1. Introduction
It would be difficult to point to an area of international criminal procedure where the amendments of the relevant Rules of Procedure and Evidence have been more responsive to developments in jurisprudence than that of the use of written witness statements in lieu of oral testimony. The deletion of the preference for orality under Rule 90(A) of the Rules of Procedure and Evidence of the International Criminal Tribunal for the Former Yugoslavia (ICTY) and the introduction of Rule 92 bis, which allows for the admission of written witness statements that do not go towards proving the acts and conduct of the accused, were clearly a reaction to two Appeals Chamber judgements in Kordić and Čerkez on the admissibility of a statement of a deceased witness.Footnote 1 Rule 92 bis, as originally enacted, was less strict than the Rule 94 ter on affidavits which it replaced, in that it did not require statements to be formally made in accordance with domestic law if the witness was dead – chambers could still rule such evidence to be admissible, considering other factors like reliability. The requirement that affidavits be introduced solely for corroboratory purposes and that the statements be admitted before the witness whose testimony would be corroborated had testified were also removed.Footnote 2 Over time, Rule 92 bis has been amended further. The requirement of giving the other party 14 days’ notice has now been removed, as has the provision that the Trial Chamber should still be able to order the appearance of the witness for cross-examination, although that option is still open to the chamber.Footnote 3 Rule 90(A) of the Special Court for Sierra Leone (SCSL) was amended to remove the preference for orality in 2003,Footnote 4 and in 2004, the Court adopted its own Rule 92 bis.Footnote 5 The International Criminal Tribunal for Rwanda (ICTR) retains a preference for orality in its Rule 90(A) although it too has adopted a Rule 92 bis on prior statements not going towards the acts and conduct of the accused.Footnote 6
In a similar trajectory, shortly after the Appeals Chamber in Milošević held that if a witness were present in court and available to attest to the veracity of a prior statement, the prior statement could be introduced under the flexible lex generalis of Rule 89(F), subject to a right of the parties to cross-examine the witness,Footnote 7 the judges of the ICTY introduced Rule 92 ter precisely to this effect. The SCSL has an equivalent Rule 92 ter, adopted in November 2006.Footnote 8
In Milutinović in 2006, the ICTY faced a request for the admission under Rule 92 bis of the prior recorded testimony of two deceased witnesses – Ibrahim Rugova, who testified in the Milošević trial, and Antonio Russo, who had given a statement to investigators in the field.Footnote 9 The former was denied admissibility, as it went towards proving the acts and conduct of the accused, whereas the latter was admitted as it related to crime-base evidence.Footnote 10 Two months later, the tribunal adopted Rule 92 quater relating specifically to the written statements of deceased witnesses, which are no longer precluded from going towards the acts and conduct of the accused.Footnote 11 The Rugova transcript was ultimately admitted pursuant to the new Rule 92 quater.Footnote 12 The SCSL adopted its own Rule 92 quater in May 2007,Footnote 13 while the ICTR retains Rule 92 bis(C) on the statements of deceased witnesses, meaning that, in principle, such testimony should not go towards proving the acts and conduct of the accused.
The most recent addition to the family of rules is 92 quinquies of the Rules of Procedure of the ICTY, which was introduced after allegations of witness intimidation in the Haradinaj Footnote 14 and Limaj Footnote 15 trials and following the report of a working group on contempt proceedings.Footnote 16 The new rule provides that, where a witness fails to attend or appears in court but fails to testify in any substantive sense owing to ‘improper interference’ with the witness, a written transcript or statement can be introduced in lieu of oral testimony.Footnote 17 As with Rule 92 quater, such evidence is not precluded from relating to the acts and conduct of the accused, although that may be taken into account as a factor in deciding admissibility. There is no equivalent rule in the SCSL and ICTR Rules of Procedure and Evidence.
We can see, then, that rules on admissibility of evidence have become more liberal over the years and have been closely tied to jurisprudential developments. Most scholarly works on these changes focus almost exclusively on the issue of admissibility, with analyses of weight being limited to perfunctory remarks stating that judges ought to be very cautious in relying on such written testimony to establish individual criminal responsibility.Footnote 18 Such caution is warranted, in order to comport with the accused's right to examine the witnesses against him or her and to have a judgment that is not based decisively on untested witness testimony.Footnote 19 This article seeks to fill a gap in the literature by establishing whether caution has been observed in practice, by tracing pieces of evidence admitted under the newer, more liberal rules on written witness testimony from admissibility to judgment, to establish what impact, if any, the liberal admissibility rules have had in practice and whether the critique that such rules would lead to a misapplication of the standard of proof has been realized.
Nine judgments will be analysed in detail. These judgments were chosen solely on the basis of their date of issue. Cognizant of the fact that the admissibility rules have become more flexible over time, and wishing to analyse trial judgments where the most flexible admissibility rules were in force, it was decided to analyse trial judgments issued from 1 January 2011 to 1 June 2012 as a representative sample. This provided a total of one judgment from the SCSL,Footnote 20 five from the ICTR,Footnote 21 and three from the ICTY.Footnote 22 This case selection brought the added benefit of providing a cross-section of chambers composed of judges from different legal traditions, which permitted the author to test the hypothesis that judges can be biased towards their own procedural backgrounds.Footnote 23 Judgments for contempt offences fall outside the scope of this article, as do appeals judgments, which tend to give deference to the fact-finding activities of the Trial Chamber due to its proximity to the evidence and the fact that it can assess demeanour, credibility, and consistency in a way that the Appeals Chamber cannot.Footnote 24 Findings of fact may only be overturned where the Appeals Chamber holds that no reasonable trier of fact would have entered a conviction on the basis of the evidence to hand.Footnote 25
There is an obvious omission in the form of the 2012 judgment in Lubanga before the International Criminal Court,Footnote 26 as the differences in the structure of that Court's Rules of Procedure and Evidence and statutory framework do not lend themselves methodologically to an accurate comparison in the same way as the broadly similar rule structures of the other three tribunals do. However, the Lubanga judgment will be referred to throughout as providing a point of general comparison.
Furthermore, a body of material entered into the record under Rule 89(C) has been excluded from the scope of this article. The reasoning behind this was that Rule 89(C) ought to be, and generally is, the preserve of documentary evidence – that is, material that is contemporaneous to the events in question, and not statements made after the fact by witnesses, whether in court or in the field.Footnote 27 The ICTY judgements studied made particular use of this type of documentary evidence, citing military orders, minutes from meetings, hospital records, and other contemporaneous documentation, usually as corroboration to witness testimony. Statements admitted under this rule should only be used to go towards proving the credibility of the witness, and should not be used as proof of their contents.Footnote 28 However, that principle has not always been observed in practice.Footnote 29
A number of key points on the operation of the more liberal rules on written witness testimony shall be illustrated. Section 2 will show that the newer rules tend to be utilized far more frequently by the prosecution than by the defence, and will also illustrate that the rules are used far more frequently in some tribunals than in others. Some chambers have taken a rather cautious approach by ordering cross-examination of witnesses where this was technically not required under the Rules. Section 3 illustrates further caution in the Chambers’ approach to corroboration of and the weight given to written testimony, but shows that some Chambers are more wary of relying on written testimony than others in this regard. It shall be further argued that the ‘totality of the evidence’ approach can leave the observer – and even later appellate judicial panels – uncertain as to the precise weight given to a piece of evidence, and calls the suitability of broad admissibility rules in this context into question.
2. Admissibility of written witness statements
Table 1 illustrates the full scope of evidence admitted, whether viva voce or pursuant to one of the rules on written witness testimony, as part of the prosecution (P) or defence (D) case. A number of remarks shall then be made about the utilization of some of the newer rules on the admissibility of evidence.
Table 1. Forms of Witness Testimony, January 2011–June 2012
A few preliminary observations may be made on the basis of the figures. The most notable trend is the conspicuous absence of written witness statements in the record of recent cases before the SCSL and the ICTR (with the exception of the Karemera trial), in comparison to the ICTY. The structure of the Rules of Procedure and Evidence accounts for some of this divergence – the ICTR has only adopted Rule 92 bis of the more recent rule amendments, and retains a preference for orality in Rule 90(A). However, this cannot be the only factor when one considers the infrequent use of Rule 92 bis, and the fact that the SCSL has adopted all but 92 quinquies of the newer rules in this area, and has shown a similar degree of conservatism. Indeed, the only request to use Rule 92 ter before the SCSL was rejected.Footnote 32 Requests to admit evidence under Rule 92 bis before the ICTR in Nyiramasuhuko et al. and Gatete were similarly denied.Footnote 33
We might ask, then, why the Karemera trial stands as an outlier amongst contemporaneous ICTR trials in its quite prolific use of Rule 92 bis statements. The argument that a chamber's willingness to admit written testimony hinges on the legal background of the judges holds some limited weight in this instance. Trial Chamber III in the case was initially composed of Judges Byron (St Kitts and Nevis), Short (Ghana) and Joensen (Denmark), two of whom are from broadly common legal system backgrounds, while Denmark has a largely inquisitorial model. In an early decision on Rule 92 bis, Trial Chamber III, so constituted, denied the prosecution request for the admission of 71 written statements of rape witnesses in lieu of oral testimony,Footnote 34 on the basis that this evidence sought to establish the widespread and systematic nature of the rapes allegedly committed by the accused's subordinates or co-perpetrators.Footnote 35 These allegations, the chamber held, were ‘so pivotal’ to the prosecution case that it would be unfair to admit the evidence without the opportunity for cross-examination.Footnote 36 This was a rather narrow interpretation of the ‘acts and conduct’ provision, which generally accepts that the acts and conduct of one's subordinates can be admitted under Rule 92 bis.Footnote 37
Later, when Judge Short was replaced by Judge Kam of Burkina Faso (which can be loosely classified as a civil-law system), the chamber did appear to become more open to the admission of written statements; indeed, the December 2006 decision on the admissibility of written statements going towards the charge of rape as a crime against humanity was ‘reconsidered’ in September 2007 as a result of a number of changes in circumstance. These changes were: that oral witnesses had testified in court on these counts, that the prosecution was now happy to accept cross-examination of these 92 bis witnesses if the Chamber so ordered (in contrast to its 2006 ‘all or nothing’ approach), and that the renewed request of the prosecution was significantly narrowed in scope as well as copper-fastened by more evidence on the trial record than had been contained at the time of its original request.Footnote 38 A reverse in position on the ‘acts and conduct’ standard for the rape witness statements was enacted, with the chamber noting that it did ‘not agree with the Defence contention that there are elements of the statements . . . which are so pivotal to the Prosecution case . . . that their admission would be unfair’.Footnote 39 Sixteen witness statements, all of which were ultimately relied upon in the Trial Judgement,Footnote 40 were admitted without being subject to cross-examination, although three had portions redacted, and one witness later testified orally.Footnote 41 In all likelihood, the change in the composition of the chamber was but one factor leading to this divergence in interpretation. The fact that corroborating oral testimony had already been produced by the time of the second decision was likely to have been as or more decisive to the Chamber's change of position.Footnote 42
Moreover, the composition of the Chamber cannot account for the preponderance of written statements as a whole in this case. It must be noted that the defence requested the admission of more Rule 92 bis statements than in any other case studied. There were issues throughout the trial with the length of the defence witness list, particularly regarding the accused Nzirorera, who had initially proposed calling 180 witnesses viva voce and submitting 47 statements under Rule 92 bis.Footnote 43 Having been ordered to reduce his list of witnesses to 55,Footnote 44 Nzirorera sought to admit 127 witness statements under Rule 92 bis in December 2008.Footnote 45 Twenty statements were admitted into evidence, and a further 60 were declared admissible, subject to certification.Footnote 46 The accused produced certified statements for 44 witnesses.Footnote 47 Several statements were denied admissibility because of their limited relevance, probative value, or reliability,Footnote 48 while 25 were rendered inadmissible by virtue of the fact that they went towards the acts and conduct of the accused,Footnote 49 but as a whole, the Chamber showed a relatively lenient approach to admissibility, which, combined with the defence's extensive request, led to a greater admission of 92 bis statements than observed in other trials. Of the 61 witness statements (excluding transcripts) declared admissible, however, only three – those statements of witnesses Kagaba, Rukerikibaye, and Kahihura – were ultimately referenced in the trial judgement.Footnote 50
The analysis further revealed that the Rules do not always operate as one might expect them to from a literal reading of the Rules of Procedure and Evidence. For instance, Rule 92 bis allows for the admission, without cross-examination, of written testimony that does not go towards the acts and conduct of the accused, although Rule 92 bis(C) of the ICTY's Rules of Procedure and Evidence and Rule 92 bis(E) of the ICTR's Rules permit the Trial Chamber to decide that the witness must appear for cross-examination. There is no equivalent in the SCSL's Rules of Procedure and Evidence. None of the relevant sets of rules indicate what factors might be taken into account in deciding whether to require cross-examination. However, the ICTY and ICTR Rules note factors like the public interest in hearing the evidence orally, fair-trial considerations, and unreliability as reasons why the Trial Chamber might decline to admit evidence under the rule;Footnote 51 the same factors could presumably be taken into account in deciding whether to require cross-examination, but this is not explicit. The argument is frequently raised that, given the fact that it does not require cross-examination, Rule 92 bis has the potential to open a Pandora's box. As many international criminal convictions are based on extended forms of liability, it has been argued that there is a risk that ‘crime-base’ evidence that may seem extraneous to the key question of the guilt of the accused may be decisive in the ultimate verdict.Footnote 52 However, of the total number of 207 92 bis witnesses across the three tribunals listed in the table above, 73 were called for cross-examination.Footnote 53 This means that over a third of Rule 92 bis testimony was in fact subject to cross-examination by the non-moving party,Footnote 54 which makes it more akin to Rule 92 ter testimony than to the form originally envisaged for Rule 92 bis.Footnote 55
It is clear that Rule 92 ter, in turn, has been greatly utilized by the ICTY, where Rule 92 ter testimony outweighed the use of viva voce testimony for the prosecution in all trials observed. Indeed, the fact that a high preponderance of testimony under Rules 92 bis, 92 ter, and 92 quater (some 65 per cent) was introduced by the prosecution in the ICTY cases studied may lead to concerns on the principle of equality of arms. In a decision on referrals to Rwanda, the ICTR held that it would be a breach of equality of arms if the majority of prosecution witnesses were heard in person and the majority of defence witnesses testified via video link.Footnote 56 This increased utilization of written statements without examination-in-chief may hamper the tribunals’ purported goal of setting a historical record,Footnote 57 as it removes much of the material that would otherwise be contained in transcripts on the tribunals’ websites and renders it inaccessible to victims and other interested parties.
Again, some flexibility in the application of the rules has been observed – in Đorđević, for example, the motion to admit the testimony of witness Haxhiu pursuant to Rule 92 ter was granted in part, but the witness was ordered to appear for examination-in-chief as regards one contentious meeting discussed in his statement.Footnote 58 While in theory, Rule 92 ter procedure ought to protect both the expedience of the trial and the rights of the accused, it has not been immune from critique. It has been criticized for limiting the orality of proceedings in the sense that the first time the court hears a witness is in the context of cross-examination, as there is no initial examination, which may in turn render credibility assessments more difficult.Footnote 59 Indeed, the Trial Chamber in Gotovina noted that while, in principle, Rule 92 ter testimony could go towards proving the acts and conduct of the accused, viva voce testimony was still preferable.Footnote 60 In Milutinović, the Trial Chamber urged caution, given that Rule 92 ter statements had been tendered just days before the witness appeared in court to give evidence, and some statements had been altered quite significantly in advance of the witness's appearance, which had the potential of hampering the other side's right of cross-examination.Footnote 61 As neither party claimed to have suffered prejudice as a result, the trial chamber did not take any further action on the issue, but it did note that such last-minute amendments to witness statements were ‘generally unsatisfactory’.Footnote 62 This issue of last-minute amendments to Rule 92 ter statements appears to have persisted in the ICTY,Footnote 63 but none of the more recent judgements mention it as a problem.
The ICTR does not have a Rule 92 quater, but its Rule 92 bis(C) allows the admission of the statements of deceased witnesses provided that such testimony does not go towards the acts and conduct of the accused. Three statements in Karemera were admitted under Rule 92 bis(C),Footnote 64 although they were not ultimately referred to in the final judgement. Rule 92 quater saw the admission of two witness statements for the prosecution in the Taylor trial before the SCSL. Admissibility under Rule 92 quater before the ICTY was an altogether more common occurrence. It is notable that one of the most criticized rules in the ICTY's procedural framework, Rule 92 quinquies,Footnote 65 was not used in any of the cases observed.
3. Weight given to written witness statements
Having discussed the volume of materials admitted under the various evidential rules, it must be noted that most of the judgments analysed emphasized the principle that admissibility had no bearing on weight; in other words, the fact that a piece of evidence was admitted to the record did not mean that the evidence in question would be relied upon in the final judgment.Footnote 66 This section seeks to adduce the ultimate value given to different forms of witness testimony introduced to the evidential record under various rules. This was not a straightforward task, not least because many of the statements are redacted on the court's database and because the witnesses are referred to by one name at one point in proceedings, and by a pseudonym at other junctures. Moreover, it was difficult to measure the precise weight given to particular pieces of testimony, given that the evidential record is assessed ‘as a whole’,Footnote 67 and there is no obligation on the Trial Chamber to address each relevant piece of admitted evidence in its judgment.Footnote 68 For the purposes of the present analysis, if a piece of testimony was referenced in the trial judgment, and the judgment did not go on to later disregard that piece of testimony or express doubts as to the credibility of the witness,Footnote 69 it was taken that the Chamber had accepted the veracity of the testimony's contents. There is a further difficulty in this endeavour, insofar as it risks prioritizing form over substance. The assumption here is not that oral testimony is inherently more reliable than its written counterpart,Footnote 70 rather that viva voce testimony allows the Chamber to assess the credibility of the witness in greater detail than if their witness statement was introduced, even with limited cross-examination under Rule 92 ter.
In spite of these limitations, there are still a number of important questions that can be answered by the present analysis, such as whether chambers view oral testimony as more valuable than written testimony, whether testimony which has not been subjected to cross-examination can be corroborated by another piece of unexamined testimony as proof of a matter, and whether the freer approach to admissibility adopted by some chambers is an adequate match to the assessment of evidence on the basis of the ‘totality of the evidence’. It will be shown that there is a large degree of inconsistency in the answers to these questions, depending on the chamber. It shall also be submitted that a failure to pinpoint the precise weight given to a piece of evidence might lead to misunderstandings at later stages of proceedings, and this is partly illustrated by the Gotovina and Perišić appeal judgements.
3.1. The continued value of oral testimony
Almost all of the judgements from the ICTR studied reiterated a preference for oral witness testimony and recalled the principle that prior witness statements should not normally be admitted as proof of their contents.Footnote 71 This position is of little surprise as the preference for orality remains in the Tribunal's Rules of Procedure and Evidence.Footnote 72 The trial chamber in Nyiramasuhuko noted that this preference was ‘general, though not absolute’.Footnote 73 The preference for oral testimony was not alien to the other tribunals which have abolished the preference for orality in their rules. In Gotovina, the case which made the most extensive use of Rule 92 ter, the Chamber ‘expressed a strong preference that . . . important evidence central and critical to the case be elicited orally from a witness in court’.Footnote 74 The Gotovina Trial Chamber adopted a meticulous approach to Rule 92 ter statements; almost universally, when a statement admitted under Rule 92 ter was referred to,Footnote 75 the part of the transcript where that witness testified live in court was also referenced, often along with other corroboratory material. When parts of the 92 ter statements were not attested to in later live testimony, they were admitted to the record under Rule 89(C) and the inconsistency would be taken into account when deciding the relevance and probative value of the prior inconsistent statement.Footnote 76
By contrast, the Perišić Trial Chamber explicitly stated that Rule 92 ter testimony was considered in the same manner as viva voce testimony would have been.Footnote 77 It is difficult to trace whether this was the case, because the prosecution's motions on the admissibility of evidence under Rule 92 ter on the ICTY's court records database lists the witnesses in confidential annexes.Footnote 78 The only exception is witness Koster, a DutchBat officer whose account from the Karadžić and Mladić transcript of the transfer of civilians from the Potočari compound by Mladić and his army and of attacks on Dutchbat forces by VRS troops was referenced without further corroboration. The witness was cross-examined on 2 and 4 December 2008, but the transcript of this cross-examination is not referred to in the Trial Chamber judgement. In the light of the often inadequate amount of time given to parties to prepare for cross-examination, as discussed above,Footnote 79 and given that the cross-examination without examination-in-chief gives the Chamber less time to evaluate the credibility of the witness, one wonders whether the Gotovina principle that important evidence should be elicited live in court would be preferable.
When it came to the weight to be attached to Rule 92 bis evidence, the Perišić Trial Chamber stated that there was no reason to presume as a general rule that Rule 92 bis testimony should bear less weight than viva voce testimony.Footnote 80 The witness statements admitted under this rule were the only eyewitness testimony of the incident and were thereby crucial to the finding of fact that a modified air bomb exploded at Bunićki Potok Street on 1 July 1995, a finding also supported by contemporaneous documentary evidence and photographs.Footnote 81 It is not suggested that the finding in this regard was erroneous or that the Chamber gave undue weight to the Rule 92 bis statements, but the Chamber could have been more explicit in stating that evidence that was not subject to cross-examination would be weighed very carefully in the light of other evidence on the record and would need corroboration for proof of its contents.Footnote 82 The Perišić Trial Chamber also stated that corroboration was not a formal requirement for Rule 92 quater testimony,Footnote 83 illustrating that the Chamber attached less significance to the principle of orality than did the Gotovina Chamber. However, the Chamber did later appear to attach some weight to corroboration through oral testimony as regards Rule 92 quater witnesses. Witness Deronjić's account was taken as a credible and reliable account, given that it was corroborated by the testimony of witnesses Nikolić and Vasić, whom the ‘Chamber had the benefit of hearing the testimony of’ in court.Footnote 84
It could be argued that the Perišić Trial Chamber's difference in approach on the weight to be given to Rule 92 bis testimony compared to the weight to be given to Rule 92 quater testimony is warranted, given that the latter can go towards proving the acts and conduct of the accused, while Rule 92 bis evidence cannot. However, it is argued that Rule 92 bis is more suitable as corroboration of oral evidence, and this tended to be how Rule 92 bis testimony was used in the judgments studied. A notable exception was observed in Karemera, however, where the only evidence of rape in the Butare préfecture that had been introduced by the prosecution was a single 92 bis witness statement.Footnote 85 The charges of rape at the other crime bases were supported by a mix of oral testimony, written statements, and adjudicated facts from the Akayesu, Niyitegeka, and Musema trials which the Trial Chamber had taken judicial notice of.Footnote 86 The Chamber held that the single statement had been ‘corroborated by the pattern of evidence from the other préfectures’.Footnote 87 The Chamber was further swayed by the fact that the defence had ‘not sought to rebut her [the Rule 92 bis witness’s] evidence’.Footnote 88 This appears to be a very dangerous precedent indeed as regards both the standard and burden of proof. It is something of a stretch to suggest that just because a pattern of conduct was proven in regions X and Y, this goes towards proving the same conduct in region Z. Furthermore, the fact that the Chamber was swayed by the defence's failure to rebut a single piece of evidence may be taken as imposing a positive burden on accused persons to counter the evidence against them, which goes beyond the presumption-of-innocence principle that the burden of proof rests with the prosecution.
The Gotovina judgement identified a number of factors for assessing the weight of Rule 92 quater testimony; these included the circumstances in which the statements were made or recorded, whether they were consistent with other statements, and whether they had been subject to cross-examination.Footnote 89 Unfortunately, perhaps, these factors were not explicitly worked through by the Đorđević Trial Chamber when it assessed witness Morina's 92 quater testimony, taken from the Milošević transcript.Footnote 90 This led the defence to claim in its appeal brief that the only evidence on the burning of a mosque at Landovica/Landovicë was Morina's 92 quater statement.Footnote 91 This was not correct, as the testimony was in fact consistent with the inspection carried out by witness Riedlmayer,Footnote 92 but perhaps this could have been made even more explicit by the Trial Chamber, and emphasis could have been drawn to the witness's oral testimony, in order to avoid such a critique of its evidentiary findings.
3.2. The ‘totality of the evidence’ approach
One of the arguments in favour of a broad admissibility regime is that it allows judges to make an assessment based on the ‘big picture’ which is sketched through thousands of individual pieces of evidence. The ‘totality of the evidence’ approach allows judges to come to a judgment based on a holistic assessment of the evidential record as a whole. The other side of the coin, however, is that it is often difficult to assess exactly how much weight is given to a piece of evidence in the absence of an explicit pronouncement on the credibility or reliability of that witness or evidence. In Đorđević, for example, following a pronouncement that, on occasion, the chamber accepted evidence that contained inconsistencies or contradictions while rejecting evidence that was apparently consistent with other pieces of evidence, the chamber claimed to have ‘acted in light of the other evidence on the issue’.Footnote 93 Similar approaches to the evidential record ‘as a whole’ were identified in Nyiramashuko Footnote 94 and Perišić.Footnote 95
The right to a reasoned verdict does not stretch to a right to know whether and on what basis a trial chamber found one piece of evidence to be more reliable or authentic than another piece,Footnote 96 and the Appeals Chamber of the ICTY has explicitly stated that the standard of proof should not be based on a piecemeal approach, but rather on the totality of the evidential record.Footnote 97 However, difficulties can arise at the appeals stage of proceedings when the Appeals Chamber expresses the view that a certain piece or pieces of evidence have not been adequately addressed. In the Perišić appeals judgement, for example, the Chamber referred to the ‘paucity’ of evidence supporting the finding on the accused's ability to issue commands,Footnote 98 and its belief that the Trial Chamber failed to adequately take the testimony of witnesses Rašeta and Orlić into account.Footnote 99 The Trial Chamber judgement had noted the testimony of both witnesses,Footnote 100 but had come to a contrary conclusion based on other evidence on file.Footnote 101 It will be recalled that:
A Trial Chamber need not refer to the testimony of every witness or every piece of evidence on the trial record, ‘as long as there is no indication that the Trial Chamber completely disregarded any particular piece of evidence.’ Such disregard is shown ‘when evidence which is clearly relevant . . . is not addressed by the Trial Chamber's reasoning.’Footnote 102
By noting the contrary evidence and basing its conclusions on the totality of the evidence, the Perišić Trial Chamber had clearly addressed the relevant evidence and had not shown complete disregard for it in a manner that would constitute an error of law. The testimony of Orlić testified that orders were not received from Perišić, but the majority of the Chamber concluded that even if the accused was not in a position to issue orders, this did not discount the existence of other means to control subordinates.Footnote 103 Yet, the Appeals Chamber was of the view that the Trial Chamber's ‘failure to explicitly discuss and analyse [these witnesses’] evidence’ constituted an error of law, and proceeded to conduct a de novo review of the evidence on that basis.Footnote 104 In a very similar manner, the Gotovina Appeals Chamber felt that the Trial Chamber's finding on the occurrence of indiscriminate attacks in Knin, in particular, was not ‘adequately supported’ by the evidence, in light of its finding that the 200-metre standard was incorrect.Footnote 105 But, as Judge Agius pointed out in his dissenting opinion, the Appeals Chamber's approach represented an ‘overly compartmentalised and narrow view’Footnote 106 that failed to acknowledge many key pieces of evidence,Footnote 107 leading to its apparent conclusion that when one finding cannot be upheld on appeal, the rest of the evidential record must collapse beneath it like a house of cards. The approach of the Appeals Chambers in both cases perhaps raises broader questions about the Appeals Chamber's function and standard of review, but for the purposes of the present article, both decisions show dubious interpretations of the ‘totality of the evidence’ approach. The Perišić appeals judgement, in particular, seems to suggest that it is not sufficient to acknowledge contrary evidence on the record and base a finding on other evidence; the Appeals Chamber required an explicit discussion and analysis of contrary evidence, with full elucidation as to why a witness's testimony on one point was not relied upon.Footnote 108 The approach of the Gotovina and Perišić Appeals Chambers would appear to constitute a move away from the well-established position that Trial Chambers are best placed to assess the evidential record as a whole, and by consequence, their findings should not be lightly overturned.Footnote 109
Later Appeals Chambers’ interpretations aside, the ‘totality of the evidence’ approach does raise a number of further questions. The first is whether the admissibility rules really align with this approach. Many of the admissibility rules on written witness testimony in lieu of viva voce evidence were introduced with a view to aiding expedience, but have ironically on occasion added an extra layer of complexity in calling for submissions on whether a statement goes to the acts and conduct of the accused, whether cross-examination would be in the interests of justice, and so forth, to such an extent that at times, it would have been more expedient to call the witness to testify in person. In the light of this paradox, we might question the wisdom of a complex admissibility process that has no bearing whatsoever on weight, and that allows pieces of evidence that have had to jump through numerous admissibility hurdles to be completely disregarded in the final judgment. The ‘totality of the evidence’ approach might be well suited to a legal culture that embraces the free admission of almost all evidence, but it appears to be at odds with a statutory framework that demands that certain criteria be met before a piece of evidence can even come before the judges.
The observer might also question the extent to which a ‘totality of the evidence’ approach can be implemented in a context where the totality of the evidence comprises tens of thousands of pages of transcript, thousands of pieces of evidence, and a trial period that often stretches into the hundreds of days. For example, as well as the witnesses outlined in Table 1, the Đorđević Trial Chamber also received 2,518 pieces of documentary evidence ‘from the bar table’ and the trial record encompasses 14,534 pages of transcript. The Lubanga judgment illustrates the practical difficulties in this regard: there was so much evidence admitted before the chamber that the parties were asked to make explicit which witnesses and pieces of testimony they wanted the Trial Chamber to focus on in making its judgement, and they were warned that if they failed to do so, they ran the risk of having relevant pieces of evidence overlooked.Footnote 110 Similarly, the Gotovina judgement placed a great degree of reliance on evidence emphasized by the parties in their closing submissions,Footnote 111 illustrating that, in spite of more liberal rules on the admissibility of evidence, proceedings continue to be party-driven, as opposed to judge-driven.
This section illustrated that the principle of orality is given varying degrees of importance, depending on the tribunal and the composition of the chamber. Some chambers placed huge emphasis on the preference that testimony should be elicited orally, while others had no objection, in principle, to testimony which had not been subject to cross-examination being given equal weight to viva voce testimony. A greater degree of consistency on this question, and on the requirements for corroboration of untested witness testimony, would assist in clarifying the evidentiary principles of international criminal procedure as the ad hoc tribunals’ mandates come to an end and observers seek to assess their legacy. When it comes to the impact of certain pieces of evidence on the final judgment, we see that it becomes difficult to adduce the precise weight given to individual pieces of evidence, given that the evidential record is said to have been assessed as a whole. This ‘totality of the evidence’ approach can give rise to issues at the appeals stage of proceedings, when the Appeals Chamber feels that a certain piece of evidence has not been duly considered and enters into a de novo assessment of the evidential record as a consequence. Furthermore, the compatibility of the ‘totality of the evidence’ approach with the more liberal admissibility rules may be called into question, given the sheer volume of evidence before the Trial Chambers and the fact that the admissibility regime falls short of a free proof-approach.
4. Conclusion
This article sought to fill a gap in the literature on written witness testimony in lieu of oral testimony at the ICTY, the ICTR, and the SCSL by assessing the broad picture of admissibility decisions on written witness statements, and the ultimate weight given to such statements. The article did not claim to give an overall assessment of the trajectory of the move towards increased use of written testimony over time, but rather a snapshot of the use and impact of new admissibility rules at one point in time. Further research would be needed to assess the overall impact of these new rules.Footnote 112
The statistics produced do show that, perhaps surprisingly, the more flexible rules on written witness testimony that have been introduced in recent years have been used relatively infrequently, particularly at the ICTR and the SCSL. This suggests that some of the criticisms of the newer rules may have been premature. Furthermore, it was shown, evidence introduced under some of the newer rules which do not technically require cross-examination were still subject to cross-examination, illustrating a cautious approach to written witness testimony.
Further caution was observed with regard to the weight given to oral testimony vis-à-vis written testimony, but the chambers’ approaches to the principle of orality was at times inconsistent. Some insisted that there was no reason to assume that written testimony should be given less weight than oral testimony, although in most instances, written testimony was corroboratory to other evidence on the record. One worrying incident from Karemera was reported, where a single Rule 92 bis statement was used as proof of a matter, seemingly influenced by the fact that it was consistent with evidence from other prefectures and the fact that the defence had not rebutted the testimony.
The evidential record is weighed by Trial Chambers as a whole, and Appeals Chambers should be slow to overturn findings of fact, given that the Trial Chamber is in the best position to come to conclusions based on an overall impression from the evidence placed before it. However, recent Appeals Chamber decisions suggest that Trial Chambers may need to be more explicit as to the precise value given to individual pieces of evidence and the reasons for doing so. Given that judgments often run to several thousand pages already, this will be a time-consuming and difficult exercise. The broad admissibility rules studied in this article may not aid expedience of trial proceedings, given the debates that tend to ensue on matters such as whether the evidence goes to the acts and conduct of the accused, whether the witness is truly unavailable, and so on. Furthermore, given the huge volume of evidence on record, chambers showed a tendency to request that parties point them to the most relevant pieces of evidence in their closing briefs or submissions. This shows the impossibility of an assessment that truly takes the totality of the evidence into account, without some direction from the parties. These and other issues discussed in the present article show that the international criminal tribunals studied have yet to strike the perfect equilibrium between trial efficiency and the most complete presentation of evidence, and may suggest a need to re-evaluate the operation of rules relating to written witness testimony in international criminal trials.
