382. The Appeals Chamber recalls that where a finding of guilt is made on the basis of identification evidence given by a witness under difficult circumstances, the Trial Chamber needs to “carefully articulate the factors relied upon in support of the identification of the accused and adequately address any significant factors impacting negatively on the reliability of the identification evidence”.[1] The Appeals Chamber stresses that this is required only when a witness’s identification was made under difficult circumstances, such as in the dark or as a result of a fleeting glance.[2] […]

392. […] The Appeals Chamber recalls that identification hearsay evidence may, depending on the circumstances of the case, require other credible or reliable evidence in order to support a finding of fact beyond reasonable doubt.[3] […]

152. The Appeals Chamber recalls that, in the case of complex issues, such as the assessment of identification evidence, the reasoned opinion requirement to be met by a Trial Chamber is higher.[1] The Appeals Chamber has held that, “where a finding of guilt is made on the basis of identification evidence given by a witness under difficult circumstances, the Trial Chamber must rigorously implement its duty to provide a ‘reasoned opinion’.”[2] In these instances, the Trial Chamber is required to “carefully articulate the factors relied upon in support of the identification of the accused and adequately address any significant factors impacting negatively on the reliability of the identification evidence.”[3] The Appeals Chamber observes that, in such cases:

[w]here the record, including the reasons for judgment, discloses a lack of appreciation of relevant evidence and more particularly the complete disregard of such evidence, then it falls upon the reviewing tribunal to intercede.[4]

155. The Appeals Chamber recalls that Trial Chambers must consider the difficulties associated with identification evidence in a particular case and carefully evaluate it before accepting it as the sole basis for sustaining a conviction. In particular, Trial Chambers must exercise extreme caution before convicting a person based on the identification evidence of a single witness made under difficult circumstances in light of the frailties of human perception and the risk that a miscarriage of justice might result from reliance upon even the most confident witnesses who purport to identify an accused without an adequate opportunity to verify their observations.[5]

156. Factors relevant to the Appeals Chamber’s determination of whether a Trial Chamber’s decision to rely upon identification evidence was unreasonable or renders a conviction unsafe include:

[…] identifications of defendants by witnesses who had only a fleeting glance or an obstructed view of the defendant; identifications occurring in the dark and as a result of a traumatic event experienced by the witness; inconsistent or inaccurate testimony about the defendant’s physical characteristics at the time of the event; misidentification or denial of the ability to identify followed by later identification of the defendant by a witness; the existence of irreconcilable witness testimonies; and a witness’ delayed assertion of memory regarding the defendant coupled with the “clear possibility” from the circumstances that the witness had been influenced by suggestions from others.[6]

Furthermore, the Appeals Chamber recalls that identification evidence may be affected by the length of time between the crime and the confrontation.[7]

113.   On whether Witness F could easily recognize Musema, the Appeals Chamber finds that Musema’s arguments are not sufficient to raise doubt as to the reliability of the contested identification testimony.  The Appeals Chamber notes that during a meeting convened by the bourgmestre of Gisovu commune, which was one of the three occasions where F had seen Musema prior to the events, F was able to observe Musema for a period of 30 minutes.[1]  Musema gives the impression that an identified suspect needs to be personally well known to the witness.[2]  This is not the case. Prior knowledge of an identified suspect is a factor that a Trial Chamber may take into account when assessing the reliability of a witness’ testimony,[3]  but that is not a sine qua non; identification may be based on other factors. In any event, the Appeals Chamber is of the opinion that it was within the discretion of the Trial Chamber to accept, in support of the evidence of identification before it, the fact that Witness F had met Musema on several occasions.

27. The Appeals Chamber agrees with both parties that no probative weight should be attached to in-court identifications.[1] As considered by the Kunarac Trial Chamber, in-court identifications are inherently unreliable “[b]ecause all of the circumstances of a trial necessarily lead such a witness to identify the person on trial”.[2] This has been affirmed in both the Kunarac and Kamuhanda Appeal Judgements.[3]

28. […] Accordingly, the Trial Chamber attached some weight to the in-court identifications and to the extent that it did, it was in error.[4] […]

30. In considering this allegation, the Appeals Chamber recalls its finding in Kupreškić et al. that a reasonable Trial Chamber must take into account the difficulties associated with identification evidence in a particular case and must carefully evaluate any such evidence, before accepting it as the sole basis for sustaining a conviction. Domestic criminal law systems from around the world recognise the need to exercise extreme caution before proceeding to convict an accused person based upon the identification evidence of a witness made under difficult circumstances. The principles developed in these jurisdictions acknowledge the frailties of human perceptions and the very serious risk that a miscarriage of justice might result from reliance upon even the most confident witnesses who purport to identify an accused without an adequate opportunity to verify their observations.

After having examined a number of domestic criminal law systems in relation to the question of identification evidence, the Appeals Chamber stated in Kupreškić et al.:

Courts in domestic jurisdictions have identified the following factors as relevant to an appellate court’s determination of whether a fact finder’s decision to rely upon identification evidence was unreasonable or renders a conviction unsafe: identifications of defendants by witnesses who had only a fleeting glance or an obstructed view of the defendant; identifications occurring in the dark and as a result of a traumatic event experienced by the witness; inconsistent or inaccurate testimony about the defendant’s physical characteristics at the time of the event; misidentification or denial of the ability to identify followed by later identification of the defendant by a witness; the existence of irreconcilable witness testimonies; and a witness’ delayed assertion of memory regarding the defendant coupled with the “clear possibility” from the circumstances that the witness had been influenced by suggestions from others.

In addition, the Appeals Chamber observes that identification evidence may be affected by the length of time between the crime and the confrontation.[7]

96. The Appeals Chamber recalls that in assessing witness testimony, “it falls to the Trial Chamber to take the approach it considers most appropriate for the assessment of evidence.”[1] A trial chamber “is […]] not obliged in its judgement to recount and justify its findings in relation to every submission made at trial.”[2] In addition, “neither the Rules nor the jurisprudence of the Tribunal oblige[]] [a] Trial Chamber to require a particular type of identification evidence.”[3] However, identifications made in difficult circumstances, such as darkness, obstructed view, or traumatic events,[4] require careful and cautious analysis by a trial chamber.[5] In addition, in-court identification evidence should be assigned “little or no credence” given the signals that can identify an accused aside from prior acquaintance.[6]

120. The Appeals Chamber recalls that in-court identification is generally permissible.[1] Contrary to the Appellants’ contention,[2] the fact that a witness fails to identify the accused during a pre-trial identification exercise does not bar a trial chamber from allowing in-court identification.[3] However, in-court identification should be given “‘little or no credence’ given the signals that can identify an accused aside from prior acquaintance”.[4] A trial chamber must therefore exercise caution in assessing such evidence.[5] The Appeals Chamber considers that, while little or no weight should be given to a witness’s in-court identification evidence,[6] a witness’s failure to identify an accused in court can be a reason for declining to rely on the identification evidence of that witness.[7] The distinction between an “identification” or “recognition” witness is thus irrelevant in the context of in-court identification. In the present case, the Trial Chamber allowed in-court identifications by several witnesses. The Appeals Chamber considers that, as a matter of law, Milan Lukić and Sredoje Lukić have not shown that the Trial Chamber erred by allowing in-court identification per se. The Appeals Chamber will evaluate the Appellants’ claims regarding each witness in the appropriate section.

136. The Appeals Chamber further recalls that “where a finding of guilt is made on the basis of identification evidence given by a witness under difficult circumstances, the Trial Chamber must rigorously implement its duty to provide a ‘reasoned opinion’”.[8] In these instances, the Trial Chamber must “carefully articulate the factors relied upon in support of the identification of the accused and adequately address any significant factors impacting negatively on the reliability of the identification evidence”.[9]

137. Factors relevant to the Appeals Chamber’s determination of whether a trial chamber’s decision to rely upon identification evidence was unreasonable or rendered the conviction unsafe, include:

[I]dentifications of defendants by witnesses who had only a fleeting glance or an obstructed view of the defendant; identifications occurring in the dark and as a result of a traumatic event experienced by the witness; inconsistent or inaccurate testimony about the defendant’s physical characteristics at the time of the event; misidentification or denial of the ability to identify followed by later identification of the defendant by a witness; the existence of irreconcilable witness testimonies; and a witness’ delayed assertion of memory regarding the defendant coupled with the “clear possibility” from the circumstances that the witness had been influenced by suggestions from others.[10]

138. Furthermore, the Appeals Chamber considers that this list is not intended to be exhaustive.[11] A trial chamber is not bound to consider all of these factors, but only those appropriate in the circumstances of the case.[12]

327.    […] [T]he Appeals Chamber finds that the witnesses’ personal knowledge of Ruzindana is not a prerequisite for identification.[1]

328.    As noted by the Trial Chamber, “prior knowledge of those identified is another factor that the Trial Chamber may take into account in considering the reliability of witness’ testimonies.”[2] The fact that some of the witnesses did not personally know the accused prior to the events is not at all a sufficient reason to invalidate the testimony of a witness who identified the Accused.[3]