Deriving Narrative Authority from the Terminated Milošević Trial
Indiana University*
Much of international criminal law’s attraction rests on the “authoritative narrative theory”—the claim that judgment creates narratives that serve as the foundation for post-conflict reconciliation. So what happens when there is no judgment? What can be salvaged from the terminated Milošević trial? One candidate for substitute judgment was the Decision on the Motion for Acquittal issued at the trial’s midpoint This chapter examines how the parties and outsiders deployed the Rule 98bis Decision, in the absence of a final verdict, to tell a story about Milošević’s guilt or innocence and craft a final judgment in the eyes of the world, if not in law. The doctrinal constraints these efforts faced suggest real limits on law’s ability to craft transformative narratives through any process short of final judgment—an indirect challenge to one of the dominant views about what international criminal law is for.
One of the most common, if most contested, claims for the efficacy of international tribunals is that they produce authoritative narratives, which in turn contribute to reconciliation.1 The key element in this claim is the final judgment. While many parts of a trial contribute to the production of narratives, only the decisions of the court carry the imprimatur of official, consequential authority. This is especially true given the largely adversarial nature of many international tribunals: Evidence submitted by one party and critiqued by the other is, by its nature, the product of partisan interpretation, and it is the role of the judge to determine that evidence’s ultimate value. It is judgment—grounded in the procedural integrity of the trial but sitting at the apex of the process—that produces the dispositive, displacing authority claimed for ICL’s interpretations of violent conflict.
A contested claim, to be sure, but assuming one finds it plausible, what happens when there is no final judgment? Assuming one believes in the transformative power of ICL and its judgments, what can be salvaged from a terminated trial? This is precisely the question that confronted the ICTY—indeed all actors with an interest in the general project of ICL or its particular effects in the former Yugoslavia—when Milošević died.
It was clear, when Milošević was transferred to the Tribunal, that the ICTY had not yet produced any demonstrable reconciliation in the former Yugoslavia. Various communities in the successor states held radically different views on the wars’ origins, on who was a victim and who a perpetrator, and what constituted justice.2 Milošević was unquestionably a critical player in the collapse of Yugoslavia and the violent wars of its dissolution, but his exact role and that of the state apparatus he headed were contested questions, which his trial promised to finally answer. It promised to be the single greatest test of the Tribunal’s—and ICL’s—ability to forge an efficacious and transformative narrative. But then, of course, Milošević died before judgment could be rendered.
The sudden termination of trials by death is not a hypothetical problem, nor is Milošević the only instance. A surprisingly large share of international defendants die before their trials can be completed, and given the nature of the conflicts that tribunals adjudicate and the continuing problems of enforcing indictments, we can expect that the problem will occur with regularity. If judgment matters, this is an endemic weakness for ICL. This chapter considers the meaning of the terminated Milošević trial in relation to the idea that courts produce consequential narratives. In particular, the chapter examines efforts to deploy and interpret the one document from Milošević that, more than any other, was thought of as a substitute for judgment: the Trial Chamber’s Rule 98bis Decision on the Motion to Acquit.3
Halfway through the trial, when the Prosecution rested, the Chamber had to decide if there was a case to answer. In their Decision, issued in June 2004,4 the judges declared the trial should go forward, though what that meant was necessarily ambiguous: The Prosecution had presented enough evidence that a court could find Milošević guilty, but the judges did not say that this court would. The Decision, therefore, had an irreducibly interim quality, but because Milošević died before a final verdict was issued, the Decision has become “the only pre-Judgement determinative ruling of the Chamber on the case.”5
This chapter examines how the Decision has been deployed to tell a story about Milošević’s guilt or innocence and craft a final judgment in the eyes of the world, if not in law. The divergent interests of the parties at trial created asymmetrical expectations about the Decision, revealed both in those actors’ responses when the Decision was issued and in the ways they tried to deploy the Decision as a quasi-judgment after Milošević’s death. These uses were limited by the doctrinal structure that produced the Decision—even actors convinced of Milošević’s guilt were constrained from using the Decision to advance that view—in precise ways. These constraints show the tenuous nature of the ICL project, one of the principal justifications for which appears radically contingent on reaching final judgment.
Going further, a close reading of the Decision indicates some parameters for an even more consequential inquiry into how real or limited the potential of even final judgments is in achieving post-conflict transformation. But although this chapter helps lay the groundwork for broader investigation of the authoritative narrative theory, its immediate purpose is forensic—a dissection of the Decision, its internal logic and institutional context, and the uses to which it was put once that context was disrupted by Milošević’s death.
Many justifications are advanced for the project of ICL. One principal rationale is ICL’s capacity to cut through endemic partisan contestation and render incontestable judgment about divisive conflicts; this authority undermines alternative claims and histories, creating space for shared understanding of truth, which may serve as a prerequisite for reconciliation.6 This rationale, appearing in many forms, is contested, but also deeply entrenched in practice and scholarship; we may call it the authoritative narrative theory of ICL.
The authoritative narrative theory begins with an uncontroversial observation: In societies that have suffered violent conflict, shared understandings about history and identity will often have been comprehensively destroyed. Theorists and practitioners commonly suggest that some form of post-conflict justice is essential to the restoration and maintenance of peace.7
The strong version of the authoritative narrative theory—which accepts that trials establish truths with consequential value—extends this line of reasoning, arguing that formal legal justice is particularly efficacious in crafting shared understandings.8 The formal authority attaching to a judgment is different from mere opinion: judgment has both the power of the state (or international community) behind it and the legitimacy of a neutral, professional judiciary; judgment follows a trial that accords different views the opportunity to be weighed in a procedurally balanced forum; and the powers of subpoena and contempt help assure that full information is available. Thus, when judgment is rendered, it exercises a special effect, displacing partisan views. As more and more discursive space is occupied by views that more and more members of society accept, the shared understandings essential to peaceful coexistence are restored.9 This is a claim about a causal chain, and the chain originates in the trial process and its judgment.
No scholar or practitioner thinks that trial’s only purpose is narrative, but for those who accept the theory, creating definitive narratives is central.10 From this perspective, the Prosecution’s expansive strategy was a laudable effort to tell the story of the entire Yugoslav crisis through the prism of Milošević’s plan for a Greater Serbia.11 Judicial narrative is not identical to writing history or a search for truth, but creates a social consensus in which history-writing and shared understanding are possible.12
The hard, optimistic edge of this view has receded in recent years.13 To retrench, however, is not to abandon, and even as scholars and practitioners sophisticate their claims, they have preserved a role for narrative.14 Individual trials may not create undeniable truths, but they introduce and vet evidentiary baselines that counter denialism and establish facts necessary to a broader narrative: For example, adjudicating a defendant’s mental state or deploying a theory such as JCE requires a court to say something about social contexts and war aims. Moreover, the Tribunal’s jurisprudence as a whole provides broader authority than any one trial. From this perspective, the Milošević Prosecution’s broad framing might not have produced a dispositive narrative on its own, but it could have contributed to one, and could still.*
Other observers question the idea that judgments catalyze transformative narratives.15 For these critics, courts’ technical procedures and legal focus make them ill-equipped to write histories.16 The fiction of combat balanced by a neutral judge arriving at fuller truth strains credulity: The extraordinary imbalance in resources between prosecutions and defense is well-known;17 for self-representing defendants such as Milošević, it evidently can reach to the limits of physical endurance.18
True justice requires a neutral or nonpartisan position, but the neutrality of court officers is often purchased through an almost total abstraction from and ignorance of the communities whose conflicts they adjudicate.19 Indeed, it is not clear that there is a simultaneously neutral and informed position: the questions most critical to those for whom an authoritative narrative is most needed are of a nature and complexity that will escape a neutral outsider.
The authoritative narrative theory also faces an empirical challenge: Quite apart from whether international courts construct informed, neutral narratives, can the narratives they do construct actually do what the theory supposes? This is a methodologically fraught proposition—how could one possibly prove the effect of this one variable?—yet the imperfect evidence we have does not support the theory: There is little evidence that reconciliation is occurring in the former Yugoslavia, or that individuals are converging on a common vision of the conflict, let alone that the ICTY has contributed to such a process.20 In the Balkans and elsewhere, it is not necessarily true that single narratives win out after conflict.
For these skeptics, the Milošević Prosecution’s strategy was a sprawling overreach that lost sight of the core purposes of having a trial and succumbed to the lure of a narrative of victimhood.21 The skeptics would prefer a narrow, forensic trial in which such aspirations are rigorously excluded when designing and managing trials.22 At most, a forensic trial might contribute indirectly to historiography, but a court’s judgment will have little value, even indirectly, if the process that produced it has not rigorously excluded the ambitious goals of narrative transformation.
It is not necessary to decide between these two views, however, because from either perspective, there will be a problem with a terminated trial. The weakest link in the narrative argument turns out to be the first one: the existence of a judgment. At rates that would constitute a scandal in a domestic jurisdiction, defendants in international tribunals die. Out of the 161 individuals indicted at the ICTY, 16 have died before the end of trial23—a mortality rate just under 10 percent. Some tribunals with smaller dockets have even higher death rates: Two of the Sierra Leone tribunal’s 13 indictments have been withdrawn due to death,24 and the ICC’s record is only slightly better.* Quasi-international trials confront similar problems owing to the antiquity of the underlying events, or because death can be strategic.†
When defendants die, trials stop. Because there are no mechanisms for rendering judgment post mortem or contra mortuum, by the authoritative narrative theory’s own logic, as the first critical link in the chain is never forged, the subsequent links never are either. Nor is the problem implicated only by physical death: Any trial that ends without a full evaluation of the substantive evidence should fail to produce the kind of judgment the theory requires.25
Final judgment is not the only part of a trial with value. The residuum of a terminated trial has uses—transcripts, exhibits, briefs, and decisions constitute an archive with a meticulous provenance any historian would value.‡ Acts and images—confession in open court, a shocking film—can have iconic or emotive value.§ The trial process, and even the very fact of trial, may encourage people to see the rule of law as an alternative to conflict.26 As Richard Holbrooke declared following Milošević’s death, “the trial was the verdict.… The Serb people came to understand the truth that he was not a nationalist, but an opportunist. A kind of rough and imperfect justice was served.”27
Yet valuable as these elements can be, they are structurally secondary. They are ancillary—their purpose, really, is to contribute to and demonstrate the integrity of judgment—and it is hard to imagine a robust argument for holding a trial if one knew it would end without judgment.28 What Holbrooke called “rough and imperfect justice” is simply the absence of authority; indeed, he called it that precisely to make up for the evident fact that its authority is deficient. Thus, the particular problem posed by Milošević is the lack of final authority and the consequent search—by those convinced judgment matters—for an alternative. For some people, that search lighted upon what looked like the nearest thing: the 2004 Decision on the Motion to Acquit.
In March 2004, after the Prosecution rested, the Amici filed a motion under Rule 98bis, requesting acquittal on a number of counts and allegations. The Prosecution filed a confidential response;29 Milošević filed no motion. The Chamber issued its Decision in June 2004, allowing each count to stand, but throwing out hundreds of individual charges; the immediate consequence was that trial continued.
At over 140 pages, the Decision goes into considerable detail about the legal standard applied and the individual allegations—much more detail than other Rule 98bis decisions.30 Defeating a motion to acquit places an extremely low burden on the Prosecution—higher than for indictment, but not by much.31 At the time, Rule 98bis provided for acquittal if a trial chamber “finds that the evidence is insufficient to sustain a conviction on that or those charges.”32 The test for sufficiency derives, as so much at the Tribunal does, from Tadić, through subsequent refinements, and confirmation on appeal in Jelisić:
“[T]he true test is not whether the trier of fact would actually arrive at a conviction beyond reasonable doubt on the Prosecution evidence if accepted, but whether it could;” or to put it as the Appeals Chamber later did in the same case, a Trial chamber should only uphold a Rule 98bis Motion if it is “entitled to conclude that no reasonable trier of fact could find the evidence sufficient to sustain a conviction beyond reasonable doubt[.]”33
Having identified a definitive test, the Milošević Chamber immediately clarifies how it will apply the test, saying it will acquit:
[w]here there is no evidence to sustain a charge… [; or w]here there is some evidence, but it is such that, taken at its highest, a Trial Chamber could not convict on it … even if the weakness in evidence derives from the weight to be attached to it, for example the credibility of a witness.… [; but w]here there is some evidence … such that its strength or weakness depends on the view taken of a witness’ credibility and reliability, and on one possible view of the facts a Trial Chamber could convict on it, [trial should continue].34
Sufficiency should be determined “on the basis of the evidence as a whole,”35 but the Chamber also considers “the sufficiency of the evidence [for a given charge] as it pertains to elements of [that] charge[.]”36 It was on this basis that the Chamber acquitted Milošević on hundreds of specific allegations while preserving each overall count.
In deciding, a chamber is obliged to take the evidence at its highest—to assume credibility unless it is utterly incapable of belief.37 Discussing liability under a JCE theory, for example, the Chamber noted that it
will not make a final determination as to the one or the other basis at this stage, that is, whether to acquit the Accused at this stage of one or the other basis of liability. The reason is that a determination as to the Accused’s liability depends to a certain extent on issues of fact and the weight to be attached to certain items of evidence, which calls for an assessment of the credibility and reliability of that evidence. These issues do not arise for determination until the judgement phase.38
The Decision consistently distinguishes between the Chamber deciding this case and “a Trial Chamber” when considering the standard of review: The Milošević Chamber will acquit only if no hypothetical reasonable trier of fact could convict—even if its judges already know they will not convict. The Chamber reiterates that “a ruling that there is sufficient evidence to sustain a conviction on a particular charge does not necessarily mean that the Trial Chamber will, at the end of the case, return a conviction on that charge[;]” it could acquit even if an accused called no evidence.39 This will sound familiar to lawyers in the common law tradition, because it is the classic motion for “no case to answer” from R v. Galbraith, which the Decision quotes at length.40
The standard of review suggests the Decision’s interim quality. If a chamber accepts a motion to acquit, this terminates proceedings—a consequential step that logically has the same effect as final judgment within the authoritative narrative model. But there is no valence to denying a Rule 98bis motion: Trial continues without prejudice to the sufficiency or meaning of the Prosecution’s evidence, apart from the minimal gloss that it was not, in the most positive light, impossibly inadequate.
This doctrinal framework was clear. How then did those various actors react to the prospect, and then the fact, of the Decision? Did those reactions change when what was supposed to be interim became terminal?
The different parties had very different interests in the Decision. Obviously, they desired different outcomes, but they, and the judges, also had different institutional roles that created not merely divergent interests, but asymmetries in how they interpreted and reacted to the Decision. Milošević’s rejectionism was predictable, but for the Amici, Prosecution, and Chamber, the calculus was different: Their roles assumed the Tribunal’s legitimacy. They needed to engage with the Decision substantively and procedurally, and had incentives to characterize it in particular ways to each other and the world.
Milošević did not need to take a position on the Decision; any pronouncement by the Tribunal had no authority because to him the institution was illegitimate ab initio.41* Despite this, Milošević took an increasingly active part in the trial42† and certainly had an interest in the Decision: Full acquittal would have been a vindication for him, and even partial acquittal would have brought the Prosecution’s case into question.
Whether or not Milošević perceived the potential to undermine the Prosecution’s case, his rejectionist strategy kept him from fully exploiting opportunities the Decision presented. For example, emphasizing that the Decision had acquitted him on some allegations would only have begged the question about all the others and the Chamber’s authority. Milošević’s grand strategy of public rejectionism encouraged him to a judicious silence;* after all, Milošević was not asking the Tribunal for judgment.
The Amici could hardly ignore the Decision; they had actively sought the power to file the motion.43 Thus, the Amici’s strategic position is conventionally contrasted with that of the Prosecution: seeking diametrically opposed outcomes, but in a structurally analogous position vis-à-vis the institution. In filing the motion for acquittal, the Amici appeared to be standing in for the Accused.
But to suppose that the Amici occupied the position normally filled by a cooperative defendant is not right because, as their name and the history of their appointment implies, the Amici were there to aid the Chamber. The Chamber first appointed the Amici when it became clear that Milošević was going to refuse counsel; their mandate was “not to represent the accused but to assist in the proper determination of the case” and to “assist the Trial Chamber[.]”44 Although the Amici could represent certain of Milošević’s interests, they were also circumscribed in that power and had obligations running to the Chamber; the Amici were far from true substitutes for the Accused, with whom they shared only an imperfect identity of interest.†
Consistent with their role as advisors to the Chamber, the Amici’s motion to acquit was restrained and neutral. They did not challenge the Prosecution’s core theory of liability and did not move for acquittal on all counts; instead, they challenged a narrower range of counts and charges for which the Prosecution had presented no evidence, which were charged under multiple counts,‡ or which relied on a particular test or interpretation the Amici contested.§ Only on Bosnia did they challenge more directly the Prosecution’s evidentiary basis—even then, apart from genocide, the Amici did not challenge entire counts, only specific allegations.45
The Amici’s brief “does not comment upon the sufficiency of the evidence called by the Prosecution on those counts upon which no submission to acquit has been made[.]”46 Perhaps the Amici thought they were refusing to concede other matters, but of course, they were the ones deciding what to seek acquittal for. By not commenting, they all but ensured that everything else would be passed through to the defense phase. Boas calls the Amici’s “reticence” to challenge the Prosecution’s core theory “curious.… Perhaps the prosecution can consider itself fortunate this avenue was not pursued further.”47
Of course, neither an amicus nor a defense attorney should necessarily challenge everything. The question on a motion to acquit is not “will the evidence suffice,” but “might it suffice if uncontradicted.” An actor playing by the rules, as the Amici were, ought not challenge evidence that, on its face, requires a defense—and it might be unwise to do so. Playing by the rules, however, does not preclude pressing one’s advantage. The Amici did not conceive of their role this way. For example, regarding Sarajevo, the Amici challenged all but one sniping and one shelling allegation; leaving those in place ensured that, even if the Chamber accepted all the Amici’s arguments—as in fact it did—the counts would survive.48 This appears, in fact, to have been the Amici’s intention.49
This restraint and neutrality—though a professional response to the Amici’s ambiguous structural position—undermines the utility of the Decision. If the Amici were constrained by their lack of identity of interest with Milošević, then it is unclear to what degree the Decision represents a robust adjudication in the way the Tribunal’s adversarial process supposes. Indeed, the consequences of the Amici’s imperfect identity with the Accused are all the more acute precisely because of the adversarial process, for which zealous defense is assumed and essential. A less adversarial process might make the disjuncture between defendant and friend of the court less disruptive of ICL’s narrative goals, but that would have required the ICTY to be formed on entirely different principles.
The Prosecution approached the impending conclusion of its case and a possible Rule 98bis motion in an atmosphere of doubt. Acquittal on even a few counts in its flagship case could have been catastrophic for the Prosecution, and many observers thought the Prosecution had failed to assemble a compelling case on the critical genocide charges.50 Still, the Prosecution could not openly contest the legitimacy of an adverse outcome; its options would have been restricted to procedural maneuvers and critiques of the Chamber’s reasoning, within a framework of institutional buy-in that compelled acquiescence. For the Prosecution, Milošević’s strategy of defiant nonchalance was simply not available.
Initially, the Prosecution placed its hopes on a Rule 98bis motion not being filed at all; they objected that the Amici were not “a party to the proceedings”51 as the Rule requires.52 When the Chamber rejected this preemptive strategy,53 the Prosecution turned to a conventional defense of its case. Its approach was technical and correct, but far less conflicted than the Amici’s—conceding the absence of evidence on numerous allegations, but still urging the Chamber to maintain all counts.*
After the Decision came down, the Prosecution undertook a rearguard action to contextualize the damage and preserve the scope of its case.† The Chamber ordered the Prosecution to produce redline indictments54—striking out allegations that were no longer part of the case—and it was against this version that Milošević and assigned counsel defended.55 The new versions56 are mostly sober, technical realizations of the Chamber’s instructions, but the Prosecution also added comments to justify marginal calls, which inevitably preserved allegations.
The Amici had challenged many specific allegations in the schedules attached to the indictment, but usually not the allegations in the main text, and the Prosecution used this fact to preserve mention of specific municipalities in the main text for several counts.57 For a persecutions count from Bosnia, for example, the Chamber ruled that allegations for three named municipalities in the schedules lacked evidence; nonetheless, the Prosecution did not line them out in the main text, arguing that “[i]n the Prosecution’s opinion[,] findings of insufficient evidence in relation to the Schedules of the Bosnia Indictment do not necessitate the deletion of [those] municipalities … as the allegations contained in [the main text] are broader than the specific allegations contained in the Schedules.”58
The most striking changes are to the Sarajevo sniping and shelling counts. As we have seen, the Amici challenged 43 sniping and 26 shelling incidents, leaving just one of each—the shooting of Seid Solak and the notorious Markale incident that killed 66 people. The Prosecution conceded that all the other incidents lacked specific evidence, but argued that “‘overview evidence’ of a shelling and sniping campaign in Sarajevo … is sufficient” for conviction.59 The Chamber rejected that argument and ordered acquittal on all except the incidents not challenged by the Amici.60
Still, although it had to redline these incidents, the Prosecution was able to preserve the original chapeau, describing the
killing and wounding [of] thousands of civilians of all ages and both sexes.… an extensive, four-month shelling and sniping attack … a protracted campaign of shelling and sniping.… [against] civilians who were, amongst other things, tending vegetable plots, queuing for bread or water, attending funerals, shopping in markets, riding on trams, [and] gathering wood.61
The chapeau even continues to note that “[s]pecific instances of sniping are described in Schedule E.… Specific instances of shelling are set forth in Schedule F.”62 The instances are still there—with lines through them. No changes appear in the main text, however, and anyone reading the indictment would not notice anything.
The Prosecution also preserved the seven overlapping counts—murder (as a crime against humanity and a violation of the laws or customs of war), inhumane acts, willful killing, willfully causing great suffering, cruel treatment, and attacks on civilians63—above an allegation set now consisting of two incidents. This is entirely correct—it is possible to characterize a single incident multiple ways—but it is also extraordinarily hollowed out. As the Chamber did not consider the Prosecution’s “overview evidence” sufficient to preserve the other sniping and shelling allegations, what exactly remained aside from the single pleaded instance of each? Yet, the preservation of all counts—and the original text—masks this, leaving the impression that nothing has changed, that no doubt or qualification has entered the juridical calculus.
The Prosecution encouraged this impression by suggesting that the redline indictments are not amendments at all, but only
working documents of value to clarify the case that the Defense has to meet. It may be thought that the Croatia and Bosnia Indictments should remain unchanged from the moment that the Prosecution case closed in order to best assist the parties and the public to be able to assess what was alleged, what was the subject of acquittal by the Trial Chamber and what (if any) counts are the subject of convictions in due course.64
This is technically unobjectionable, but also asserts a mystifying continuity between the original indictment—containing allegations on which Milošević had been acquitted—and the subsequent revision. At the time, all this was done to preserve options for trial, but later it would serve a different purpose.
For the Chamber, the Decision’s legitimacy was not in question; the judges had no interest in minimizing the effects of a document they had written. Their concern was with its impact on the trial and the ICTY’s jurisprudence—in particular, how the Decision affected other Rule 98bis proceedings. One judge’s subsequent behavior suggests he was appalled at what they had produced, and took steps to avoid a repetition.
The most significant tension arising in the Decision involves the role of judges as the ultimate trier of fact. As we have seen, the Decision discusses at length the relationship of Rule 98bis to the claim of no case to answer. This is an example of common law influence, yet the rule produces a very different dynamic with very different implications at the ICTY.
The judges felt it important to distinguish Rule 98bis from its antecedent, owing to a key structural difference between the ICTY and the common law: the lack of a jury. In the common law, the jury is normally the trier of fact,65 while the judge acts as the trier of law. The Decision notes that “an essential function of the [common law] procedure is to ensure that at the end of the [p]rosecution’s case[,] the jury is not left with evidence which cannot lawfully support a conviction[.]”66 But this protective function has limits: “a judge must not allow a submission of no case to answer because he considers the prosecution’s evidence to be unreliable, [because] by doing so he would usurp the function of the jury.…”67
Even though the Decision discourses at length how the common law rule protects the jury’s role, it does not actually mention the obvious disjunction that there are no juries at the ICTY. Only in his separate opinion does Judge Robinson note that
in principle, there is far less danger of an unjust conviction at the Tribunal than in criminal proceedings in common law jurisdictions.… [S]urely the fact that a Trial Chamber is composed of professional judges, whose need to be insulated from weak evidence is not as great as a lay jury, must make a difference.…68
Robinson adduces the possibility of a different rule from the ICTY’s different structure. As it turns out, his agenda is to cut back Rule 98bis, leaving more to the judgment phase:
Charges at the Tribunal are multilayered to a degree that is generally not present in indictments at the domestic level.… Is it useful to devote the Tribunal’s resources to an exercise which may result in the elimination of a dozen of these hundred or more individual allegations or details … while the charge or count remains intact? Is there any prejudice to an accused in leaving those dozen individual allegations for consideration … at the judgment phase?69
He therefore suggests restricting Rule 98bis to submissions
designed to eliminate a charge or count rather than individual allegations … [and] that allege that there is no evidence, as distinct from insufficient evidence[.]70
Robinson’s critique rests on claims about judicial economy and judges’ professional perspective.71 We might question this: Although much effort would be saved at the midpoint of trial, the defense would feel compelled to address every allegation in its phase, as it would not know if the judges thought them strong or weak. Perhaps Robinson’s view is a function of his frustration at having just waded through the entire Rule 98bis process: The enormous, tedious* Decision took months to write, but then Robinson never got to final judgment, when—had his preferred approach been in place—those claims would have reappeared.†
Surely the more relevant factor is that ICTY judges are the ultimate trier of fact. Concern with usurpation of the jury’s function is irrelevant in the ICTY’s unitary model—not, as Robinson supposes, because of judges’ skill, but rather the total identity between judge and jury. The Chamber would only be usurping itself, so the Rule 98bis “no case to answer” motion is, in a sense, a solution without a problem. It is not the judges’ professionalism, but their position, that makes Rule 98bis pointless. Still, well thought out or not, this signals that the judges are not signaling their own view about the evidence’s sufficiency.72
The Decision preserved all counts, but Judge Kwon dissented on genocide under a JCE I theory. JCE I requires that an accused actually possessed the special intent necessary to commit genocide. The Chamber accepted that the Prosecution had supplied sufficient evidence to continue with that charge, but Kwon disagreed.73 He did not dissent on other theories (such as JCE III, aiding and abetting or complicity) that contemplate liability even if an accused merely knows about or foresees genocidal harm, but he did not believe it was reasonable to infer Milošević’s own intent from the evidence.74
Kwon’s dissent did not affect the outcome of the Decision, but it delimits elements of a putative final judgment, because of the asymmetrical information embedded in a Rule 98bis decision. Denying a motion continues the trial, but acquittal has a clear, incontrovertible meaning. We cannot know how the other two judges would ultimately have ruled on JCE I liability, but unless there was new evidence in the defense phase, we can be certain about Kwon: It is difficult to see how Kwon could not have acquitted on any theory requiring Milošević to have actual genocidal intent. Whether Kwon’s view would have commanded a majority or not, we cannot know; the very limited information in the Decision does not work in that direction.*
…
Though neither they nor anyone else could know it then, the Decision was—or rather turned out to be—the judges’ last major statement on the evidence. In time, this made it into something more than its authors originally anticipated. Its appeal as a source for narrating the events of the Yugoslav wars turned out to be as considerable as it was doctrinally implausible—for the very feature that makes the Rule 98bis procedure pointless is the same feature that later made it seemingly attractive for the authoritative narrative theory.
It is precisely the total identity of interest between the ICTY’s triers of fact and law that makes the Decision potentially appealing. A decision to continue a case at the ICTY might be thought to signal a likely final outcome with greater confidence than a similar decision in the common law. Because the judges hearing a Rule 98bis motion are the same ones who will decide the case, it was, and is, tempting to suppose one could read a great deal about the meaning of a terminated trial into an interim pronouncement. But as almost everyone who later attempted to use the Decision this way was compelled to conclude, this seeming value is deceptive.
When Milošević died in March 2006, the Decision—an interim document that ordinarily would be superseded—suddenly gained a potential prominence in the vacuum of indecision and inarticulation. For some, it became a kind of judgment.
A realistic assessment of the Decision’s effects must take into account all uses, and the Decision’s deployment in other cases and by other courts—however logically suspect—is its own proof of influence. Still, the doctrinal weakness of these uses—evident in the tentativeness with which even those most eager to deploy the Decision do so—exercises its own constraint on the Decision’s utility as an ersatz judgment. Normally, realism encourages a skeptical attitude toward doctrine, but here we see how doctrine constrains the uses for which the Decision can be plausibly deployed; not just any argument will do.
In her press statement following Milošević’s death, the Chief Prosecutor invoked the Decision in the context of a mountain of evidence and a nearly completed trial:
I deeply regret the death of Slobodan Milosevic. It deprives the victims of the justice they need and deserve.… During the prosecution case, 295 witnesses testified and 5000 exhibits were presented to the court. This represents a wealth of evidence that is on the record. After the presentation of the prosecution case, the Trial Chamber … rejected a defense motion to dismiss the charges for lack of evidence, thereby confirming, in accordance with Rule 98bis, that the prosecution case contains sufficient evidence capable of supporting a conviction on all 66 counts. The Defense was given the same amount of time as the prosecution to present its case. There were in total 466 hearing days. 4 hours a day. Only 40 hours were left in the Defense case, and the trial was likely to be completed by the end of the spring.75
Almost everything here is accurate.* Claiming the trial was nearly over, besides heightening the poignancy and sense of waste, accords the Decision an aura of considerably greater finality than even Del Ponte would have assigned it just the morning before, let alone when it was issued. Everything needed was available—just a few hours missing. This is a half-truth, yet what is interesting is the implication: Del Ponte simultaneously acknowledges what she must—there will be no verdict—and claims that nonetheless there can be. Observers could make their own.
Considering the Prosecution’s earlier, assiduous efforts to render the Decision as invisible and inconsequential as possible, this public invocation is ironic. It also suggests the Prosecution understood the Decision’s strategic possibilities—its potential to lend an air of finality—even while being compelled to acknowledge the Decision’s doctrinal limitations.
Just under a year later, the ICJ rendered its judgment in the Bosnian Genocide case,76 which had acquired even greater salience as an implicit second chance to demonstrate Milošević’s responsibility.77 The ICJ acknowledged the “unusual feature” that “[m]any of the allegations before this Court have already been the subject of the processes and decisions of the ICTY.”78 Thus, when the ICJ ruled that Serbia had not committed genocide,79 its failure to acquire all the evidence from Milošević—in particular, minutes of the VSO—became a focus for frustration.80 But what, if anything, could the ICJ have done with the Milošević Decision’s review of evidence?
The Bosnian Genocide judgment extensively discusses the value of evidence from the ICTY. Although not a criminal jurisdiction, the ICJ applied a high standard of proof: “claims against a State involving charges of exceptional gravity must be proved by evidence that is fully conclusive.”81 It considers the ICTY’s “fact-finding process” to be of the kind that “falls within [its preferred] formulation, as ‘evidence obtained by examination of persons directly involved,’ tested by cross-examination, the credibility of which has not been challenged subsequently.”82
The ICJ also expressly discusses the evidentiary value of Rule 98bis decisions.83 It accurately paraphrases the Jelisić (and R. v. Galbraith) standard that such decisions only mean a court could convict, not that it would, and notes that the Krajišnik Chamber ultimately acquitted on genocide after having dismissed a Rule 98bis motion.84 The ICJ then draws the critical conclusion: “Because the judge or the Chamber does not make definitive findings … the Court does not consider that it can give weight to those rulings. The standard of proof which the Court requires in this case would not be met.”85 Evidence from the ICTY is not enough, interim evaluations are not enough—only final judgments are definitive.86
Having set the bar high, the ICJ promptly limbos under it, citing the Milošević Decision several times. Here the judges consider if members of a group protected under the Genocide Convention were killed:
In the Milošević Decision…, the Trial Chamber found that many Muslims were detained in Luka camp in May and June 1992 and that many killings were observed by witnesses, it held that “[t]he conditions and treatment to which the detainees at Luka Camp were subjected were terrible and included regular beatings, rapes, and killings.” [and] “The witness personally moved about [twelve to fifteen] bodies and saw approximately 100 bodies stacked up like firewood at Luka Camp; each day a refrigerated meat truck from the local Bimeks Company in Brčko would come to take away the dead bodies.”87
Three paragraphs later, they reach their finding:
On the basis of the facts set out above, the Court finds that it is established by overwhelming evidence that massive killings in specific areas and detention camps throughout the territory of Bosnia … were perpetrated.… The Court thus finds that it has been established by conclusive evidence that massive killings of members of the protected group occurred and … the requirements of the material element … are fulfilled.88
Similarly, the ICJ invokes the Decision to discuss the Manjača Camp in deciding if Serbia “caus[ed] ‘bodily or mental harm’ within the meaning of the Convention.”89
These two camps were hardly the only incidents the ICJ considered. It is the fact that the ICJ’s judges used the Decision at all, however, and precisely in this pedestrian way, that is of note. Surrounding the ICJ’s citation of the Decision’s Luka Camp review are citations to judgments in Brđanin, Krnojelac, Stakić, Nikolić, Sikirica, and Jelisić, the report of the Commission of Experts, and General Assembly and Security Council resolutions.90 The ICJ treats the Decision exactly as it does final judgments: To the degree the ICJ’s finding “by overwhelming evidence” is built upon the Decision, it overreaches, but also grants a retrospective authority to the Decision.
The ICJ even reached right past the Decision, directly back to testimony from the trial:
The Applicant has drawn attention to certain evidence given by General Wesley Clark before the ICTY in the Milošević case. General Clark referred to a conversation that he had had with Milošević during the negotiation of the Dayton Agreement.…
…
General Clark gave it as his opinion, in his evidence before the ICTY, that the circumstances indicated that Milošević had foreknowledge of what was to be “a military operation with a massacre.” The ICTY record shows that Milošević denied ever making the statement …, but the Trial Chamber nevertheless relied on General Clark’s testimony in its Decision.…91
The Chamber indeed discussed General Clark’s testimony, but it is a clear misreading of the Rule 98bis process to say that the Chamber “relied on” it in any definitive sense; yet the ICJ assumes the testimony has probative value because it has been laundered through the Decision.*
The ICJ first announces a strict, skeptical standard forbidding reliance on anything but final judgments, but then reads profligately from Rule 98bis decisions.† In a way, it is a formalistic game: The ICJ could cure the defects in its discussion simply by considering the Decision’s evidence as its own, rather than as a derivative product of another court; it could have simply asked itself what it thought Clark’s statement meant, or heard Clark directly as a witness—or, for that matter, demanded its own copy of the VSO minutes from Serbia.92 But this solution merely returns us to the theme that evidence either has intrinsic value or not, but acquires none at trial until that process characterizes it in some definitive way. The ICJ standard confirms precisely that logic, which it then fails to follow.‡
In a scholarly article, a former ICTY prosecutor, Dermot Groome, offers a searching critique of the ICJ’s failure to more fully consider evidence from Milošević.93 Groome’s critique is premised on the significance of the Decision, however, demonstrating the creative, if limited, repurposing to which its “sufficient evidence” standard can be put.
Groome surveys the evidence concerning Milošević’s role as the pivotal figure in Serbian politics and the relationship of Serbia to genocide in Bosnia, and, while admitting that the Decision “carries none of the weight of a final judgment[,]”94 sees significance in the fact that the Prosecution’s case met the Decision’s standard of review. The evidence should have similar potential for the ICJ case, and that possibility implies an imperative:
[G]iven the Milošević trial chamber’s finding that there was sufficient evidence upon which a court could find Milošević guilty of the crime of genocide…a thorough inquiry into Bosnia’s claims before the ICJ required the ICJ to examine the evidence referred to in the [Decision] to adjudicate the case before it.95
This is possibilitative argument—that a thing is necessary merely because it is possible—making the interim Decision a mandatory writ for another review. Groome, an accomplished lawyer, admits what he must; his desire to deploy the Decision leads him to adopt an essentially procedural move, as the doctrinal structure leaves no other option. Groome rightly notes that the ICJ would have been hard-pressed to find Serbia at fault if the Decision had acquitted Milošević;96 the Decision’s actual, ambiguous outcome—continuation of trial—does not yield an equivalent, opposite obligation.
The judges who wrote the Decision were also concerned with its institutional effects on the Rule 98bis process, so one way of reading the Decision is as an influence on ICL jurisprudence. A conference of the International Law Association (ILA) illustrates how one can extract authoritative text from a terminated trial, but equally suggests the limits of that effort.
The ILA Conference Report, on the rules for the use of force, relies on the Decision to formulate the definition of armed conflict.97 The Report notes that the Decision favors the approach taken in Tadić over the more restrictive view in the ICRC Official Commentary to Common Article 3.98 This is a consequential difference—but the point is that in building its case, the Report is not constrained by the Decision’s interim nature because when the Milošević judges discuss legal tests, they speak in their own voice, and the strictures of Rule 98bis relax.
In the section the ILA Conference Report relies on, the Chamber—after noting the Amici’s contention that there was no armed conflict in Kosovo before 24 March 1999—discusses the Tribunal’s jurisprudence for seven paragraphs, during which a single, trivial mention is made of the Prosecution’s and Amici’s views, and none of evidence.99 The analysis is not couched in the hypothetical language of what a trial chamber might decide; instead, “[i]t is settled in the International Tribunal’s jurisprudence,” “[t]he Trial Chamber makes the following observations on the Tadić test” and so forth.100
Other sections of the Decision demonstrate a similar pattern. In the discussion of deportation and forcible transfer, there are more mentions of Prosecution and Amici disagreements, but the overall tenor is the same—the Chamber is speaking in its own voice.* Likewise, the Decision notes the Amici’s contention that the mens rea for genocide and command responsibility cannot be reconciled, but then says that “[o]n the basis of the Decision of the Appeals Chamber in … Brđanin, this submission is unmeritorious.”101 The Chamber is not simply considering if the Prosecution’s preferred rule could be applied by a reasonable judge; instead, it is dismissing the Amici’s view on the merits. Other Rule 98bis decisions display a similar confidence when deciding legal tests.102
This is the most confident language in the Decision, and consequently the most authoritative basis for relying on the Decision because the Chamber is expressing its actual view.103 Of course, where the Decision is most authoritative is also where it is most anodyne, regurgitating legal standards with long pedigrees in other cases. In the discussion of mens rea, for example, the Chamber dismisses the Amici by citing Brđanin,104 and this is typical.105
Logically, Rule 98bis decisions are the best place to change the law on how Rule 98bis decisions are made. But this only shows the Decision’s authority is greatest on a circumscribed set of legal and procedural issues that do not engage with the factual evidence of responsibility, let alone questions of shared narrative—which is to say, with the putative purpose of the entire exercise. The authoritative narrative theory relies on the procedural integrity of the trial process, but this does not mean it is concerned only with process—in the end, the theory rises, or falls, on the stories it tells about what happened. These are claims about the substance of judgment.
The postmortem deployments of the Decision have been limited and cautious. Some actors have tried to mobilize the Decision as a kind of ersatz judgment, but these efforts feel halfhearted, constrained by the doctrinal straitjacket Rule 98bis creates; even its most ambitious advocates acknowledge that the Decision is not a verdict. For anyone who has ever felt the pull of the most skeptical, critical views of law—that nothing drives legal analysis other than actors’ preferences—the Decision’s restrained deployment is a bracing riposte, a reminder that text, process, and doctrine matter. The very fact of this restrained nonuse—by actors with evident convictions about Milošević’s guilt—suggests the limited utility of anything other than final judgment in constructing claims that rest on judicial authority.
In particular, there is a notable absence of claims that the Decision—really, the Milošević trial in general—has contributed to an authoritative narrative, despite the hopes ringed round what was the most important of the ICTY’s trials. That advocates of the authoritative narrative school did not pick up the Decision and make more of it suggests they did not think it would advance their goals. Either they were uncomfortable with its specifics—concerned that it might point toward acquittal*—or they recognized that this less-than-judgment was structurally inadequate.
If a trial reaches final judgment, no one is terribly concerned with interim decisions; it is only when a trial is terminated that they assume retrospective importance. This suggests a mismatch between an interim document’s initial design and its later deployment. The Decision was not built to bear the load of final judgment, and using it for that purpose stresses the unfinished edifice. Logically, this is true for any evaluation of fact produced in a trial that ends before judgment.† Formal, legal authority may or may not do what its advocates claim; in terminated trials, it surely cannot.
For what does rejection of a claim under Rule 98bis mean? Accepting a motion to acquit is undeniably consequential, yet rejecting a motion to acquit does not have an equally unambiguous impact: Apart from its discussions of legal standards, not a single piece of factual evidence deployed in the Decision can be confidently assigned a definitive value, except those the Chamber rejected as insufficient. All we can know about this evidence—as an element of the Decision—is that it was not so insufficient, so ludicrous, as to compel the Chamber to acquit. If it is to contribute to a transformative narrative and to reconciliation, that evidence will have to be deployed in other ways, according to other theories about how communities reach consensus and reconcile after war.
The radically limited utility of the Decision does not mean a terminated trial cannot contribute to reconciliation in any way. Advocates of the narrative theory rarely argue that a single trial yields the full truth. The events of Milošević have been adjudicated in other cases before the ICTY that both rely on related evidence and cast light on issues Milošević did not resolve.‡ More broadly, processes outside of the legal system may have greater flexibility in deploying the evidence and Decision: Independent analysis and historiography produce narratives that can do as much informational and persuasive work as legal judgment.
The Human Rights Watch report Weighing the Evidence provides a strong example of the claim that the trial process has independent value. Weighing the Evidence cites extensively from the trial evidence to build an argument about what happened during the Yugoslav wars and who was responsible. The report describes its own standard of review:
Human Rights Watch did not attempt an exhaustive review of the evidence introduced a [sic] trial. Human Rights Watch did consider Milosevic’s cross-examination and defense and we did not include evidence where we felt Milosevic had raised valid questions in rebuttal as to the value of the evidence.*
The report reviews the evidence de novo, not filtering it through the Decision, although it often could have.† This asserts the autonomous value of the evidence, evaluated without reference to the Chamber’s formalistic standards. In theory, an outside observer could disagree with the Decision: Even if the Chamber threw out an allegation, Human Rights Watch could in effect reinstate it through its own analysis; equally, it could acquit Milošević on a count the Chamber had upheld.‡
Likewise, accounts of the trial, such as Boas’ and Armatta’s books,106 which draw on the formal evidence but also analyze the trial process, produce accounts that may in time embed themselves in popular memory as definitive retellings of Milošević’s responsibility. So far these works have been produced by insiders or close observers, but the Milošević archive also is—or might be§—a rich source for historians to construct accounts that might acquire a kind of authority. Through these the trial could contribute to the processes the authoritative narrative theory contemplates: “In time, evidence introduced in the Milošević case may go some way toward vindicating these hopes [for recognition of crimes in Serbia]. Scholars and non-governmental organizations have begun what will likely be a long process of reclaiming that evidence and establishing non-judicial processes of learning from it.”107 Even if the theory in its pure form cannot extract value from a terminated trial, other narrative forms may be efficacious.
But whatever their effects, advocacy and scholarship will of necessity contribute to a different genus of authority than the specific kind supposedly afforded by judgment. Such work demonstrates both what can be achieved through private evaluation of evidence and its limits, which are defined precisely by the lack of authoritative imprimatur. Anyone can review evidence; it is not the facts but their legal characterization that matters to the particular kind of authoritative narrative ICL is supposed to produce. Historians and legal scholars can reinterpret evidence and even contest authoritative rulings, yet claiming too much for lay reinterpretation undercuts the original argument for trials, which is that legal decisions contribute specially to narrative and reconciliation. An outcome arrived at without going through the forensic processes of trial does not, by definition, create the particular effect with which we are concerned. In this sense, other approaches simply recapitulate the problem of authority.
The constricted horizons of the Milošević trial certainly suggest the tenuous nature of ICL’s immature project: so much depending on a single trial whose outcome depended on the health of a single defendant. Still, precisely because ICL is a young project, it is difficult to reach confident, empirically grounded conclusions about its effects. A terminated trial such as Milošević clearly cannot accomplish what the authoritative narrative theory expects; what is still not clear, however, is whether even final judgments can.
The two most critical features of the Decision are its interim quality and the trial’s termination; it would overreach to derive from them a claim about the effect of final judgment. However, this chapter’s inquiry into the relationship between terminated trials and the authoritative narrative theory does make two useful contributions to framing an investigation of that broader question, by removing some confounding factors from the puzzle of narrative reconciliation’s absence.
First, a close explication of the Decision shows just how structurally insufficient anything other than final judgment is for the creating authoritative narrative. If narratives of the kind the theory predicts were generated by other parts of a trial, their influence should be visible even in terminated cases, but it is not. Final judgments are the only plausible locus for further investigation of the narrative theory’s effects.
Second, this chapter further clarifies that investigation’s proper scope, because termination cannot be the only reason for the more general failure of reconciliation in the former Yugoslavia and the paucity of evidence for the effects the authoritative narrative theory predicts. If most trials ended prematurely, we might suppose this was limiting ICL’s reconciliatory effects. But although terminated trials are an endemic problem, they are not the norm; most trials reach final judgment, and therefore should have done something.
We ought to be skeptical about the special ability of international trials to create or contribute to transformative narrative. This chapter has focused on the specific problem of terminated trials’ incapacity to produce authoritative narrative, but in doing so it also defines the parameters of the general theoretical question toward which the Decision points us, which concerns the authoritative narrative theory itself. The Decision—that imperfect, interim document—by its very imperfection suggests the need to examine the sufficiency, not of its own transformative potential, but of the idea itself.