided that it is not unnatural, although, if assessed from the pragmatically reasonable angle, quite absurd indeed. But there are caveats. Firstly, these instruments are meant, on the whole, for statutory and constitutional interpretation and not for that of doctrinal provisions of the courts ’ case-law. Another caveat pertains to the factual side of adjudication: in the Court ’ s case-law the “not wholly improbable” clause is at times taken exactly for what it literally says (for a recent example see P ă tra ş cu v. Romania , no. 7600/09, §§ 38 and 49, 14 February 2017). Even if it might have been worded, in Ramanauskas v. Lithuania (cited above), so strictly with no special intention, the later ordaining, even if infrequent (contrast Bannikova v. Russia , no. 18757/06, 4 November 2010), of the letter of the tenet to its rigid meaning does not warrant dismissing the criticism of the wording of the clause in question as a mere carping at trifles.

13. One may wish (as I do) that, in lieu of the absolutist, even somewhat extremist formula “not wholly improbable”, a temperate down-to-earth and middle-of-the-road formula “not reasonably improbable” (or “impossible”, “implausible”, etc.) was coined in Ramanauskas v. Lithuania (cited above) or fine-tuned in the subsequent case-law thus remedying the fait accompli . The doctrinal guidance for endowing with the benefit of doubt, but in fact with the “judicial belief”, formally not unnatural, but nevertheless ludicrous stories which very few (if anyone at all) can believe on the basis of the shared human experience thus would have been eliminated, at least minimised. And vice versa : if the clause in question is applied literally in the case, in which the contention regarding the “truth” not yet “uncovered” boils down in essence to “his word against theirs” contraposition, then not the prosecution, but only the alleged victim of an incitement can effectively enjoy the benefit of doubt; such a defendant must be ultimately acquitted.

14. In the disputes of “his word against theirs” type the whole “truth” is hardly attainable. The conviction of the accused, whose unrealistic story the court has not bought, may still leave some doubt (even if minuscule ): what if the events did in fact happen contrary to how things normally are in life? On the other hand, t he exculpation of the accused on the sole ground that the prosecution failed to provide cast-iron proof that there was no incitement, whereas the defendant furnished the story which could not be rebutted with one hundred percent certainty because it indeed could hold true under some extremely unlikely amalgamation of circumstances, may also be far from “uncovering the truth”. As the “absence of any such proof” may not be unequivocal, which side to take is a matter of balancing of proof by the Court – at times not an easy task. Still, it is the accused who must benefit from any doubt. If his or her exculpation on the basis of the “absence of any such proof” leaves too much room for doubt, the “not wholly improbable” clause can be succoured by the “uncovering the truth” clause: with hindsight it is easy (and tempting) to hold that the domestic courts had not taken all the “steps” that might have been “necessary” for “uncovering” the “truth”, even if in practice it would have been disadvantageous or hardly possible to take every “step” which one might envisage. What latitude for discretion: though it is not explicitly required that all the “necessary steps” are taken, it is in fact one and the same thing to say that the “necessary steps” had not been taken or that not all of them had been taken. For the finding of a violation of Article 6 § 1 of the Convention it may suffice to highlight one single “necessary step” and to mark it as “not taken”, even if its usefulness or practicability could reasonably have been seen differently at the material time.

III

15. Ramanauskas v. Lithuania (cited above) was an almost typical “his word against theirs” case, save that “their words” were partly supported by other evidence , including the authorised secret recordings of the applicant ’ s conversations with one of the alleged instigators. Conflicting testimonies as to “ who incited whom to give or accept the bribe ” were provided: the applicant alleged instigators attributed the mentioning of the money to the applicant, while he blamed them . T he Supreme Court admitted the difficulty of “establish[ing] who was the instigator of giving and accepting the bribe ”, but held that “[e]ven assuming that [Mr] Ramanauskas was incited by [other persons] to accept a bribe, ... the incitement took the form of an offer , and not of threats or blackmail”, and [h]e was therefore able to decline (and ought to have declined) the illegal offer ... [however, he] accepted [the bribe] of his own free will” ( ibid., § 27; my emphasis). That was sufficient for the conviction. ECtHR summed up that approach by stating that “[o]nce [the applicant ’ s] guilt had been established, the question whether there had been any outside influence on his intention to commit the offence had become irrelevant” and observed that a “confession to an offence committed as a result of incitement cannot eradicate either the incitement or its effects” ( ibid., § 72). Given that the Supreme Court itself conceded that “the incitement took place”, although it clearly meant the alleged instigators ’ request to secure the acquittal of a third person and not the offer of the “reward” to the applicant, the Government ’ s case was hopeless. At the heart of its failure to convince t he Court that the offence would have been committed without the outside influence was not that the Supreme Court erred by not establishing with one hundred percent certainty that the applicant had not been incited to take the bribe (that part of the “truth” did not lend itself to “uncovering”), but that it was satisfied – in line with domestic criminal law – with the mere fact that he had taken it. The Court, however, was not satisfied with that.

16. I am prepared to accept that for the finding of a violation of Article 6 § 1 it was enough that the Supreme Court held it to be “irrelevant” whether there had or had not been an incitement. (It also went into the terminological considerations on the similarities and differences between “provocation”, “incitement” and “inducement” (ibid., § 27), which obscured the reasoning or, rather, its understanding in Strasbourg ). The judgment leaves little doubt that this was a fatal slip: under the Convention, as interpreted by the Court, one cannot be convicted on the sole basis, in principle, of the commission of a crime without the possibility of an incitement being excluded.

17. There is a difference between not excluding a possibility and stating that it had materialised. While the Supreme Court held it to be irrelevant for the conviction of the applicant that he had been asked to perform illicit actions on behalf of a third person and in this respect had been incited, but admitted its inability to establish “who was the instigator of giving and accepting the bribe ”, The Court, however, was not sophisticated at all in distinguishing between the request to perform the said actions and the offer to “remunerate” them. Both would constitute an incitement. The question remains whether they should be seen as one “compound” initiative or two (related, of course) intiatives. It was accepted even by the Supreme Court that the applicant had been solicited to commit the first set of illicit actions. But it could well have been that he would not have agreed to commit them if the alleged instigator had not agreed to his request for “reward”. The Supreme Court admitted the “first” incitement and left the question open as to the “second” one, but held that they were irrelevant for the conviction. In contrast to that, the Court ’ s judgment is worded so as to make it clear that also the initiative regarding the “remuneration” had come from the “outside”. E.g., it is stated that the applicant “ had apparently agreed to seek to have a third person acquitted in return for a bribe of USD 3,000”; or that “the actions of [the alleged instigators] had the effect of inciting the applicant to commit the offence of which he was convicted ”, i.e. accepting a bribe , that is to say, a “reward” (ibid., §§ 62 and 73; my emphasis). Where the domestic courts, which directly examined all the evidence before them, failed to “uncover the truth”, that mission was accomplished from a distance . While the Supreme Court employed the fiction of division of the incitement to commit a criminal offence into two elements (not a meaningless analytical enterprise as such), ECtHR no less fictitiously employed the inductive inference and discovered that what was true of a part was true of a whole. Not only that was not necessary for the substantiation of the finding of a violation of Article 6 § 1 – that diminished the strength of the whole reasoning.

18. Another fiction employed in Ramanauskas v. Lithuania (cited above) also amounting to establishing of the fact, is worthwhile mentioning. T he Supreme Court established that one of the alleged instigators was a “police driver” who initially acted “in a private capacity”. The Court “promoted” that person to the rank of an “officer of a special anti-corruption police unit” (this is the same Special Investigation Service, as in the instant case, only then its name was translated loosely ; ibid., §§ 13 and 27 ). Well, being a police driver does not make one an “officer”, even if one acts as an undercover agent. Why was such a sham transmutation resorted to at all? One could surmise that otherwise the Court would not have had a sufficiently solid basis for inferring that the “Lithuanian authorities ’ responsibility was engaged ... for the actions of [the alleged instigators] prior to the authorisation of the [criminal conduct simulation] model”. This factual inference is drawn from the fact that “no satisfactory explanation has been provided as to what reasons or personal motives could have led [the driver-officer in question] to approach the applicant on his own initiative without bringing the matter to the attention of his superiors, or why he was not prosecuted for his acts during this preliminary phase”. True: no such explanation has been provided (ostensibly because the relevant file had been destroyed upon the expiry of the period established in the law, which presumably preceded the communication of the case to the Government). The Court might have inferred rightly – but it might have erred. There is a difference between the Government not proving their case before the Court and the Court discovering itself that the facts were different from those presented. The reasoning cited above erases that difference. Without going into a greater detail, I would only note that in order to additionally support the inference regarding the “authorities ’ responsibility ... for the actions ... prior to the authorisation of the model”, which belongs to the domain of “is”, the Court resorted to an argument from the domain of “ought to be”: “To hold otherwise would open the way to abuses and arbitrariness by allowing the applicable principles to be circumvented through the “privatisation” of police incitement” (ibid., § 65). It is a moralistic fallacy in its prime.

19. In general, the Court ’ s whole assessment of the situation addressed in Ramanauskas v. Lithuania (cited above) is permeated with distrust of the authorities ’ every submission: they all are rejected without mercy. To single one out: in the Court ’ s words, the alleged instigators ’ contact with the applicant had been a “blatant prompting ... to perform criminal acts, although there was no objective evidence – other than rumours – to suggest that he had been intending to engage in such activity” (ibid., § 67). While the domestic courts are criticised for not having “establish[ed] ... the reasons why the operation had been mounted” (ibid, § 71), it appears that the Court itself established that there were no serious reasons (the operative information being derogatorily dubbed “rumours”). The rest we know: violation of Article 6 § 1 and financial compensation → acquittal → new career → bragging about the victory in Strasbourg → another bribe → Strasbourg again; but all that is a posteriori . The applicant ’ s conversations dealt with in the instant case (by the way, what is reproduced in the judgment from his speech had to be expurgated and some parlance has not been reproduced so that the judgment does not become an 18+ reading) demonstrate that one must think twice before taking as gospel truth that the applicant himself would never-never seek being bribed and that it is (usually?) someone else who first offers him the “remuneration”.

20. That much to the application of the “not wholly improbable” clause. But the finding reached on this basis was also supported by praying in aid the “uncovering the truth” clause. As the applicant ’ s story, in the Court ’ s opinion, was “not wholly improbable”, it merited the “necessary steps” leading to the direction of “uncovering” of the “truth”. One “step” seems to have been playing truant. The domestic courts did not call to testify in court one of the two alleged instigators, the intermediary of the “driver-officer” discussed above, whom the Court has not labelled an “officer” of a special service, but still rightly held to “app[arently] hav[ing] played a significant role in the events leading up to the giving of the bribe” (ibid., § 71) . His confrontation with the applicant in the courtroom might have shed more light on the matter. But he could not be traced . The trial court was therefore cautious; it “did not take into account [that witness ’ statement at the pre-trial stage] in determining the applicant ’ s guilt” and based the conviction on other evidence (ibid., § 24). However, the Supreme Court held that it was unnecessary to exclude that evidence, because according to domestic criminal law, regardless of who was the instigator, the crime had nevertheless been committed and entailed responsibility (see paragraphs 15 and 17 above). The non-exclusion of the impugned evidence might have been nominally in line with the Court ’ s requirement that “all evidence obtained as a result of police incitement must be excluded” ( see Ramanauskas , § 60 , cited above ), because it was not established that there had been an incitement regarding money changing hands (however critically this approach was assessed by ECtHR).

21. Equality of arms is a serious matter. I am ready to agree with the Grand Chamber that the presence, in the courtroom, of the witness in question was desirable: even if his confrontation with the applicant in court would not necessarily have helped to get much closer to the “truth”, the courts at least would have taken one more “necessary step” in that direction. I am also prepared to agree that , in principle, the authorities should have shown as great a diligence in tracing that person as possible. But what was possible at the material time and what diligence would have sufficed ? Could the authorities reasonably have foreseen that that person would become untraceable? If so, should they have imposed restrictions on that person ’ s freedom of movement in order to secure his presence at the trial? What restrictions? Would they have not infringed his rights? It is easy to judge with hindsight. However, if at the time of the trial the witness ’ whereabouts had indeed not been known to the authorities ( this was not rebutted by anyone ), the Court ’ s ex post facto consideration that his presence should nonetheless have been secured amounted to the predetermination of the Government ’ s case being doomed to failure . It also appears that if that witness could not be traced, the case would have had to be closed, with the option of allowing the applicant (if he insisted) to continue working as a prosecutor. It would have been some accomplishment in vindicating the individual ’ s “rights” at any cost.

22. To enhance the cogency of the “necessary step” exercise, the latter is presented in context . The domestic courts are criticised for not having undertaken a “thorough examination ... of ... whether or not [the authorities] had incited the commission of a criminal act” by “establish[ing] in particular the reasons why the operation had been mounted, the extent of the police ’ s involvement in the offence and the nature of any incitement or pressure to which the applicant had been subjected” (ibid, § 71). Some of these elements were in fact examined by the Supreme Court. For instance, although the trial court ’ s judgment indeed “did not contain any discussion of the authorisation and implementation of the [criminal conduct simulation] model”, that was dealt with by the Supreme Court ( ibid., §§ 24 and 27). It is a matter of assessing whether the “examination” was “thorough” enough. But even if it was not, all the points of criticism mentioned above boil down to one single point – that the person in question “was never called as a witness in the case since he could not be traced”. It is hardly possible to agree with the straightforward inference that because of that person ’ s absence the defendant was deprived of the “opportunity to state his case on each of these points ” (ibid., § 71; my emphasis), especially as all “other” points are not “other” at all but derivatives of the one “necessary step” not taken.

23. What is striking in Ramanauskas v. Lithuania (cited above) is that the Court ’ s reasoning passes over in silence one circumstance which the domestic courts saw as having been of some importance (which does not bail out their domestic law-based, but nevertheless ill-fated stance on the “irrelevance” of incitement ). T he fact is overlooked (or deliberately ignored?) that the applicant was not just anybody, but a prosecutor . Prosecutors are supposed to be immune to incitements, aren ’ t they? If such attempts are made, they must report the matter so that the appropriate measures can be taken. If a prosecutor had not reported the illicit offer, but had accepted it, why should his or her version be given effectively more credence than that of the one who had reported? With hindsight, now it is clear (also for the Court) that the applicant ’ s first publicly disclosed “entrapment” had hardly been set for anyone ’ s sick or squint-eyed amusement, so there must have been reasonable grounds for not giving his version preference over that of the prosecution. So what if the Supreme Court had not fairly admitted that it was not able to “uncover the truth” and had not made it clear that it did not care whether there was an incitement, but had deflated the applicant ’ s story as “wholly improbable”, the one backed by no evidence other than circumstantial, whereas that of the prosecution was supported by some (not only “rumours”) ? Would it have satisfied the Court ’ s rigid standard of “not wholly improbable”? Who knows .

IV

24. As has been mentioned, in the instant case the applicant again failed to provide any coherent explanation as to why he attracted the special services ’ attention. Also his explanations regarding the purpose for which he had accepted the money (“legal services”?! what “services” precisely ?! had he already provided any ?!) were incomparably more grotesque than those to which he resorted in his first case. (In addition, lightning never strikes twice in the same place, as the saying goes; however, that would indeed be secondary.)

25. But was the applicant ’ s story “wholly improbable” in the literal sense of the “not wholly improbable” clause? Of course not, because it is not unnatural . There will always remain a probability (even if it will amount to a fraction of a unit where the decimal separator is followed immediately by several zeros) that the applicant did indeed blather this and that to his collocutor, the prisoner (as well as to his intermediary), without really contemplating doing anything illegal, or that he indeed intended to provide his “client” with certain “legal services”, albeit disclosed neither to the domestic courts nor to this Court, for which he was grossly, even disproportionally (see paragraph 68 of the judgment ) “remunerated” in advance without any formal contract. The formalistic approach would suggest that the applicant might have acted carelessly and irresponsibly, but nonetheless not illegally: the fact that people sometimes act carelessly or irresponsibly is a fact of life. Nothing, virtually nothing in the applicant ’ s version of events contradicts the laws of nature (or of society, for that matter).

26. What that version is at odds with as “wholly improbable” is the reasonable mind, the shared knowledge of how things are in life. “ Wholly improbable” in the instant case (exceptionally?) has been understood as “ reasonably improbable”. The Court thus has not applied the “not wholly improbable” clause uncritically – and for a good reason. When the applicant had gone to see the prisoner who wanted to be released on probation, he clearly knew the reason for which they had to meet. The applicant had been the first to mention the “price” to his collocutor. Unsolicited, he had also mentioned an earlier instance of bribery attempted by him (whether it was true or not). His phone conversations with his “client” are most revealing. And so on, and so forth. Finally, the applicant had accepted the cash. All this (in a nutshell) renders his story about entrapment contrary to common sense. Hardly anyone, t herefore, would be able to believe it, not even ECtHR, even though it was the Court ’ s Grand Chamber which authored the extremely restrictive (from the point of view of the respondent governments) or extremely permissive (from the point of view of the applicants) “not wholly improbable” clause which translated, for the purposes of application in the agent provocateur cases, the pragmatic adage in dubio pro reo in an utterly peculiar way. This time, however, the Court was not overly legalistically na ï ve.

V

27. Since Ramanauskas v. Lithuania (cited above), the Court ’ s case-law on agents provocateurs has undergone at least one important development. No, the Court has not formally renounced the “not wholly improbable” clause. It is, perhaps unfortunately, not in the habit of the Court to explicitly indicate that a certain part of its doctrine has been effectively overruled by the subsequent case-law; that somewhat antiquated case-law is therefore fragmentarily cited in later judgments and decisions, as if it still maintains the same jurisprudential force. (As already mentioned, the present judgment also cites the Grand Chamber ’ s “not wholly improbable” clause, which was never formally rephrased.)

28. The development discussed here pertains to the new methodology for examining agent provocateur complaints. In 2017, the Court crystallised something which up to then had been present in its case-law only in a sketchy, fractional and rudimentary way: the so-called Matanović methodology (see paragraph 62 if the judgment ). In Matanović v. Croatia (no. 2742/12, 4 April 2017), the Court found that the establishment “with a sufficient degree of certainty” that the “domestic authorities investigated the applicant ’ s activities in an essentially passive manner and did not incite him or her to commit an offence” (substantive test of incitement) dispenses the Court from the need to proceed with the examination of the way the domestic courts dealt with an applicant ’ s plea of incitement, including compatibility with the Convention standards pertaining to a fair hearing (procedural test of incitement). If the Court is convinced that no entrapment took place, “that will normally be sufficient for [it] to conclude that the subsequent use in the criminal proceedings against the applicant of the evidence obtained by the undercover measure does not raise an issue under Article 6 § 1 of the Convention (ibid., § 133; my emphasis; also see Grba v. Croatia , no. 47074/12, 23 November 2017 (not yet final) ).

29. The Court used to dismiss complaints regarding alleged incitement on the grounds that no incitement had taken place even before the Matanovič methodology was set out (see, for example, Eurofinacom v. France (dec.), no. 58753/00, 24 June 2003; Kuzmickaja v. Lithuania (dec.), no. 27968/03, 10 June 2008; Trifontsov v. Russia (dec.), no. 12025/02, 9 October 2012; and Lyubchenko v. Ukraine (dec.), no. 3460/05, 31 May 2016). But since Matanovič v. Croatia (cited above), not only has the possibility of such dismissal become more veracious, but the tendency may emerge that such complaints are dismissed as manifestly ill-founded (in accordance with Article 35 §§ 3 (a) and 4 not only by a seven-member Chamber, but by a three-member committee (see, mutatis mutandis , Mills v. Ireland (dec.), no. 50468/16, 10 October 2017). If, however, the application is not manifestly ill-founded, the Court proceeds with its examination on the merits under the substantive test of incitement, but having established that there was no incitement, does not carry out the procedural test.

30. Mr Ramanauskas ’ second application could be declared manifestly ill-founded in the same way as the applications mentioned in the previous paragraph. But the Matanović methodology leaves leeway for examination on the merits, as in the instant case, of even a poorly substantiated claim that a provocation had been staged against the incitement-prone, but otherwise allegedly law-abiding applicant, and the adoption not of a decision as to the (in)admissibility of the application but of a judgment where no violation of Article 6 § 1 is found. On the other hand, if there had been no incitement, a further examination of the complaint would have been excessive. Further (though this is not decisive), the notion of “normality” contained in the formula “ that will normally be sufficient” perhaps covers the abnormality ( notably, but not exclusively, in the eyes of societies and institutions faced with real difficulties in countering corruption from places other than ivory towers) of the same applicant ’ s follow-up case regarding the similar complaint, after his victory in the first case effectively braced him to stay incitement-prone.

31. The Court limited itself to the substantive test of incitement and did not undertake the procedural one. No examination undertaken of the procedural facets could in any way have altered the finding that there had been no incitement. But had the Court nevertheless decided to proceed with the procedural test, the applicant ’ s allegations would have appeared to have been no more convincing than those examined under the substantive test. The applicant pleaded not guilty before the domestic courts (see paragraphs 24 and 25 of the judgment ). He accused two other persons of inciting the commission of the offence and claimed that one of them, the prisoner, had not been allowed to use a voice-recording watch, by which their conversations had been recorded, in prison and that he must have acquired one from the authorities. These allegations have been verified by the domestic courts (see paragraphs 27 and 29 of the judgment ). Although the applicant had an effective opportunity to challenge the use of the evidence acquired from that prisoner ’ s watch, he did not put forward any argument against the authenticity or veracity of the information obtained from it, but limited his objection exclusively to the formal use of such information in evidence during the proceedings. These arguments were given due consideration by the courts (see paragraphs 25, 27 and 29 of the judgment ). The fact that the applicant was unsuccessful at each procedural step does not alter the fact that he had an effective opportunity to challenge the evidence and oppose its use (see, among many authorities, Dragojević v. Croatia , no. no. 68955/11 , § 132, 15 January 2015 ). The domestic courts are, in principle, better placed to judge the reliability of evidence and it s compliance with domestic law. What is of no less importance is that the impugned evidence was not the only evidence on which the conviction was based. The trial court took into account the applicant ’ s statements and examined them against the testimony of the alleged instigators and of the officers of the Special Investigation Service, but also against those of the applicant ’ s co-accused (who pleaded guilty and whom the applicant for some reason had not accused of incitement), as well as against the evidence obtained after the authorisation of the actions simulating criminal conduct (see paragraphs 27 and 29 of the judgment ); it also had regard to the applicant ’ s entrapment plea, heard witnesse s, and analysed voice and video ‑ recordings and transcripts of the conversations between the applicant and the alleged instigators (compare Lagutin and Others v. Russia , nos. 6228/09 and 4 others, § 101, 24 April 2014). The court of appeal responded to the applicant ’ s complaints of entrapment by once again analysing the audio-recordings, questioning the witnesses, who could testify on the issue of incitement, and they were heard in court and cross-examined by the applicant and his lawyer (see paragraph 29 of the judgment ). The Supreme Court analysed the applicant ’ s arguments and provided relevant reasoning for refusing his contentions (see paragraph 29 of the judgment ). To sum up, the applicant ’ s plea of incitement was adequately addressed by the domestic courts, which took all the possible “necessary steps” to “uncover the truth” and to eradicate the doubts as to whether the applicant had committed the criminal offence as a result of incitement by an agent provocateur .

32. It appears that the Lithuanian authorities, including the courts, learned the lesson taught by Ramanauskas v. Lithuania (cited above). They took cognisance and made good use of the general principles set out in that judgment. The applicant, who significantly contributed to the setting out of the relevant principles by providing the Court with an opportunity to look into his first publicly disclosed story of bribery and the way in which Convention law should be applied to his ventures, seems to have drawn other conclusions from his first case. But impunity should have its limits.

VI

33. The instant judgment also teaches lessons – not only the domestic authorities or the potential succumbers to provocations. The Court itself should also be drawing conclusions from its case-law. “A time to cast away stones, and a time to gather stones together” (Eccles. 3:5).

34. One conclusion may be that the general principles applicable in agent provocateur cases must be revisited. No fundamental overhaul of the doctrine is necessary. But the “not wholly improbable” clause must be tempered, its wording must be toned down. Meant to provide guidance in the alleged entrapment cases, that clause, in its literal reading, has, so to say, entrapped the Court itself, with its great reliance on repeated verbatim citations from its own case-law. By mechanically migrating from one case to another, it reinforces the ivory tower recommendation to take at face value even the reasonably improbable stories of the alleged victims of incitement and thus to mock of justice.

35. If (and when) the general principles applicable in agent provocateur cases are revisited, there should come into being another leading judgment (or decision), preferably, of the Grand Chamber, on which the subsequent agent provocateur cases should be modelled. Ramanauskas v. Lithuania (cited above) can no longer fully perform this function, in particular because now it can be seen, at least by some, as besmirched by the fact of being that judgment which in fact reassured the applicant of his impunity and planted in him a hope (though eventually a baseless one) of the Court ’ s na ï vet é . The fine-tuned doctrine should send a very clear message that the Court does not shut its eyes to the real difficulties which the domestic authorities encounter when countering crime in general and corruption in particular. A mere declaration (words which also migrate from case to case) that the Court is “ mindful of ... the difficulties of the task of investigating offences” or that it is “aware of the difficulties inherent in the police ’ s task of searching for and gathering evidence for the purpose of detecting and investigating offences” (see paragraph 9 above) does not suffice. Worded in most general terms, this wishy-washy assurance does not even mention corruption, which, as a rule, is a clandestine activity. In this context, the Court should ask itself in each and every case pertaining to alleged incitement: what if the authorities had not performed the impugned operation? Or: what results would their inaction have brought about (on this I refer to my dissenting opinion in P ă tra ş cu v. Romania , cited above )?

36. If there are indeed factual and legal grounds for finding, in an agent provocateur case, a violation of the Convention, so be it. A breach of rights is a breach of rights. It would be unfair and unprofessional to defend the authorities a priori , also in view of the fact that provocation against opponents in particular and political justice in general are a growing reality in some states. Nevertheless, in the assessment of “his word against theirs” contrapositions in agent provocateur cases, the fact of commission of crime should be given some prominence. Also, it should matter, in particular in corruption cases, whether the alleged incitement (provided that it is established that it took place) included the offer of “reward” or not. These circumstances certainly cannot attain the status of evidence decisive for the determination of whether there was an incitement to commit a criminal offence , but they should not be completely dropped off the scales and treated as being no evidence at all. This is especially pertinent to cases where the alleged victim of an incitement is, say, a prosecutor (a judge, a law enforcement officer, etc.), whose obstinate refusal and/or inability to resist incitement undermines the very raison d ’ ê tre of his or her being in the respective position and whose treatment, in the Court ’ s case-law, as a “victim” on a par with an incitement-prone “man in the street” frustrates and erodes the individualised examination of facts by the courts as an inherent feature of modern Western (in particular European) law and brings us one step back to its earlier condition of the formal (mandatory and indiscriminate) assessment of evidence (on this I refer to my dissenting opinion in Lisovskij v. Lithuania , no. 36249/14, 2 May 2017) and a step away from the ideals of substantive justice. This calls for a more nuanced approach, commensurate with the professional and civic function of alleged “professional” victims. If the undifferentiated application of the “not wholly improbable” clause is the standard, then, in the context of the instant case, why shouldn ’ t the Court be no less clement to an advocate ’ s depravity than to that of a prosecutor and not give an advocate ’ s version of events the same benefit of doubt as to that of a prosecutor? However, t he well-meant belief that human beings are equally inci tement-prone must not be so all ‑ embracing. The Court ’ s big-heartedness should allow for at least some differentiation between a “man in the street” and state officials (prosecutors, policemen, judges, politicians, etc.), where the latter are bound to the requirement to resist incitement to commit a criminal offence in some stronger way than the former and must be more incitement-resistant, more unyielding to “outside influence”. If being a professional office-holder has any added value, then one ’ s added responsibility is yet another facet (out of many) of that added value, and its logical consequence. If this is so, the clause which (as I would like to hope) replaces the “not wholly improbable” clause, even if more or less easily applicable to a layman, should have some in-built reservations when applied to officials.

37. In particular, where the incitement is found to have taken place against, say, a judge, a prosecutor, etc., who had succumbed to it, all the pros and cons of awarding an outstanding amount in respect of damage to the victim of that incitement (who is, in fact, a victim of his own irresponsibility, cynicism and greed) should be considered – as comprehensively as possible. In such cases a formalistic approach is the enemy of a just one, and “equal justice” may turn into a caricature of justice. Even admitting that the finding in Ramanauskas v. Lithuania (cited above) was, overall (but not in every passage of argumentation), a reasoned one, the amount awarded to the applicant was barely explicable (to put it mildly). It drew gasps from many in the law enforcement and the judiciary. True, part of the amount awarded was compensat ion for “indisputabl[e] ... non ‑ pecuniary damage, which cannot be compensated by the mere finding of a violation” ( ibid., § 87), and the other part was compensation for the loss of earnings sustained by the applicant, because owing to his conviction he could no longer work as a prosecutor (at least while he was serving his sentence). Did the Court see the non-termination of the applicant ’ s work as a prosecutor as a value which had to be compensated, and the non-pecuniary damage sustained by him because of his falling victim to a provocation not counterbalanced by any non-pecuniary or pecuniary damage sustained by society? Raising these questions and finding fair answers to them is yet another lesson to be learned (also) by the Court, albeit at some price and somewhat belatedly. Time will show whether or not this is mere wishful thinking.