Stephen Karganovic | Besides “satellite photos” (which turned out not to be that at all and are unavailable for expert scrutiny anyway), autopsy reports prepared between 1996 and 2001 (when exhumations of Srebrenica-related mass graves were abruptly terminated because there were none left to open and the numbers generated up to that point did not look good at all), the notoriously flawed evidence of “Star Witness” Dražen Erdemović (whose latest performance was at the Karadžić trial on 27 and 28 February 2012), and ICTY’s latest gimmick –independently unverifiable DNA matches (introduced at the Popović at al. trial and publicised vigorously ever since), intercept evidence is another very important evidentiary component of the Hague trials.
A thorough analysis of the spurious use and rampant abuse of “intercepted conversations” to buttress allegations of criminal guilt before various ICTY chambers is yet to be written. But the notes that follow should leave no doubt that this form of evidence, on which the Prosecution routinely and heavily relies to impart an appearance of undeniability and precision to its accusations, like most things that go on in that strange judicial institution, is not all that it is cracked up to be.
The simulacrum of solemn deliberation in which some ICTY chambers indulge for the record would suggest even to a confirmed cynic that rigorous professional standards are practiced and that great care is taken to observe every procedural nicety for the benefit of the accused. A recent ruling rendered in the Karadžić case appears to support such an impression:
On 17th of January, 2012, during the testimony of Witness Pyers Tucker, the Chamber admitted associated exhibit at P4230, a summary of three intercepted conversation from March 1993, between, number one, General Mladić and the accused; number two, Vinko Pandurevic and General Milovanovic; and number three General Mladić and Vinko Pandurevic. In the relevant portion of his amalgamated statements discussing Exhibit P4230, Mr. Tucker says “I have been told this is the record of three conversations produced by Croatian intelligence services” and then proceeds to discuss how the intercepted conversation relate to a meeting he had on the 11th of March, 1993, with General Morillon, Branko Grujic and Vinko Pandurevic.
The Chamber recalls its previous rulings that intercepts are a special category of evidence which, before being admitted, requires further evidence about their authenticity and reliability from sources such as the relevant intercept operator or a participant in the intercepted conversation. The Chamber considers that summaries of intercepts such as those in P4230 fall into the same category and therefore that Exhibit P4230 was not sufficiently authenticated for it to be admitted through Pyers Tucker. The Chamber thus reconsiders its decision of 17th of January, 2012, to admit P4230 and orders that the documents be marked for identification as MFI P4230. [Prosecutor v. Karadžić, Transcript, p. 23968-9]
A similar impression of scrupulous correctness is cultivated by the Karadžić Chamber a bit further on. The Prosecutor, Mrs. Edgerton, tries to introduce some intercept evidence allegedly involving General Mladić, but is sternly rebuked by Judge Kwon for failing to produce a satisfactory foundation:
MS. EDGERTON: The second conversation that’s no longer on the screen in front of us, Your Honour, involves General Mladić speaking with an intermediary from the VRS Main Staff. General Mladić is not heard, but the interlocutor has been identified as General Mladić. [Prosecutor v. Karadžić, Transcript, s. 25316]
And further on:
JUDGE KWON: Ms. Edgerton, last year on 4th of February in the decision of judicial notice decision, we ruled as follows: “Therefore, declaration from persons who are neither participants in the conversation themselves nor intercept operators are not sufficient for the purpose of establishing an intercept’s authenticity. The Chamber is thus not satisfied that the authenticity of the following intercept was sufficiently established.”
I think this case should be no different. In order for you to tender that intercept, it should have been discussed through [Prosecution investigator] Mr. Blaszczyk, who at least could have provided at least some foundation as to how this intercept came into the possession of the Prosecution. [Prosecutor v. Karadžić, Transcript, s. 25324]
Fair enough. But have ICTY chambers been universally reluctant to accept dubious intercept evidence? What real guarantee do we have that even the Karadžić Chamber is not merely posturing in its apparent adherence to high standards? The verdict will tell.
In the meantime, we do have some indications of how this type of evidence has been received and treated by ICTY chambers in past cases. A short review is in order.
In the Krstić trial judgment, already in the fourth paragraph the Chamber sets the stage for the benevolent consideration of every conceivable variety of evidentiary materials:
“The Trial Chamber draws upon a mosaic of evidence that combines to paint a picture of what happened during those few days in July 1995.” [Prosecutor v. Krstić, Par. 4]
A concrete, and in the specific context of the Krstić case vitally important application of that “mosaic” principle, appears in due course as the Chamber discusses how it reached its conclusion that the Bosnian Serb Army had in its custody the number of prisoners that would have been required to make the mass slaughter that was attributed to it possible:
There are also fragments of information from VRS communications about the possible magnitude of the executions. An intercepted conversation, at 1730 hours on 13 July 1995, indicates that about 6,000 men had been captured from the Bosnian Muslim column by that time… Other intercepted VRS conversations reveal that, on 15 July 1995, midway through the executions, at least 3,000-4,000 Bosnian Muslim prisoners were being detained by the VRS.171 Further, on 18 July 1995, two unidentified Bosnian Serbs were heard in an intercepted conversation reflecting on the recent events in Eastern Bosnia, including matters relating to the Bosnian Muslim column.172 One participant said that of the 10,000 military aged men who were in Srebrenica, “4,000-5,000 have certainly kicked the bucket”. [Ibid., Par. 83]
This evidence, based, as the Chamber says, in great part on intercepted communications leads in the next paragraph to a very dramatic conclusion:
“The Trial Chamber is satisfied that, in July 1995, following the take-over of Srebrenica, Bosnian Serb forces executed several thousand Bosnian Muslim men. The total number is likely to be within the range of 7,000 -8,000 men.” [Ibid., Par. 84]
The serious cumulative impact of that “mosaic”, including its intercept component, is clearly reflected in a further paragraph of the Krstić Judgment:
“The Trial Chamber finds that, following the takeover of Srebrenica in July 1995, the Bosnian Serbs devised and implemented a plan to execute as many as possible of the military aged Bosnian Muslim men present in the enclave.”[Ibid., Par. 87]
Of course the Krstić Chamber was obliged to consider, pro forma at least, the reliability of the evidence on which it relied so heavily to form its “picture mosaic”. That is done in Pars. 105 et passim where the issue of reliability of intercept evidence is addressed.
To paraphrase the Chamber’s narrative, intercept records were handed over to the OTP by the Bosnian government. VRS did have secure means of communication, but their use was too cumbersome, so they often used unsecured lines for expediency. Intercept evidence was relied on by OTP for key elements of its case.
After attributing such great weight to intercept evidence in the presentation of the prosecution case, the Chamber concludes reassuringly:
“The Trial Chamber was told that all possible measures were taken to ensure the accuracy of the transcribed conversations.” [Ibid., Par. 109]
Noting that Defence expert, General Radinović, had expressed some doubts about the reliability of this type of evidence, the Chamber takes the position that it “accepts that the OTP did in fact diligently check and cross-reference the intercept material as part of the ‘intercept project’,” which should be sufficient to allay the general’s concerns. The Krstić Chamber goes on to say that “The Trial Chamber accepts that the OTP did in fact diligently check and cross-reference the intercept material as part of the ‘intercept project’. In order to determine whether the material was reliable and genuine, the OTP looked at the internal consistency between the notebooks and the printouts of each conversation. Transcripts of a single conversation, which was recorded by two or more interceptors, were also compared. The OTP also embarked on a process of corroborating the intercepts with information obtained from other sources, such as documents acquired from the VRS, the RS Ministry of Defence and UNPROFOR, as well as aerial images.” [Ibid., Par. 114]
It is difficult to avoid the impression that when it comes to presenting such important evidence the Chamber leaves it to the Prosecution to monitor itself and that it is quite happy with the results of that arrangement.
Should there be any residual doubts advanced by sceptics such as General Radinović, the Chamber draws its trump card:
“A former OTP employee assigned to the ‘intercept project’ testified that, as a result of this corroboration process, she became convinced that the intercepts were ‘absolutely reliable’… the former OTP employee [helpfully identified as Mrs. Frease] who appeared before the Trial Chamber testified with ‘absolute certainty’ that the dates ascribed to the individual conversations were accurate.” [Ibid., Par. 114]
That surely should suffice to put all doubts to rest. Unsurprisingly, Prosecution military expert Richard Butler endorses the Chamber’s view. [Ibid., Par. 115]
To make things even more iron clad, it was not just Prosecution personnel at The Hague who took great care to guarantee the integrity of the intercept evidence gathering process. It turns out that Bosnian Muslim technicians at the source were equally professional and conscientious:
“All possible measures were taken by the Bosnian Muslim interceptors to ensure the accuracy of the recorded conversations, as would be expected in any prudent army. This fact was reinforced by the measures taken by the OTP to verify the reliability of the intercepted evidence as part of the ‘intercept project’.” [Ibid., Par. 116]
To what degree that is really true we shall soon see when we review the statement one of those intercept operators gave to the Office of the Prosecutor.
In the Blagojević and Jokić case, we find similar dicta. Most notably, the Chamber announced up front that “the Trial Chamber is convinced that the intercept-related evidence admitted is a reliable source of information.” [Prosecutor v. Blagojević and Jokić, Par. 30] Defence objections to that conclusion are summarized, but they are promptly overridden: “The Defence of Dragan Jokić argued that the intercept transcripts were taken down by unknown personnel or personnel with a history of unreliable transcriptions lacking sufficient training, that substandard equipment was used, that by not providing original tape recordings the Prosecution was effectively submitting hearsay evidence, which ought not to be admissible.” [Ibid., Footnote 72]
It is nevertheless important to remember, when the Chamber says that “bearing in mind the testimonial evidence and the very large amount of documentary evidence, the Trial Chamber cannot find that it is necessary to have access to the original audio recordings of the intercepts”, that of more than 100 intercepts used in the Krstić case only one was audio.
In the Popović et al. judgment rendered in June of 2010, the chamber considered various factors in regard to 213 intercepts admitted prior to concluding that “Trial Chamber has found the intercepts to be overall probative and reliable” [Prosecutor v. Popović et al., Par. 66]
The Trial Chamber’s procedure was to examine whether, “based on the totality of the evidence, a reasonable trier of fact could find the intercepts to be what the Prosecution purports them to be—a contemporaneous record of intercepted VRS communications.” The Trial Chamber says that in the process it “considered the testimony of several witnesses relating to the intercepts, such as intercept operators, an expert in radio relay communications, and a Prosecution analyst. It considered all challenges made by the Defence, including the theory that the intercepts had been fabricated, evidence relating to the chain of custody, and the general lack of audio recordings. In sum, the Trial Chamber concluded that the Prosecution had established that the intercepts as a whole were prima facie relevant and probative.” [Ibid., Par. 64]
It seems that all those challenges turned out to be without foundation and that the record keeping practices of the Bosnian Muslim intercept operators in besieged Srebrenica were meticulous and satisfactory in every way. In fact, the Chamber made its determination to view that evidence favourably “particularly in light of the evidence given by the intercept operators” [Ibid., Par. 65]
Two examples will put into bold relief some of the issues raised by the high level of receptivity shown by various ICTY chambers to intercept evidence. They suggest that in regular criminal cases in most national jurisdictions the way this evidence was gathered would almost certainly be found questionable by non-political judges.
The reference in Par. 383 to a key purported intercept resulting in information that was vital to the construction of the factual underpinnings of the Popović judgment and therefore to the credibility of the judgment as a whole is an apt illustration. It concerns a 13 July 1995 intercept indicating the capture by Serbian forces of about 6,000 Srebrenica Muslim POWs. It is the only clear reference to the number of POWs captured at that point. Without establishing such a fact, the case against the defendants would be seriously undermined because without the prior capture of thousands of Muslim prisoners, executions on such a huge scale could not have occurred. This is the Trial Chamber’s summary of the intercept’s content:
“A conversation intercepted at 5:30 p.m. on 13 July indicates that approximately 6,000 Bosnian Muslim prisoners were detained in the Bratunac area at three locations, with about 1,500 to 2,000 men in each location. One of the locations appears to be the football field at Nova Kasaba, another was ‘up there where the checkpoint at the intersection is,’ and a third was ‘halfway between the checkpoint and the loading place.’ In this context, the Trial Chamber is of the view that one of the places is Sandici Meadow and the other Nova Kasaba.” [Ibid., Par. 383]
But a review of the actual intercept, as presented by the Prosecution and available as a trial exhibit and a document from the Tribunal data base, raises some serious concerns. The interlocutors are designated as X and Y, meaning that they are completely anonymous and that even their existence cannot be verified, not to mention the possibility of cross examining them. Other than the sheet of paper with some writing on it, purporting to be an intercept of their conversation, we do not have any objective evidence from first hand sources that such an interchange even took place or, if it did, that the participants were in a position to know what they were talking about. The latter point is of critical importance. A key conclusion about the number of prisoners is drawn exclusively from the conversation attributed to them and is based on the assumption that they were competent reporters of the relevant facts. (See Exhibit 1)
But even if we were to credit this piece of evidence, it is still susceptible to variable interpretations. According to the Prosecution, and the Chamber accepts that, anonymous individuals X and Y had a conversation at 5:30 pm on 13 July 1995 where Y informs X that at each of three different locations there were “about 1,500 to 2,000” prisoners, or a total of “6,000”. Even if we accept the conversation’s authenticity, the conversation does not support “beyond a reasonable doubt” the interpretation that the Prosecution and the Chamber attribute to it about the total number of captured prisoners. For each of the cited locations, a range of 1,500 to 2,000 captured prisoners is given. Assuming that the Chamber chose, for whatever reasons, to give credence to the information contained in the intercept of nameless participants, it still had the option of choosing the lower total of about 4,500 POWs. Since in the purported intercept there is no claim of an accurate headcount, the Chamber would have acted reasonably by erring on the side of caution. But no, four and a half thousand captured prisoners, though a considerable number, will not do because it is still too far from the required total of 8,000 “victims of genocide”. The court therefore simply adds up maximum numbers from an unsubstantiated document and uses that as the basis for its conclusion. Mass murder now is possible because approximately the projected number of victims were already under the control of the executioners.
Another curious use of “intercept evidence” can be cited. We are turning our attention again to the Krstić trial, and the evidence of prosecution military expert Richard Butler about a July 18 1995 intercept from which he draws equally significant conclusions. The relevant section of his testimony is in the Krstić trial transcript, p. 5205. Essentially, using an English translation, Butler offers his interpretation of the July 18 intercept where he claims that the execution of several thousand Muslim prisoners is described in coded terms as their having “kicked the bucket”. “I can only assume,” Butler testified, “that this was a reference to Muslim men who were transferred to the Zvornik brigade zone of responsibility, where they were executed.” [Ibid., Transcript, p. 5205] The issue is important because the interlocutors refer to 4,000 to 5,000 persons.
Two important observations are in order. First, Butler admitted that he does not speak Serbo-Croatian and therefore would have been unable to follow the conversation in the original language of the speakers if it were shown to him. Second, there is no record of the existence of a Serbo-Croatian original in the ICTY data base.
In his expert opinion, based on the version of this intercept that was shown to him by the Prosecution (and, as it turns out, accepted by the Chamber in its judgment), Butler advanced the view that the phrase “kicked the bucket,” which is used in the English version, signifies a mortal outcome. But both Serbo-Croatian and English speakers may question that. First, there is no expression equivalent to “kick the bucket” that native Serbo-Croatian speakers in the intercept might have used that comes to mind. Since there is not even a purported Serbo-Croatian original of this key conversation, what they actually said to each other is something that we will never know. Second, from the standpoint of the English language, in which Butler presumably is fluent, “to kick the bucket” is not customarily used to describe violent death. So at a minimum some serious questions can be raised not just about the authenticity of this intercept, but about the interpretation attached to it as well.
Which brings us to the central issue: how reliable are the intercepts that are used at ICTY? A corollary question is how trustworthy are the judicial conclusions that are drawn based on such evidence.
Dramatic, but completely ignored, answers to these questions were provided by Emir Osmić, one of the Bosnian Muslim Army’s intercept operators who were keeping tabs on Srebrenica radio traffic within the Army of the Republika Srpska [VRS]. In a statement given to Office of the Prosecutor investigators on 6 May 1999, Osmić describes in detail his duties as a BH Army intercept operator and the way he and his colleagues performed their job. This is the way he depicts that process:
When my shift on duty was over, I would hand my notebooks over to the commander who would then type them up, and return them to the next shift to continue to use. When the notebooks were filled with notes the commander would take them and, I believe, carry them over to the division headquarters, after which they would send them to the archive or something like that. I had nothing to do with what went on with them after I turned them over to the commander. The tapes that we used we kept reusing because we did not have enough tapes. We used tape-recorder tapes and we would tape over the previously recorded material if during the shift the tape ran out. I am not sure if a single one of the tapes on which we recorded important conversations was preserved. The one thing I do recall is that we had to use the same tapes over and over again because we did not have enough of them. I have no idea what happened to those tapes. [Statement of Emir Osmić, EDS file number 0084 8061]
The situation we have here, according to operator Osmić, is that no physical evidence for the incriminating conversations between VRS officers and personnel is available. The same tapes were used repeatedly, and with each use what was previously recorded would be erased. Written notes of what was heard and before erasure supposedly recorded were made, but they ended up in some black hole at the headquarters and “the archive”. Between then and their popping up in court at The Hague, there seems to be no verifiable chain of custody, no assurance that they were not tampered with by the Bosnian Muslim authorities, who had them under their control and were not a neutral party in the ICTY proceedings.
These facts should be assessed against the backdrop of pious protestations by some ICTY chambers about “authentication” that were cited at the beginning. How consistent is the judicial branch of ICTY in adhering to the enunciated principles? If they were consistent, would that not be reflected at key junctures where intercepts were being used to buttress major elements of the prosecution case? We saw some examples which indicate that it may not have been.
We may, for the moment, set aside the issue of purported audio intercepts made by foreign intelligence agencies during the Bosnian war. They were used also in the Hague proceedings but to a relatively minor extent (just as in the case of “aerial photographs” considerations of national security were adduced to thwart the subjection of this form of signals intelligence to thorough independent evaluation). But from both the quantitative and qualitative standpoint, their courtroom role was not nearly as prominent as that of locally produced intercepts, originating from the monitoring sources of the Army of Bosnia and Herzegovina. To the extent that they have been used (e.g. at the Popović et al. trial) it was under limitations which constituted a severe handicap for the defence. It is important to bear that in mind because just as with the DNA evidence, modern technology makes it extremely easy to falsify audio material. When effective expert analysis of the proffered data is precluded, the purported evidence is as good as useless.
We believe that we have offered sufficient reasons for the inherent untrustworthiness of intercept data used at ICTY (and, by extension, before its Sarajevo clone, the State Court for War Crimes of Bosnia and Herzegovina, which uses identical material) to warrant the sounding of alarm bells. We need more than the disingenuous assurances of intercept operators of one of the warring parties and the “absolute certainty” of Mrs. Frease. The trial records of these courts, in particular with respect to Srebrenica where the greatest concentration of intercept evidence abuse may be noted, should be carefully combed and the authenticity of all intercepts the court relied on should be subjected to thorough professional scrutiny. ICTY (and Sarajevo court) intercepts which fail to meet fundamental standards of admission in ordinary criminal cases in national jurisdictions should be excluded from consideration. The verdicts of both courts should be modified as necessary to reflect the exclusion of such evidence.
Stephen Karganovic is the President of Srebrenica Historical Project.