DECLARATION OF JUDGE SHAHABUDDEEN

1.         I support the judgement but propose to state my understanding of two points, one concerning the reliability of evidence, the other concerning the test for deciding on the effect of additional evidence.

A.   Reliability of Evidence

1.         The Problem

2.         The point here relates to the reproduction, in paragraph 46 of the judgement, of the holding of the ICTY Appeals Chamber in Kordić[1] that -

the reliability of a statement is relevant to its admissibility, and not just to its weight. A piece of evidence may be so lacking in terms of the indicia of reliability that it is not 'probative' and is therefore inadmissible.

3.         This proposition was adopted in paragraph 286 of the appeal judgement in Akayesu.[2] I believe that that judgement was correct, but will note that that case, like Kordić, was concerned with the question of the admissibility of an out-of-court statement, and not with evidence generally.

4.         My hesitation is that the Kordić proposition may be given a wider application than may have been intended: it could be understood as meaning that evidence of all kinds must be shown to be reliable before it is admitted. I do not think it was meant in that universal way. In general, I agree with the view expressed by J.R.W.D Jones that -

… whilst evidence may be excluded because it is unreliable, it is not required that evidence be shown to be reliable before it is admitted. The evidence need only be shown to be relevant, in order for it to be admissible.[3]

5.         Jones was not dealing with possible grounds of inadmissibility other than unreliability, and he accepted that evidence which was in fact shown to be unreliable at the admissibility stage might be then excluded as inadmissible. His focus was directed to the question whether there was a requirement that evidence must be shown to be reliable as a pre-condition of admissibility. With exceptions, I do not think that there is such a requirement.

2.         In general, at the admissibility stage, the credibility of evidence (including reliability) has to be assumed; reliability goes to weight and is assessed later

6.         Under the system of the Tribunal, whatever may be the situation in particular national systems, the principle is this: reliability is a component of credibility, credibility goes to weight, and weight is assessed at the end of the proceedings.

7.         Rule 89(C) of the Rules of Procedure and Evidence (“Rules”) provides that a “Chamber may admit any relevant evidence which it deems to have probative value”. As has been repeatedly and correctly pointed out, relevance implicitly requires some component of probative value: evidence is relevant if it is probative, that is to say, if it has a tendency to make the existence of a fact that is of consequence to the determination of the case[4] more probable or less probable.[5]  Evidence which does not have this tendency to prove what has to be proved is not probative; it is therefore not relevant and is not admissible. This applies to all evidence, whether hearsay or direct.[6] 

8.         But a distinction has to be made between a judgement that evidence is probative and the basis on which the judgement is made. A judgement that evidence is probative is made on the basis that it is credible, including a finding that it is reliable. At the admissibility stage it is assumed, rather than found, that the evidence is credible. It is on the basis of that assumption that it is determined, at that stage, whether the evidence can advance the proof of the fact which has to be proved and is therefore probative. Evidence which cannot do that (even if it is assumed to be credible) is not probative; it is therefore not relevant and is not admissible. If, on the basis of an assumption that it is credible, it is determined that the evidence can establish the fact to be proved and is therefore admitted, the next question (to be answered at a later stage of the proceedings) is to what extent it does indeed establish the fact to be proved. It is this next question which raises the point whether the evidence is credible, including the issue whether, even if the witness is speaking truthfully, he is for one reason or another mistaken. And it is here that the presence or absence of reliability matters.

9.         In general, then, a decision to admit assumes that the evidence is credible: it assumes matters, such as reliability, which go to credibility. The assumption that the evidence is credible is then verified after the making of a decision to admit it; this is part of the exercise concerned with the assessment of the weight to be assigned to the admitted evidence.[7] If the evidence is then judged not credible, it is simply given no weight and eliminated from the proof, even though it was earlier admitted.

3.         The foregoing general rule may be displaced in some cases but not in all

10.       What appears to be a general rule that credibility (including reliability) is assumed at the stage of admissibility is, however, inapplicable where a different rule has been laid down by or under the Rules; further, the assumption stands rebutted if it in fact appears at that stage that the evidence is indeed unreliable. Nothing needs to be said on the latter branch; something may be said on the former.

11.       In respect of hearsay, the existence of a different rule has come into being, and for good reason. Granted that hearsay evidence is considered to be admissible under Rule 89(C)[8], its nature and provenance call for special care when deciding to admit it. There may be cause for not admitting it where it has passed through a multitude of intermediaries before reaching the court; these are matters that can often be sufficiently explored at the stage of admissibility of the particular piece of hearsay evidence to justify non-reception on grounds of unreliability.[9] Subject to the qualification mentioned below, the developed jurisprudence, as it is evidenced by Kordić and other cases, accepts that reliability must be established before hearsay evidence is admitted. A rule to that effect could be founded on Rule 89(B), reading:

In cases not otherwise provided for in this Section, a Chamber shall apply rules of evidence which will best favour a fair determination of the matter before it and are consonant with the spirit of the Statute and the general principles of law.

12.       It is recognised that in Delalić[10] an ICTY Trial Chamber rejected a defence submission “that a determination of reliability should be seen as a separate, first step in assessing a piece of evidence offered for admission”.[11] As a general matter, the rejection was right. I consider, however, that the rejection is today to be regarded as qualified in the particular case of hearsay evidence: the accumulated jurisprudence demonstrates a requirement for proof of reliability before such evidence is admitted.[12]

13.       But this is said with the following qualification: it may not be practicable to make a full exploration of all the circumstances relating to the reliability of an out-of-court statement at the admissibility stage. In consequence, a Chamber may not be in a position to decide that the reliability of such a statement has or has not been definitively established; it may, however, be able to find that there are indicia of reliability.  In such a case, it may admit the evidence, deferring a final decision for a later stage of the proceedings.

14.       Thus, in Delalić, the Trial Chamber admitted certain out-of-court documents on the basis that there were “sufficient indicia of reliability”.[13] In doing so, it “emphasised that this decision does not in any way constitute a binding determination as to the authenticity or trustworthiness of the documents sought to be admitted”. It added that these “are matters to be assessed by the Trial Chamber at a later stage in the course of determining the weight to be attached to these exhibits”. In effect, since it treated “authenticity” as covered by “reliability”, it treated “reliability” as a matter of “weight” which could be “assessed at a later stage”. On this basis, it considered that definitive proof of reliability as a condition of admissibility of out-of-court statements was not necessary; provisional proof was all that was required at that threshold stage.

15.       As to direct evidence, it may be even less feasible to explore questions of reliability at the admissibility stage; reliability may depend on the totality of the evidence and may only be capable of definitive determination at a later stage of the proceedings. A party may not always be in a position to show that its direct evidence is reliable at the admissibility stage; if, on the ground that reliability is not shown at that point, the evidence is then shut out, the court deprives itself of the opportunity of later finding that the evidence was in fact reliable.

16.       In such cases, the general principle should therefore apply: reliability should be left for assessment as part of weight. In the present matter, it is observed that it was in the final judgement that the Trial Chamber considered whether the direct evidence of certain witnesses was or was not “reliable”[14], the evidence having been admitted earlier. In my respectful view, that was the correct approach.

4.         Where the Rules intend reliability to be a condition of admissibility, they say so

17.       It is recognised that Rule 95 of the Rules bars admissibility in the case of unreliability, but only in particular circumstances. The Rule reads:

No evidence shall be admissible if obtained by methods which cast substantial doubt on its reliability or if its admission is antithetical to, and would seriously damage, the integrity of the proceedings. [Emphasis added].

18.       The second branch of the Rule excludes evidence as not admissible if, even where the evidence is perfectly reliable[15], its admission is antithetical to or would seriously damage the integrity of the proceedings. It need not be considered further.

19.       The first branch of the Rule excludes evidence as not admissible if it was obtained by methods which cast substantial doubt on its reliability.[16]  On a contrario reasoning, the Rule implies that, in cases not within the scope of the Rule, the principle is that proof of reliability is not a condition precedent to admissibility; reliability is to be later determined as a matter going to weight.

20.       With the exception referred to, the Rule establishes no linkage between admissibility and unreliability.

5.         General legal thinking

21.       Is such a linkage to be found in general legal thinking? Rule 89(A) of the Rules provides that the Chambers “shall not be bound by national rules of evidence”. It does not prohibit a Chamber from consulting national rules on the subject, and that has indeed been done in other cases.

22.       Accordingly, it may be noted that, in some systems, reliability is linked to admissibility, but only in limited circumstances. Thus, in one jurisdiction, legislation provides that if a “confession was or may have been obtained ... in consequence of anything said or done which was likely, in the circumstances existing at the time, to render ₣itğ unreliable ..., the court shall not allow the confession to be given in evidence …”.[17]

23.       So, there, the courts have been required not to admit evidence of a specific kind, namely, confessions, on the ground of unreliability arising from particular circumstances. In part, the courts of the jurisdiction concerned might already have had such a duty under the law relating to voluntariness.[18] The important thing, however, is that, in the case of other types of evidence, the assumption of the legislation is that, in the absence of exceptions, reliability has to be left to be considered as part of the weight of the evidence and does not have to be established before the evidence is admitted.[19]

24.       It is also useful to bear in mind the statement of the United States Supreme Court in United States v. Matlock[20] that “the rules of evidence normally applicable in criminal trials do not operate with full force at hearings before the judge to determine the admissibility of evidence”.

6.         Conclusion

25.       The general principle appears to be that reliability goes to weight and not to admissibility and is to be assessed only when weight is evaluated. The general principle is displaced only by exceptions made by or under the Rules. Where exceptions do not apply, the general principle does. Accordingly, in the normal situation there is no requirement for proof of reliability as a condition of admissibility; reliability is to be left for later evaluation as part of weight.

B.   The Test For Deciding On The Effect Of Additional Evidence

26.            Paragraph 185 of the judgement adopts the following statement from the Kupreskic judgement of the ICTY Appeals Chamber:

The test to be applied by the Appeals Chamber in deciding whether or not to uphold a conviction where additional evidence has been admitted before the Chamber is: has the appellant established that no reasonable tribunal of fact could have reached a conclusion of guilt based upon the evidence before the Trial Chamber together with the additional evidence admitted during the appellate proceedings? (“reasonable conclusion criterion”).[21]

27.       Supporting references were not given for that statement. None had to be, but the absence excuses inquiry.

28.       The basis on which additional evidence is admitted is not the same as that on which evidence of a new fact is admitted. Otherwise, they have the same characteristics: they both represent evidence which was not before the Trial Chamber and they both involve a determination by the Appeals Chamber of their impact on the judgement of the Trial Chamber. It would appear that the criterion for this determination should be the same in both cases.

29.       As to what is the criterion, Article 25 of the Statute of the Tribunal speaks of a new fact “which could have been a decisive factor in reaching the decision” (“decisive factor criterion”). Evidently, this is the criterion to be applied by the Appeals Chamber in determining the impact of a new fact on the judgement. It would appear that the same criterion should apply to the determination of the impact of additional evidence on the judgement.

30.       It may be said that the decisive factor criterion yields the same result as the reasonable conclusion criterion. But perhaps not quite. These are the reasons for hesitation.

31.       In normal appellate practice, the reasonable conclusion criterion applies where all the evidence has in fact been assessed by the trial court and where the conclusion reached by the trial court on that evidence is known. The test then is whether the known conclusion reached on the assessed evidence was one which no reasonable tribunal would have reached on that evidence. Where that test is met, what is being said is that something went wrong in the handling of the case by the court below.

32.       In the case of additional evidence, the evidence in question was never before the trial court and the latter never came to a conclusion on it: it is not being said that anything went wrong in the handling of the case by the court below. All that can be said by an appellate court is that the additional evidence could or could not have been a decisive factor in reaching the decision which was reached by the court below. The stress is to be laid on the word “could”. What this looks to is the capacity of the additional evidence to function as a decisive factor. The lower court might or might not in fact have considered it to be a decisive factor: one never knows, for the lower court (even if it could be reconstituted) is not being interrogated. But that is not the question. The question is whether the appellate court judges that the evidence had the capacity to function as a decisive factor. 

33.       There is ground for apprehension as to whether the two tests yield different results in marginal but real situations. On the decisive factor criterion, it may be possible for the Appeals Chamber to reverse the conviction in circumstances in which it may have to maintain it on the criterion of reasonable conclusion. The decisive factor test is thus more favourable to the accused. And so it should be, for what is being dealt with is additional evidence which was not before the Trial Chamber and on which its thinking is therefore not known. It is right to make extra allowance for the possibilities involved in that circumstance. In my view, the decisive factor criterion is to be preferred.

Done in English and French, the English text being authoritative.

_____________________

Mohamed Shahabuddeen

Dated this 16^th day of November 2001

At The Hague

The Netherlands

[Seal of the Tribunal]