. 6
( 11)


focus of their efforts will be “ at least initially “ directed to redressing systemic
criminality in non-Western spaces, the end result is the exclusion of the local in
those places where atrocity is most likely to be criminalized.
There is a second behavioral permutation. Not all states may pursue mimicry.
Some states simply may acquiesce in an ICC exercise of jurisdiction or may self-
refer a matter to the ICC. The decision by the Ugandan government to self-refer
atrocity committed by a rebel group, the Lord™s Resistance Army, in northern
144 Atrocity, Punishment, and International Law

Uganda illustrates this phenomenon. The Ugandan situation is one of a num-
ber of cases, largely unexpected, where national and territorial jurisdiction join
in the same state and that state refers what is largely an internecine con¬‚ict,
albeit one with transnational implications and of serious concern to the inter-
national community as a whole, to the ICC.117 The Lord™s Resistance Army
has engaged in a nearly twenty-year-long con¬‚ict in which practices of crimes
against humanity, child abduction and soldiering, and sexual violence abound.
Much of this con¬‚ict is centered among the Acholi people of northern Uganda,
who are implicated on both sides of the violence. In the Ugandan context, it is
unclear whether the domestic court system is unable to prosecute or whether
domestic sociolegal institutions are unable to hold perpetrators accountable.
Yet, the ICC is investigating and has issued indictments.
Why did the Ugandan government self-refer this situation to the ICC? One
reason may well have been control. Fearing that the ICC Prosecutor may have
exercised his proprio motu power to investigate in any event, the Ugandan gov-
ernment may have self-referred out of an anticipatory hope that, were the ICC
to assume jurisdiction based on a self-referral instead of acting independently,
this would give the Ugandan government greater control over the situation. The
Ugandan government may have craved control for eminently rational reasons:
namely protecting itself and promoting its interests. Speci¬cally, the Ugandan
government may wish to obfuscate atrocities allegedly committed by its own
armed forces. Although the ICC has not precluded investigation of these spe-
ci¬c allegations,118 it is unclear how seriously they will be pursued. The arrest
warrants issued in October 2005 by the ICC “ its ¬rst “ were for the leader
and four members of the Lord™s Resistance Army (it has been reported that one
indictee was subsequently killed by the Ugandan army). At this juncture, it bears
mentioning that the Ugandan government is an illiberal regime with an uneven
human rights record. This is the same government that, while requesting ICC
intervention, was condemned by the International Court of Justice for violations
of international humanitarian law “ and ordered to pay reparations “ with regard
to its responsibility for unlawful armed activity, plundering, and massive human
rights violations in the DRC between 1998 and 2003.119
Ugandan political elites also may have turned to the ICC in the hopes it
would promote their own interests by targeting rivals120 and allowing elites to
manage, and dissuade, calls by local community leaders to settle matters through
traditional forms of dispute resolution. One example of a traditional practice is
mato oput (drinking bitter root herb). Another is nyouo tong gweno (a welcoming
ceremony incorporating eggs and twigs). To the extent that these forms of dis-
pute resolution gain currency, they portend a decentralization of power from
the centralized apparatus of the state or the state™s delegation of power to an
international organization.
Ugandans, particularly residents of the northern parts of the country most
affected by the violence, have expressed reservations to the idea of ICC interven-
tion.121 A number of parliamentarians and religious leaders from northern
Uganda in fact have traveled to The Hague to oppose the ICC investigation.122
Legal Mimicry 145

Joanna Quinn notes that “to the people of northern Uganda, the international
legal process is almost completely foreign.”123 Instead, members of victimized
communities value traditional approaches such as mato oput and nyouo tong
gweno.124 Many community members feel that these social institutions respect
the fact that the line between victimizers and the victimized, particularly in the
case of child soldiers, is opaque.125 In local eyes, the fact that the ICC was invited
by the Ugandan government spoils its putative impartiality.
Predictably, both mato oput and nyouo tong gweno have been subject to the
pressures of the internationalized legal paradigm. Desperate for some semblance
of these mechanisms to be invoked in Uganda™s settlement of these terrible atroc-
ities, “Acholi parliamentarians have drafted an addendum to the ICC bill, the
implementing law, to attach penalties to their traditional justice mechanism in
an effort to fall within the complementarity principle and prevent criminal pros-
ecution of such cases.”126 This suggests how complementarity initiates a drive
toward homogenization by massaging the traditional into the neotraditional.
Similarly, William Burke-White reports that, although the Congolese president
may have referred atrocity in the DRC to the ICC to discredit his political oppo-
nents, these opponents, who prefer that any prosecutions remain within the
domestic judiciary, in turn push judicial reform of the domestic courts so as to
reduce the likelihood that the ICC will admit these cases.127 The ICC, by virtue
of the complementarity regime, therefore plays a role in changing domestic
sociolegal structures, instead of serving as a temporary substitution or stopgap
for these structures.128 This brings the discussion back to the hypothesis that the
complementarity regime encourages mimicry.
Returning to the Ugandan situation, in addition to their sense that ICC justice
will be externalized justice (and hence that the justice payoff will be low), local
constituencies, comprised of people who actually live in the area ravaged by
violence, have expressed concern with the political effects of ICC indictments.
A Ugandan delegation actually implored the ICC not to indict the leaders of
the Lord™s Resistance Army because delegees felt that doing so removes the
bargaining chip of amnesty for such individuals in settling the country™s civil
war.129 Although amnesty tends to be a dirty word in the lexicon of international
criminal lawyers, it may not be so in the lexicon of local populations; moreover,
it appears that amnesties have more appeal130 and favorable long-term results
than international criminal lawyers may care to admit. Local communities in
northern Uganda also express concern that the ICC will not be able to guarantee
the security of those witnesses called to testify. Now that indictments have been
issued, as a matter of formal law amnesties seemingly have been pushed off the
table “ while, paradoxically, local pressures for peace discussions that actively
contemplate the prospect of amnesty increase in stridence. In 2006, the Ugan-
dan government guaranteed Lord™s Resistance Army leader Joseph Kony™s safety,
and even amnesty, in exchange for serious efforts to negotiate peace, which it
then participated in. The ICC and its Western backers promptly expressed deep
reservations about such a proposal, insisting that Uganda apprehend Kony and
refusing to drop its arrest warrants. Tension therefore is emerging between the
146 Atrocity, Punishment, and International Law

pursuit of peace, which is relevant to communities torn by strife, and satisfaction
of international arrest warrants, which is relevant to the functionality of inter-
national criminal law. At the time of writing, the situation in Uganda delicately
remains in ¬‚ux, although the prospect of amnesty apparently is what has stopped
the violence.131
In summary: even in cases where they self-refer, states may be animated by a
complex array of motivations, including the prospect of avoiding proprio motu
intervention by the Prosecutor. Consequently, instead of pursuing mimicry by
reforming domestic institutions, the state may invite the ICC to investigate. This,
however, leads to a similar (and perhaps even more troubling) result, namely
the simplistic superimposition of selective exogenous criminal law on terribly
complex con¬‚icts. The existence of the ICC may offer illiberal governments
a tool to consolidate power and avoid enfranchising the policy preferences of
af¬‚icted local populations by providing these governments an option to refer
matters to a distant institution focused on the reductionism of punitive criminal
law. The ICC thereby creates an option of exit for national governments to
externalize complex processes of justice onto a foreign entity. This availability
of exit creates a risk that little is done genuinely to place domestic sociolegal
structures in the service of postcon¬‚ict transition. In the case of Uganda, this risk
has been attenuated only by dint of the vigor of the Acholi community, which
has been unexpectedly forceful in the bottom-up articulation of its preferences.
Partly as a result of Acholi mobilization and partly as a result of learning that
it cannot control the ICC intervention, Ugandan authorities, despite having
self-referred the matter in the ¬rst place, have begun to openly hedge regarding
the exact role that they wish the ICC to play.132
One reaction to my concerns over the remodeling effects of complementarity
is that I overstate these effects. After all, the ICC targets “the most serious crimes
of concern to the international community as a whole.”133 At ¬rst blush, the ICC
appears to be designed only to pursue a handful of leaders. Arguendo, the ICC
would have limited interest in the vast majority of perpetrators, namely those
lower-level offenders who undertake the dirty work of atrocity. Consequently, it
follows that the ICC would have little to no interest in how national and local
institutions deal with such offenders.
In response, it is not altogether clear that the ICC never would prosecute
lower-level offenders. In some places, such prosecutions may be necessary in
order to begin to compile the judicial record with regard to higher-level offenders
and develop familiarity with the facts of the atrocity. After all, the ICTY™s initial
prosecutions involved low-level thugs, such as Dusko Tadi´ , and one individual
of diminished mental capacity; even a more recent conviction of great jurispru-
dential value “ namely, criminalizing and punishing sexual violence “ involved
a relatively low-level paramilitary commander named Dragoljub Kunarac. Fur-
thermore, the East Timor Special Panels essentially only prosecuted lower-level
offenders. But, assuming arguendo that the ICC™s focus will be on high-level
defendants, I would agree that the ICC should prove to be predisposed to act
generously toward national and local institutions with regard to how they process
Legal Mimicry 147

low-level offenders. And I would certainly agree that an international system
keyed to a handful of perpetrators is less invasive in its migrations than one that
explicitly captures all perpetrators.
However, national and local actors will take their cues, and model their
behavior, from how international institutions process those deemed most respon-
sible for atrocity. Gauging by international responses to gacaca in Rwanda, and
traditional mechanisms in Uganda, there is in fact palpable concern over pro-
cessing lower-level perpetrators in a manner that deviates from the norms of
international institutions. The fact that international institutions exert in¬‚u-
ence over the modalities of accountability for individuals in whom they have
little, if any, interest merely attests to the in¬‚uence these institutions wield.
In sum, although by virtue of complementarity “the majority of prosecutions
for international crimes are expected to take place in domestic courts,”134 the
form of these proceedings may become monochrome. Admittedly, it is probably
inevitable that international institutions exert some downward pressures on the
process of domestic law. It also is probably unavoidable that domestic law has
some trickle-down effect on local law. In both cases, these pressures also may
create considerable good. However, should these pressures lead to the external-
ization of justice and the creation of democratic de¬cits, then they no longer
serve salutary ends. Therefore, international lawmakers ought to consider how to
control these pressures so as to minimize their undesirable effects. In this regard,
Chapter 7 proposes to substitute quali¬ed deference for complementarity as a
pluralistic interpretive lens for the vertical application of authority. Quali¬ed
deference gives more leeway to local variation from the trial and punishment
modalities of contemporary international criminal tribunals.

(vi) conclusion
The international community increasingly is holding atrocity perpetrators
accountable. That said, the accountability process remains narrowly oriented to
incarceration following liberal criminal trials. It is not a broader process that is
yet comfortable with meaningful restorative initiatives,135 indigenous values,136
quali¬ed amnesties, reintegrative shaming, the needs of victims, reparations,
collective or foreign responsibilities, distributive justice, or pointed questions
regarding the structural nature of violence in the international system. Interna-
tional criminal law pursues some individuals “ cast as enemies of humankind “
in some places. In so doing, it punishes these individuals and, thereby, cleanses,
puri¬es, and salves. This process, however, conveniently or unwittingly swad-
dles the myriad structural factors that permitted the guilty to perpetrate evil on
such a large scale. With pronouncement of sentence comes a rush to closure,
absolution for the acquiescent, and the evaporation of collective responsibility.
This results in the punishment of certain individuals but does not lead to the
reform of criminogenic conditions. Scholars of international crime have not
yet satisfactorily examined the relationship between these conditions and the
long-term peaceful resolution of disputes within and between af¬‚icted societies.
148 Atrocity, Punishment, and International Law

When the ascendancy of criminal trials discourages the development of alternate
approaches to accountability, the result is a troubling reductionism. Moreover,
given the dif¬culties criminal trials experience in attaining their stated retribu-
tive and deterrent objectives in contexts of collective cataclysm, to which I
turn in the next chapter, there is some urgency to the investigation of alternate
rationales for and modalities of sanction.
Research suggests that lasting social order in societies roiled by internecine
con¬‚ict is restored by a “forgiveness process characterized by truth telling, redef-
inition of the identity of the former belligerents, partial justice, and a call for a
new relationship.”137 Assuredly, criminal trials could form an element of some
of these goals, notably partial justice and truth telling.138 But restorative mech-
anisms also could form an element of these, and other, goals as well. Victimo-
logical surveys indicate that aggrieved individuals seek polycentric mechanisms,
including those consonant with restoration.139
This is not to say that restorative initiatives by de¬nition always are salutary.
Some restorative justice initiatives, such as the Sierra Leone Truth and Rec-
onciliation Commission, served ritualistic importance, but did not actualize
local reconstructive practices nor stimulate much in the way of truth-telling.
Restorative modalities are no panacea; local justice must not be sentimental-
ized. Restorative modalities that draw parallels from mechanisms used to rein-
tegrate ordinary deviant transgressors in settled times will likely run afoul of the
complexities of reintegration in situations of mass atrocity. Restorative shaming
theory predicated on a majority of the community™s disapproval of the impugned
conduct may not be directly transposable to contexts where a majority of that
community may not have actually disapproved of atrocity. Moreover, restorative
mechanisms that inject alien methodologies will likely prove of limited effect.
We cannot blithely assume the suitability of a truth commission whose logos is
one of Western psychoanalytic theory generalized from the single patient to an
entire society. It is critical not to implement restorative mechanisms that may
be faulted for the same kind of externalization and transplantation that shadow
internationalized criminal process.
Postcon¬‚ict justice is terribly and terri¬cally complex. There are no sim-
ple solutions. Chauvinism that views truth commissions as a one-size-¬ts-all
hegemonic remedy succumbs to the same frailties as judicial romanticism.
Consequently, one important lesson is the need to avoid methodological parsi-
moniousness. Instead, consideration should be given to consolidating diverse
mechanisms more closely attuned to the social geographies of the af¬‚icted
chapter 6

Quest for Purpose

The stated values of the punishment of extraordinary international criminals
principally are retribution, deterrence, and expressivism. In this chapter, I con-
sider whether extant sentencing modalities at the local, national, and interna-
tional levels attain these aspirations. I conclude that, although these modalities
go some way to meet retributive and deterrent goals, they fall well short of oper-
ationalizing these goals in any meaningful sense. Extant modalities experience
greater, albeit still limited, success in attaining expressive goals.
At the outset, it is important to underscore that the three theories of pun-
ishment discussed here are not mutually exclusive. Despite the potential for
tensions among these theories, courts often refer to them overlappingly when it
comes to punishing a single defendant. This creates some tautness insofar as the
goals of deterrence (to punish to prevent future crime) at times may con¬‚ict with
those of retribution (to punish because the criminal deserves it). On the other
hand, this overlap also may generate synergies. For example, retribution may
have some positive utilitarian effect in deterring violence by discouraging vig-
ilantism among the general public during periods of political transition.1 That
said, the French prosecution of 100,000 collaborators following liberation from
the Nazis did not quash vigilantism, insofar as thousands of individuals believed
to be collaborators were privately killed. Furthermore, despite the existence of
the ICTY, Kosovo Albanians carried out “revenge killings” against Serbs in 1999
following NATO™s Operation Allied Force; despite the existence of the ICTR,
the RPF committed violent reprisals against Hutus.2 Retribution and expres-
sivism also share connections: a public that sees a wrongdoer punished in a
manner that accords with perceptions of that individual™s just deserts can aug-
ment the value of the legal system in the eyes of that same public. This prospect,
in turn, has given rise to the notion of expressive retribution, which has emerged
in recent ICTY jurisprudence.
Punishing atrocity perpetrators on occasion purports to promote other goals,
such as rehabilitation, incapacitation, and reintegration. Insofar as these goals
remain on the penumbra of sentencing practice, I do not consider them here.3
This is not to deny their normative worth. They are, in fact, immensely important
to any project of transitional justice. International criminal law, however, has
150 Atrocity, Punishment, and International Law

not yet accorded these goals much in the way of jurisprudential emphasis and,
in cases where such emphasis ¬‚eetingly has been given, the goals remain “ as
set out in Chapters 3 and 4 “ rather poorly operationalized.
Reconciliation, another goal, has been given some attention in the judg-
ments of the ad hoc tribunals and a little more so in those of the East Timor
Special Panels. Much of this attention, however, is rhetorical, in that inter-
national legal institutions expend little effort in practically (and consistently)
thinking about how their punishment schematics actually can be made to help
victims and offenders (not to mention victim and offender communities) rec-
oncile. National and local institutions, perhaps because they are more deeply
embedded in domestic transitional frameworks, often are forced to engage more
practically with the operationalization of reconciliation. To this end, reconcil-
iation has somewhat greater currency in certain in situ institutions, such as
gacaca; on the other hand, gacaca™s reconciliatory potential is crimped by its
operational structure. Overall, there is insuf¬cient evidence to support the inclu-
sion of reconciliation among the principal objectives that existing institutions
ascribe to the imposition of sentence upon extraordinary international crimi-
nals. There is much to be said in favor of reconciliation as an objective, both in
theory and in practice, especially given its on-the-ground importance in many
af¬‚icted communities.4 The challenge, however, is to assess how much reconcil-
iation actually can be generated by criminal trials. In Rwanda, although national
criminal trials have developed a jurisprudential record and have involved ten
thousand defendants, the promotion of national reconciliation is not among
their successes.5 Laurel Fletcher notes from her research in the Balkans that,
although under certain circumstances trials can contribute to what really is a
very private personal decision to reconcile, these circumstances are not common
and, hence, it is dif¬cult to generalize correlations between trials and collective

(i) retribution
Immanuel Kant understood retribution to mean that criminals should be pun-
ished because they deserve it.7 For the retributivist, criminals are not mere cogs
in a process of social engineering. Rather, they are ends in themselves “ actors
deserving of condemnation. G.W.F. Hegel, although generally holding to a
restorative view of justice, also recognized the merit of retribution. Hegel noted
in Philosophy of Right that “as the criminal has done, so should it be done to
Retribution is the dominant stated objective for punishment of atrocity per-
petrators at the national and international levels.9 In practice, though, extant
punishing frameworks experience a number of challenges in attaining their
retributive ambitions. These challenges, which operate at both the national
and international levels, include three distinct phenomena: (a) selectivity;
(b) severity of sanction and discretion of sentencing judges; and (c) plea bar-
gaining. Although each of these phenomena may promote a variety of salutary
goals, each also hampers the ful¬llment of retributive aspirations.
Quest for Purpose 151

a. Selectivity
The retributive function is hobbled by the fact that only some extreme evil gets
punished, whereas much escapes its grasp, often for political reasons anathema to
Kantian deontology. Assuredly, I recognize that criminal law always is contingent
on politics. Selectivity is inevitable in the operation of law even in a robustly
ordered and purportedly egalitarian domestic polity. However, as Robert Cryer
notes, selectivity poses a greater challenge to international criminal law than it
does to national criminal law.10 The contingency of international criminal law
is pronounced, including when enforced by international institutions.
Only a few atrocities ever become judicialized. Diane Marie Amann notes
that “[a] random con¬‚uence of political concerns produced ad hoc tribunals for
just two out of a number of con¬‚icts that warranted such treatment.”11 These
inconsistencies do not eviscerate the retributive value of punishing the guilty in
Bosnia or Rwanda. Rather, they underscore the dif¬culty in ascribing retributive
purposes to international criminal law as a whole when a “con¬‚uence of political
concerns,” and not the inherent gravity of the crimes, prods the punishment of
offenders. Owing to these concerns, and the contingencies they sow, plenty of
perpetrators of extremely grave crimes simply avoid any entanglement with a
punishing institution.
Even when a punishing institution is established, however, the reach of the
criminal law only attaches to a small subset of alleged perpetrators. The ICTY
Prosecutor, for example, has been forced to select a modest number of cases
from many thousands of targets.12 Selection decisions often are discretionary in
nature.13 Prosecutorial discretion tends to be exercised in favor of those cases
where there is a better chance of securing a conviction. In some cases, the
better chance arises because of the inherent gravity of the crime, in that the
crime was planned, organized, brutal, and vast in scope and, therefore, left
a deep evidentiary footprint. In these cases, an overlap with retributive goals
may arise, in that discretion is exercised in favor of the worst cases where the
perpetrators most deserve to be punished. However, in other cases, the exercise
of prosecutorial discretion is contingent on variables (such as the cooperation of
states, utility of convicting a low-level thug for strategic purposes, and availability
of material resources) that have little to do with the inherent gravity of the alleged
At the ICTR and the East Timor Special Panels, selectivity arises insofar as the
jurisdiction of these punishing institutions is formally or practically limited to
an arti¬cial and politically convenient time frame. Large numbers of killers and
killings are therefore left unexamined. Katzenstein offers the following explana-
tion for the truncated temporal jurisdiction of the Special Panels:

Limiting the investigations exclusively to referendum-related violence of 1999,
despite a mandate that provides for jurisdiction over acts committed during
a much broader time frame, was not simply a decision based upon resource
constraints. Rather, it was also motivated by a concern that a more expansive
inquiry could lead to the indictment of U.S. of¬cials who countenanced the
152 Atrocity, Punishment, and International Law

Indonesian invasion and helped to equip and train the Indonesian military
both prior to and throughout the occupation.14

At ¬rst blush, there is less selectivity in a permanent institution, such as
the ICC, than in ad hoc institutions created by the UN Security Council.
And, indeed, to a large degree the ICC operates independently of the Security
Council.15 However, selectivity intractably affects, and will continue to affect,
the ICC™s work. The Rome Statute places considerable power within the of¬ce
of a single individual: its Prosecutor, currently Luis Moreno-Ocampo.16 The
ICC Statute “is almost totally silent with respect to the larger policy questions
about which potential accused should be pursued by the Prosecutor.”17 There is
limited judicial oversight of Prosecutorial decisions not to investigate. Although
there may be very important and eminently justi¬able reasons for the Prosecutor
to decline to investigate or prosecute “ reasons that I would avidly support “ it
remains that these reasons, however laudable, if applied to extremely grave cases
undermine retribution as a principled objective.
Moreover, regardless of institutional oversight of the ICC Prosecutor™s dis-
cretion, it is impossible to squeeze out the political contingency of criminal
liability in the ICC™s practice.18 Looking ahead, there will likely be a large dis-
parity between the cases the ICC could potentially prosecute and those that it
will effectively prosecute.19 The ICC Prosecutor, for whom resources remain
limited, will face “competing situations of crisis.”20 Ineluctably, this means that
only some crises will be selected for investigation and prosecution. The Rome
Statute provides limited guidance regarding how to comparatively evaluate cri-
sis situations.21 Pragmatically speaking, in order for the ICC institutionally to
maintain resource support, it is incentivized to investigate wrongdoers in polit-
ically powerless places.22 Decisions whether or not to investigate or prosecute
therefore become contoured by concerns over how they affect the ICC™s polit-
ical standing, funding, and support among states. Cases may be turned away
because of politics and initiated because of politics, instead of cases initiated or
turned away solely because of the gravity of the alleged violations of international
law that they actually present. In the end, the permanent ICC could de facto
resemble an ad hoc institution contingent on international political consensus.
However, even when such consensus exists “ as is the case with the Security
Council™s referral of the Darfur situation to the ICC “ resource availability will
affect the ICC™s ability to do its work. In the Darfur situation, for example, it
does not appear that the Security Council is paying for investigatory or prose-
cutorial costs occasioned by the referral. Contingency and selectivity triggered
by funding vagaries may lead to a situation where culpable individuals evade
To his great credit, Moreno-Ocampo has expressed interest in examining the
broader context in which mass violence occurs, in particular links to interna-
tional economic dynamics and corporate behavior. This development would be
salutary (although it is limited by the fact the ICC only has jurisdiction over
natural persons). After all, prosecuting only a small number of individuals in
Quest for Purpose 153

cases of massive levels of violence leads to a very partial print of justice. Atrocity
is often the result of structural factors. Slobodan Miloˇevi´ , Saddam Hussein,
and Pol Pot emerged from deeply globalized forces, including acts and omis-
sions of international agents and foreign governments; the sources of genocide at
Srebrenica and Rwanda are complex and multicausal. Frankly, for many Rwan-
dans and Bosnian Muslims, retribution might well include accountability for
the UN and foreign governments, whose peacekeepers were ineffective while
genocidal massacre occurred in their midst. Just because these entities are not,
or cannot be found to be, criminally guilty does not mean that they are in no
way responsible for genocide. In the end, however, the operation of interna-
tional criminal law occasions a retributive shortfall in that too few people or
entities receive just deserts while many powerful states and organizations are
absolved of responsibility. So, too, are bystanders “ many of whom are not so
innocent. Although it may seem counterintuitive, restorative justice modalities
and institutions that push reintegrative shaming could in fact augment overall
retribution by capturing a far greater number of individuals and organizations
in the accountability process, albeit not to the severity or depth characteristic of
the criminal conviction.
At the national level, courts that adjudicate extraordinary international crim-
inals face many of the same selectivity challenges that hinder their international
counterparts. National courts that assert jurisdiction based on nationality or ter-
ritoriality additionally face their own proximity to the violence and their own
susceptibility to domestic political pressures. They may be preoccupied with
maintaining their own legitimacy during periods of political transition. These
concerns contour decisions regarding who to prosecute and can disaggregate
prosecutorial decisions from the gravity of the underlying offenses. This cer-
tainly appears to be the case in national courts throughout the former Yugoslavia,
where many proceedings are corroded by ethnic bias, thereby undermining the
principled attainment of retributive objectives.
Furthermore, the retributive value of punishment at the national level can be
compromised by two diametrically opposite phenomena that do not arise at the
international level: (1) overcapture and (2) intentional undercapture through
Overcapture most acutely involves pursuing individuals for atrocity when
that pursuit is motored by concerns other than the gravity of the alleged atrocity
offense. For example, denunciations have been lodged with gacaca that appear
to be motivated by ulterior purposes of land acquisition, romantic disputes,
political vendettas, and relationship breakdowns. In such cases, one way to get
even with an uncooperative colleague or unfaithful lover is to denounce him
or her as having been involved in genocide. In some cases, accusations may be
entirely false, in other cases they may be partially true, and in some they may
be fully true but brought not to seek justice for acts committed in 1994 but,
rather, acts “ often not criminal “ initiated much later that have nothing to do
with genocide. In each of these scenarios, there is a dilution in terms of the
retributive value of the punishment that is meted out: this dilution is greatest
154 Atrocity, Punishment, and International Law

when punishment is undeserved, but also arises when punishment is pursued
because of concerns unrelated to the gravity of the initial atrocious conduct.
Due process can ¬lter out those denunciations that lack foundation and, hence,
serve a gatekeeping or corrective function with regard to overcapture.
Undercapture is trickier. National courts often must face the prospect of
amnesties. These erode the retributive value of prosecution and punishment for
mass atrocity or politically related violence. Generally, amnesties are unavail-
able for ordinary domestic crime, yet they arise frequently in situations of mass
atrocity.23 In some cases, amnesties are unquali¬ed. In other cases, they are
quali¬ed, in that they require the person seeking amnesty to tell the truth, apol-
ogize, or make amends. Whereas the amnesties (self-)accorded to Argentine and
Chilean military leaders were unquali¬ed, amnesty in South Africa was quali-
¬ed. Regardless, even in the case of South Africa, “[f]rom a retributive point of
view, it is not immediately clear why a murderer who kills for political reasons
should be entitled to amnesty in return for the truth, while one who kills out
of passion or greed should not.”24 This disparity in treatment is particularly vex-
ing for the retributivist because the gravity of the conduct of the extraordinary
international criminal is supposed to be greater than that of the ordinary com-
mon criminal. Assuredly, as noted in Chapter 5, much can be said in favor of
amnesties. Amnesties have many justi¬cations that can be coherently grounded
in moral theory, popular will, and pressing political realities. That said, these
justi¬cations “ however attractive “ do not attenuate the reality that amnesties
selectivize punishment of extraordinary international criminals at the national
level in a manner that hampers retribution as a principled penological goal.

b. Severity of Sanction and Discretion of Sentencing Judges
Retribution requires proportionality between the gravity of the offense and the
severity of sanction. In this section, I consider challenges to the retributive
metric posed by three realities revealed by the data presented in Chapters 3
and 4 regarding the sentencing of extraordinary international criminals. These
realities are:

(1) Regardless of the level at which punishment is imposed, sentences for
extraordinary international crimes are not generally longer than for serious
ordinary common crimes;
(2) Sentences for extraordinary international crimes are not as a rule longer
when pronounced by international tribunals than when pronounced by
national courts (nor are conditions of imprisonment harsher or stigma weight-
ier), even though international tribunals exercise jurisdiction over the most
serious offenders; and
(3) There is signi¬cant disparity within and among institutions when it comes
to the severity of sentence, and this disparity is not consistently explainable
on the basis of the gravity of the offense.
Quest for Purpose 155

Let us consider each of these realities in turn, and investigate how they
obstruct the retributive goals of international criminal law.

(1) Overwhelming Gravity of the Crimes. The data reveal that, at both the
national and international levels, sentences for multiple international crimes
are generally not lengthier than what national jurisdictions award for a single
serious ordinary crime. The length of a term of imprisonment is, obviously, not
the only possible indicator of retributive value. Nor is it evident that the mere
addition of several years to a sentence necessarily augments its retributive force;
or that shortening a sentence by several years guts that force. However, length
of sentence constitutes the central “ and, basically, only “measurement device
that liberal legalist institutions practically avail themselves of when it comes
to operationalizing punishment in extant sentencing frameworks. According
to the proportionality metric: the graver the offense, the longer the term of
imprisonment. Therefore, the length of a prison term is used as a meter for
retributive value. As such, these frameworks must be judged by their own terms.
Some positive law instruments at the national level, such as in many Western
countries, provide longer sentences for extraordinary international crimes than
for ordinary serious common crimes; this also was the case in certain jurisdictions
that punished atrocity committed in Europe or the Paci¬c Rim in the immediate
aftermath of World War II. On the other hand, this is not a universal practice “
far from it “ among positive law instruments in all national legal orders. What
is more, the practice of contemporary courts that punish offenders for extraordi-
nary international crimes, for example in the former Yugoslavia, reveals that, for
the most part, sentences for multiple international crimes range from as severe
to less severe than for a single serious common crime.25 As the states of the
former Yugoslavia develop specialized war crimes chambers to process atrocity
cases, average sentences might increase through the ordering of harsher mini-
mum sentences. Such an upward “ and, for the moment, largely conjectural “
trajectory in the severity of sentence, however, would do no more than place
such sentences in the same ballpark as sentences for serious ordinary crime.
The East Timor Special Panels appeared in their practice to be support-
ive of greater retribution for international crimes than for ordinary crimes. As a
hybrid tribunal with dual jurisdiction over ordinary and extraordinary crimes, the
Special Panels constitute an interesting case study. The data reported in Chapter
3 demonstrate that, with regard to mean sentences, the ratio between ordinary
crimes and international crimes was 1:1.58. Therefore, the mean sentence for
extraordinary international crimes was about 50 percent longer than for ordinary
common crimes. However, the data interpretation remains subject to a num-
ber of important caveats. First, the mean sentence for serious ordinary crimes
was 6.3 years. The median sentence was ¬ve years. These sentences are very
modest when compared to the treatment that serious ordinary crimes receive in
the domestic law of many states, thereby suggesting that the disparity between
sentences for ordinary and international crimes in the practice of the Special
Panels emanated in part from lightly punishing ordinary crimes. Second, three
156 Atrocity, Punishment, and International Law

sentences of 33 1/3 years issued by the Special Panels for crimes against human-
ity deviated considerably from the median sentence and, thereby, arti¬cially
boosted the mean. These sentences eventually were reduced by ex post Presi-
dential Decree to twenty-¬ve years. Third, over time the trend in East Timor
arced toward more lenient sentences, including extremely modest sentences in
the two- to seven-year range for crimes against humanity. These three caveats
cloud the apparent practice of the Special Panels to pursue greater retribution
for extraordinary international crimes.
One major impediment to the retributive aspirations of international crim-
inal law is that widespread crime cannot be re¬‚ected in punishment owing to
human rights standards, which cabin the parameters of sanction. In particu-
lar, these standards limit the amount of pain that institutions can in¬‚ict upon
detainees. The gravity of atrocity crimes can quickly become overwhelming “ so
much so that, from a retributive perspective, gravity becomes unintelligible and
immeasurable. How, then, to make punishment proportionate to the amplitude
of harm caused?26
Faced with the prospect of “trying” the former Romanian dictator Nicolae
Ceausescu and his wife, Elena, the prosecutor bitterly noted that his entire
professional code of ethics became upended. He famously remarked:
[A]s a lawyer, [I] would have liked to oppose the death sentence, because it is
inhuman. But we are not talking about people. I would not call for the death
sentence, but it would be incomprehensible for the Romanian people to have
to go on suffering this great misery and not to have it ended by sentencing the
two Ceausescus to death.27

The evil was simply so overwhelming that the prosecutor had no way to punish
the perpetrators other than resorting to a sanction that fell outside the values
he associated with the law. For the Romanian prosecutor, the wrongdoing “ to
borrow from Arendt “ simply exploded the limits of the law.
But the wrongdoing can explode the limits even of a legal process that
favors the death penalty as an ultimate retributive sanction. If retribution
truly were to re¬‚ect the gravity of extraordinary international criminality, death
might even fall short. As the Supreme Court of Israel frustratedly observed in
We know only too well how utterly inadequate the sentence of death is as
compared with the millions of unnatural deaths he decreed for his victims.
Even as there is no word in human speech to describe deeds such as the deeds
of [Eichmann], so there is no punishment under human law suf¬ciently grave
to match [his] guilt.28

In correspondence with her mentor Karl Jaspers, Arendt observed that, for
extraordinary international crimes, “no punishment is severe enough [ . . . ] this
guilt, in contrast to all criminal guilt, oversteps and shatters any and all legal
systems. [ . . . ] We are simply not equipped to deal . . . with a guilt that is beyond
Quest for Purpose 157

crime [ . . . ].”29 If the retributive value of punishing extraordinary international
criminals truly were to be engaged, perhaps punishment would have to exceed
anything ordinary.30 Truly proportionate sentences then might involve torture
or reciprocal group eliminationism. That is a terrifying path. In such a scenario,
survivors would become as depraved as their tormentors.
In sum: for those who commit the most egregious crimes of concern to the
international community as a whole, sanctions tend to range from less severe
to as severe as the punishments for ordinary murder in many countries.31 But
extraordinary international crimes are supposedly graver than serious ordinary
common crimes. The fact that punishment does not match this enhanced gravity
weakens retribution™s credibility as a penological goal for international crimi-
nal law.

(2) Treatment of High-Level Offenders. When it comes to punishing extraor-
dinary international criminals, although the retributive value of international
convictions is supposed to be greater than that of national convictions,32 the
sentences of the international criminal tribunals are not predictably lengthier
than those meted out in those territorial jurisdictions where atrocity is over-
lappingly prosecuted as extraordinary international crimes through national or
local institutions.33 Nor are the conditions of imprisonment at international
institutions more onerous; nor is the stigma of conviction weightier.
Let us begin with length of sentence. The overall evidence is inconclusive
regarding the existence of differences in terms of absolute length of sentence
between international and national institutions. Although IMT sentences were
harsher than those of many national civilian courts that prosecuted World War II
atrocity, they were not more severe than the sentences issued by certain national
military instrumentalities. The enhanced severity of international sanction is
even less apparent in the case of contemporary institutions, particularly when
factoring in that international tribunals assert jurisdiction over the most serious
offenders. Elemental retributive theory suggests that these offenders propor-
tionately deserve harsher punishment. Therefore, the fact that the evidence is
inconclusive with regard to whether contemporary international tribunals issue
harsher sentences than national or local institutions, which generally process
lower-level offenders, is of concern to the viability of retributive theory.
Evidence introduced in Chapter 4 suggests that national courts in the states
that emerged from the former Yugoslavia have punished extraordinary interna-
tional criminals less harshly than the ICTY. This, however, is not due to the
particularly lengthy nature of sentences issued by the ICTY. Rather, it is largely
due to the prevalence of ethnic bias and unprofessionalism in national courts
(e.g., in Croatia), which often results in modest sentences for low-level offenders.
Looking ahead, though, the situation in these national courts is set to change.
I predict an increasing alignment of the practice of these courts (in particular,
specialized war crimes chambers) with that of the ICTY as these courts receive
referrals from the ICTY and begin independently to prosecute higher-pro¬le
158 Atrocity, Punishment, and International Law

cases as a matter of course. This alignment would diminish sentencing differ-
ences between the two levels of judicialization.
The ICTY does not as a matter of course sentence its convicts to terms of
imprisonment that exceed what it determines to be available under domestic
law. As a benchmark, the ICTY often adopts twenty years as the maximum term
available in the sentencing practice of the former Yugoslavia (this seems to come
from the fact that, under the SFRY Criminal Code, the most serious offenses
that were eligible for the death penalty could be transformed by the court into
a twenty-year sentence).34 Many ICTY sentences dip below twenty years (and
the mean and median sentences are well below that ¬gure); on the other hand,
some sentences have exceeded that maximum and, in fact, the ICTY has as
a matter of law af¬rmed that it has the discretion to exceed this maximum.35
However, as set out in Chapter 4, current sentencing frameworks in the states
that emerged from the former Yugoslavia, although abolishing the death penalty,
permit maximum sentences in the forty- to forty-¬ve-year range depending on
the state. Although in some cases the ICTY acknowledges this maximum range,
it rarely issues such sentences (as of May 2006, it has only done so twice “ under 5
percent of its total sentences). In the end, it is not surprising that survey research
demonstrates that residents of af¬‚icted communities in the former Yugoslavia
view ICTY sentences as lenient.36 Moreover, ICTY defendants tend to resist
referral of their cases to national courts in the former Yugoslavia. The prospect
of referral apparently played a part in one ICTY defendant™s decision to plead
guilty.37 The fact that perpetrators demonstrate greater fear of punishment at
the hands of national authorities seems at odds with the supposedly enhanced
retributive value of punishment at the ICTY.
Because Rwandan domestic law still provides for the death penalty, the defen-
dants found guilty by the ICTR (mostly senior of¬cials) in theory receive sen-
tences lower than what they likely would receive under Rwandan law. Overall,
the Rwandan national courts issue death sentences to about 10 percent of all
defendants, although that percentage has progressively declined since 2002 and
no individuals have been executed since 1998. To be sure, Rwanda™s practice of
no longer enforcing death sentences suggests that, de facto, the death penalty
may no longer be a practical sentencing option in Rwanda. That said, as an
ICTR Trial Chamber recently noted in Prosecutor v. Muhimana, the death sen-
tence does remain on the books.38 This reveals a paradox: namely, leaders of the
genocide are formally punished less severely than lower-level offenders. This
paradox also is evident in Sierra Leone, which “retains the death penalty under
its domestic law [leading to a situation] where the worst offenders are eligible
for lower punishments because they are tried at the international tribunal.”39
The ICTR sentences slightly under half of its convicts to life terms. Because
it has acquitted three individuals so far, this means that just over 40 percent
of all ICTR defendants who have gone to trial receive life sentences. This is a
higher proportion of life sentences than that issued by the Rwandan national
courts, including the Specialized Chambers. But when the number of death sen-
tences in the national courts is added to the mix, the result is that approximately
Quest for Purpose 159

40 percent of defendants receive life or death. The proportion of life or death
sentences has been dropping annually in the Rwandan national system (there
is no evidence of such a trend regarding life sentences at the ICTR). With
regard to ¬xed terms of imprisonment, my research, set out in Chapter 4, iden-
ti¬ed a median term of 11 years and a mean term of 15.25 years in the Rwandan
courts; to be contrasted with the ICTR™s practice among ¬nalized sentences
of a mean ¬xed term of 20.9 years and a median ¬xed term of 15 years (among
un¬nalized sentences, the mean increases to 23.5 years and the median jumps to
25 years).
Given the ¬gures regarding ¬xed-term sentences, and the trends in the domes-
tic courts, the overall length of sentence is therefore higher at the ICTR than
in the domestic courts. However, the defendants prosecuted before each insti-
tution are not similarly situated. Account must be had that, when it comes to
high-status offenders, sentences in Rwandan courts are stiffer. For example, in
Prosecutor v. Semanza, the ICTR Appeals Chamber found that, although the
defendant™s sentence may have been more severe in Rwandan courts, “the Trial
Chamber acted within its discretion when it imposed a lesser sentence.”40 Under
Rwandan law, Semanza would have received at least life imprisonment (maybe
even a death sentence). In Prosecutor v. Bisengimana, a case from 2006, an ICTR
Trial Chamber recognized that the gacaca law subjects a person of the stature
of the defendant who pleads guilty to crimes against humanity to a sentence
between twenty-¬ve years to life.41 It then sentenced the defendant to ¬fteen
years™ imprisonment. In the case of gacaca, preliminary evidence from Rwanda
suggests a tremendously wide range of sentence. The maximum sentence under
gacaca for Category 2 offenders is thirty years™ imprisonment. This dips below
what the ICTR can issue. However, the ICTR would have virtually no interest
in prosecuting an individual whose culpability is alleged to be tantamount to
that of a Category 2 offender.
On a different note, as set out in Chapter 3, the ICTY permits convicts access
to conditional and early release. Eligibility for pardon or commutation of sen-
tence hinges upon the domestic criminal law of the state where the prisoner
serves sentence.42 ICTY convicts are imprisoned in Western European states
whose domestic law permits application for commutation or early release to be
made after two-thirds of the sentence has been served. This process disempow-
ers the af¬‚icted society by superimposing the ordinary common criminal law
of a faraway state (and judgments of state of¬cials in these faraway places) as
a template to attenuate punishment after the fact. Interestingly, however, the
ordinary domestic laws of some states emergent from the former Yugoslavia
have been revised to provide for early and conditional release, even in the case
of extraordinary international criminals. This reveals the emergence of some
consensus in these states in favor of permitting atrocity perpetrators access to
the same early release possibilities that are available for common criminals.
I posit that this domestic law reform is in part in¬‚uenced by the perceived
need to mimic modern international methodologies. Regardless, it is unclear
how the putatively enhanced retributive value of punishing extraordinary
160 Atrocity, Punishment, and International Law

international criminals is satis¬ed by permitting these criminals access to (and in
many cases granting) early release in the same manner as for ordinary common
Early or conditional release has not yet begun at the ICTR. But it soon might,
given that the ICTR has concluded agreements to house prisoners in Western
European countries. If ICTR convicts begin to serve sentence in these countries
and then become entitled to early release provisions that neither Rwanda nor
Mali (where ICTR convicts currently serve sentence) contemplate, the retribu-
tive gap between the treatment accorded those most responsible for genocide
and those less responsible will grow. In the case of the East Timor Special Panels,
the retributive value of punishment, already threatened by conditional release,
is additionally undercut by the exercise of Presidential Decrees that reduce the
duration of imprisonment. The application of these Decrees promotes an added
layer of selectivity and unequal treatment among convicts.
When it comes to assessing whether the retributive value of international pro-
ceedings exceeds or falls short of that of national proceedings, a truly purposive
comparative analysis must transcend strict quantitative measurement of length
of imprisonment. In this regard, I consider two other aspects of the retributive
value of punishment: (1) conditions of imprisonment and (2) stigma.
Qualitatively speaking, conditions of incarceration arranged at the interna-
tional level tend to be much less harsh than those available to defendants sen-
tenced nationally. When compared to the domestic Rwandan prisons, the ICTR
detention unit is luxurious. Although prisons in Mali, where most ICTR defen-
dants serve sentence, are not as comfortable as the ICTR detention unit, they
are superior to options available in Rwanda.
Moreover, defendants awaiting trial at the ICTR detention unit receive a
quality of health care that exceeds that accorded to defendants in national tri-
als or gacaca and, more starkly, victims living in Rwanda. ICTR defendants
have access to treatment, medication, and services that few victims can claim.43
These disparities are galling given the prevalence of HIV/AIDS in Rwanda gen-
erally and in particular among genocide survivors.44 Victims™ groups in Rwanda
have made antiretrovirals available to some members of the public; nonetheless,
affordable and accessible medical treatment is scarce. Prosecuting and punish-
ing perpetrators is supposed to voice retribution. However, in the case of ICTR
defendants, the fact they are accused of extraordinary international crimes iron-
ically may keep them alive and healthy to enjoy a quality of life that exceeds that
of victims and probably exceeds that which they would experience were they
not to be “punished” at all.
A similar concern arises regarding the retributive value of the pain and pun-
ishment in¬‚icted by the ICTY. In the recent plea-bargained sentence of Biljana
Plavˇi´ (a Bosnian Serb leader known as the Serbian Iron Lady), “victims reacted
with predictable outrage” at the fact that “Plavˇi´ was sent to serve her term in
a posh Swedish prison that reportedly provides prisoners with use of a sauna,
solarium, massage room, and horse-riding paddock, among other amenities.”45
While in jail, she was even “presented a birthday cake on her birthday.”46
Quest for Purpose 161

Similarly, research on popular attitudes toward punishment in East Timor
reveals profound externalization of justice concerns as well as a deep retributive
shortfall. Legal scholar Nancy Amoury Combs concludes: “[I]n the eyes of many
East Timorese, detention constitutes precious little punishment since prisoners
are fed and housed in jail, and in some cases can avoid their compensation
If severity of sanction is construed to include level of stigmatization, then there
is cause to believe that international sanctions are more denunciatory.48 Inter-
national proceedings reach a worldwide audience. They are broadcast on televi-
sion screens everywhere. The stigma is spread widely. Paradoxically, though, the
broadcasting often is more accessible outside of the af¬‚icted locality than within
it. But it is important not to underestimate the stigmatizing value of national
proceedings.49 Sometimes, the shaming value of sentencing is more acute when
carried out by a community of one™s immediate peers.
And, ¬nally, other differences persist between international and national
modalities of punishment that, in certain cases, render the retributive value
of punishment at the international level less onerous than at the national level.
Rwanda once again presents an example. In Rwanda, the national courts through
the partie civile process routinely award very substantial (although largely uncol-
lected) ¬nancial damages to victims and survivors; also, the accused, if found
guilty, will be ordered to pay the costs of the proceedings. Both of these additional
remedies represent an interesting diversi¬cation of the accountability paradigm
that adds to the retributive weight of punishment. Although the ICTR Statute
contemplates some restitutionary and reparative possibilities, these have not
been pursued in practice. My review of the Rwandan national jurisprudence
notes cases where, in order to execute these orders, the convicted person™s assets
are auctioned off. The prospect of a lifetime of work or community service to
pay off these civil damages can be of a high punitive force and, hence, can go
far in pursuing retributive goals.

(3) Variability in Sentencing. At the international level, trial judges have
unfettered discretion to af¬x the period of imprisonment for convicted extraordi-
nary international criminals. Moreover, appellate judges, who also bene¬t from
broad discretion in sentencing, actively intervene as well. Assuredly, sentencers
at the international level are assisted by a typology of aggravating and mitigating
factors. But the predictability or clarity this typology provides is limited. Parties
do not have much of a sense of what evidence to present in sentencing hearings
(if there even is a separate hearing) and which aspects to emphasize. There
is considerable inconsistency “ both cardinally and ordinally “ in terms of the
sentences issued. Although there is some indication that the sentencing jurispru-
dence of international criminal tribunals is deepening in depth and rigor, it still
remains confusing, unpredictable, and without the ordering bene¬ts of a viable
heuristic. There is also, dating back to Nuremberg, confusion with regard to the
mixing of factors implicating liability with factors to consider in aggravation of
162 Atrocity, Punishment, and International Law

At the national level, there is much greater diversity in terms of the discretion
accorded sentencers in determining the period of imprisonment. Some national
frameworks tie the hands of sentencers. Some, such as Rwanda™s, set parameters
within which some discretion is retained. Others are very permissive.
Discretion in sentencing carries with it certain advantages, such as ¬‚exibility
and the opportunity to individualize punishment. Discretion, however, poses
challenges to the attainment of the claimed retributive purpose of punishment.
It can lead to a lack of consistency in sentencing that, in turn, could cloud the
public™s ability to assess the gravity or seriousness of crime. This obfuscation
might diminish public respect for the legal system.
As outlined in Chapter 3, at the international level considerable variability
persists within institutions in terms of the length of sentences meted out to simi-
larly situated defendants. Moreover, sentences vary considerably not only within
but also among the various international tribunals. For example, the sentences
of the East Timor Special Panels for extraordinary international crimes are the
most lenient. Does this mean that atrocity in East Timor is of the least gravity?
The most proximate comparison is between the two ad hoc tribunals. ICTR
sentences are longer than ICTY sentences; in addition, the ICTY welcomes
early release, which has not (yet) been operationalized at the ICTR. In order
for the retributive justi¬cation to explain this disparity, it might assume that the
gravity of the Rwandan violence exceeds that of the former Yugoslavia.
Intuitively, making comparative assessments of the gravity of systemic vio-
lence does not seem patently unreasonable. That said, these are dif¬cult com-
parisons to make and can degenerate into hairsplitting. Punishing institutions
have not yet articulated any framework of comparative assessment that deter-
mines the yardsticks by which to measure the greater gravity of, for example,
Rwanda™s tragedy over Bosnia™s, or Sierra Leone™s over East Timor™s. No punish-
ing institution has justi¬ed the enhanced or diminished length of its sentences
on the basis of the more repugnant nature of one nation™s atrocity over another™s.
Two other rationalizations could coherently explain why ICTR sentences are
longer than ICTY and East Timor Special Panel sentences. The ¬rst involves
the incorporation of the norms of the af¬‚icted community. Domestic sentences,
in particular those maximum sentences that would be imposed on high-level or
notorious convicts, are more punitive in Rwanda (death, life imprisonment) than
in the former Yugoslavia (long-term ¬xed imprisonment).50 The second ratio-
nalization posits that sentences by the ICTR appropriately are harsher because
the ICTR has convicted much more frequently for genocide, which has been
described as the “crime of all crimes,”51 and, therefore, for those who accept
this description, it follows that perpetrators of genocide simply deserve harsher
In principle, I would welcome a policy whereby international institutions
sentenced differently based on incorporation of national norms. Such a pol-
icy militates against a democratic de¬cit. This policy is particularly desirable
when national positive law instruments, or court activity, represent what popu-
lations on the ground envision as legitimate sentencing practice. In my opinion,
Quest for Purpose 163

accommodating representative national sentencing practices is intimately con-
nected to the meaningfulness of sanction. This accommodation, however, would
not necessarily be grounded in the retributive value of punishment, but, instead,
in other penological justi¬cations and, even, broader justi¬cations such as demo-
cratic legitimacy.
That said, the international criminal tribunals have not explicitly recognized
differences in national sentencing practices to justify the longer sentences at
the ICTR. In fact, the incorporation of national sentencing practices in the
decisionmaking of international punishing institutions remains unpredictable.
The ad hoc tribunals refuse to view national practices as in any way binding. In
fact, an ICTY Trial Chamber recently held that national sentencing practices
are “purely indicative.”52 The ICTR has held it has no obligation to conform to
general practice regarding prison sentences at the national level, although it is to
refer to this practice.53 The East Timor Special Panels took a similar approach to
the incorporation of domestic law.54 The Special Panels sporadically integrated
adat, a traditional notion regarding taking responsibility and paying respects, into
the sentencing framework. However, they never predictably stated when adat
ought to be referenced. The ICC does not appear to be under any obligation to
recognize national or customary practices. In sum: although the incorporation of
national and local law into international sentencing practices is a salutary goal,
the extant process of incorporation seems to undermine consistency without
providing the real legitimizing bene¬ts of local and contextual involvement.
With regard to the second rationalization, it is true that many of the convic-
tions at the ICTR have been for genocide, whereas nearly all ICTY convictions
have been for crimes against humanity and war crimes. Looking more carefully,
however, the ICTY™s actual genocide convictions do not incur as stiff a penalty
as the ICTR™s. An ICTY Trial Chamber sentenced Blagojevi´ to eighteen years
(on a count of complicity to commit genocide) and the Appeals Chamber
sentenced Krsti´ to thirty-¬ve years (for aiding and abetting genocide).56 The
ICTY™s harshest sentences have been for crimes against humanity. Factually,
many perpetrators in Rwanda were convicted as primary perpetrators of geno-
cide, not as aiders and abettors, and this more serious level of responsibility does
problematize the comparison somewhat. Fundamentally, though, the interna-
tional tribunals have not consistently stated that, ceteris paribus, genocide merits
a more severe sanction because its inherent gravity exceeds that of crimes against
humanity or war crimes.57

c. Plea Bargaining
Plea bargains present another challenge to the retributive value of punish-
ing extraordinary international criminals. This challenge operates at both the
national and international levels.
Plea bargains can take several forms. The confession and pure guilty plea con-
stitutes one form. Here, an offender unilaterally confesses, admits guilt, and fore-
goes the trial. The matter proceeds directly to sentencing.58 Other forms involve
164 Atrocity, Punishment, and International Law

bilateral negotiation. The plea agreement, for example, arises from negotiations
between prosecution and defendant. This quasicontractual process, which has
gained traction at the international criminal tribunals, incorporates pragmatic
elements that are commonplace to U.S. domestic criminal law. For example,
prosecutor and defendant can conclude a sentence bargain, where they agree
upon underlying facts and negotiate regarding a sentencing range. This range is
then presented to the international tribunal for approval together with the guilty
plea. Another form is the charge bargain. In charge bargaining, the prosecutor
may drop certain charges (including serious charges) as part of the plea nego-
tiations. In some cases, the defendant may only plead guilty to one omnibus
charge that generalizes the speci¬cs of the factual record (and may exclude
facts regarding certain dropped charges and details regarding others). Charge
bargaining has occurred at the international tribunals, although less frequently
than sentence bargaining.
Plea bargains have many bene¬ts. The confession and guilty plea often
re¬‚ects atonement, apology, and recognition of responsibility on the part of
the offender who self-convicts. In a case involving a controversial charge bar-
gain, ICTR judges noted the offender™s genuine desire to tell the whole truth,
ask for pardon, and publicly express remorse.59 Plea bargains are cited for their
ability to promote reconciliation.60 Moreover, plea bargains promote manage-
rial ef¬ciency. Accordingly, they appeal to prosecutors facing tight completion
strategies, funding shortfalls, or political pressures, each of which has dogged
international institutions. Plea bargains ensure that some justice is seen to be
done. They also can facilitate the acquisition of evidence that implicates other
That said, negotiated plea bargains compete with the notion that perpetrators
deserve to be punished. In this quasicontractual exchange, punishment becomes
disconnected from desert or gravity and contingent on what the convict knows,
who else the convict is willing to implicate, and the vulnerability of the pun-
ishing institution. Perpetrators having information on others will likely be given
a better bargain than those with nothing to offer. A perpetrator involved in a
joint criminal enterprise with high-level accused could bene¬t the most from
the liability or sentencing discount regardless of the egregiousness of the crimes,
the perpetrator™s ability to encourage recidivism among others, or the expressive
value of stigmatizing that perpetrator through public denunciation.
Assuredly, these disparities also are found in domestic criminal law, in partic-
ular regarding the sentencing of drug offenders and criminal syndicates (where
they have prompted a broad array of critical commentary). Concerns, how-
ever, are even more pronounced when a plea-bargaining model designed to
process ordinary crimes in select adversarial domestic criminal law systems is
extended to the context of extraordinary international crimes. An institutional
policy that differentially punishes extraordinary international criminals based
not on the gravity of their offenses but, rather, on judicial economy, strategic sys-
tem interests, and bureaucratic contingencies splinters the deontological basis
of retribution.
Quest for Purpose 165

The ICTY has noted that another bene¬t to plea bargains is that they protect
victims from having to testify.62 Although experience indicates that testifying is
traumatic for many victims, this is not the case for every victim. In fact, for some
victims, testifying might have signi¬cant cathartic value. Accordingly, if one of
the purposes of retribution is for individual victims to see punishment in¬‚icted
on the criminal, victims should play a role in determining whether or not a plea
should be accepted and, if so, on what terms.
Plea bargains of diverse forms have occurred at each of the ICTY, ICTR,
and East Timor Special Panels. Prosecutors from each institution have pushed
plea bargains. Defendants at each institution have responded differently. ICTY
defendants have been receptive to plea agreements owing to an appetite for
sentencing concessions.63 Admittedly, the actual mitigating discount that plea
bargaining will obtain is dif¬cult to measure and remains unpredictable. As of
November 2004, the mean and median sentences of those ICTY defendants
convicted following a plea bargain were 2.6 and 6 years shorter than the mean
and median sentences of those convicted following trial.64
The willingness of ICTY defendants to plead guilty recently has abated owing
to decisions by the Trial Chambers to impose sentences that exceed the range
that the ICTY Prosecutor and defendant had agreed upon.65 In these cases, the
Trial Chambers were motivated by the perceived excessive lenity of punishment
in these agreements.66 The ICTY positive law instruments stipulate that the Trial
Chambers are not formally bound by a sentence recommendation contained
in a plea agreement.
The Momir Nikoli´ case presents an example. Nikoli´ was a security and
intelligence of¬cer at Srebrenica, where he played a command role. He was the
¬rst Serb of¬cer to admit to participating in the Srebrenica massacre. The Trial
Chamber sentenced him to twenty-seven years.67 This exceeded the recommen-
dation of the plea agreement (where the ICTY Prosecutor had agreed to recom-
mend a ¬fteen- to twenty-year sentence and the defense a ten-year sentence).
Nikoli´ had pled guilty to one count of persecutions as a crime against human-
ity. The Trial Chamber had expressed a number of reservations with regard to
plea bargains for cases of extraordinary international criminality, and did not
follow the plea recommendation, although it still found the guilty plea to be
signi¬cant and to constitute an important factor in mitigation. Nikoli´ appealed.
The Appeals Chamber did not quarrel with the Trial Chamber™s approach to
the guilty plea. However, for other reasons it reduced the sentence to twenty
years.68 This meant that the sentence fell within the range the Prosecutor had
Although there have been other cases in which ICTY judges have voiced some
reserve regarding the general suitability of plea bargaining for serious crimes of
concern to the international community, the Trial and Appeals Chambers have
accepted many of the plea agreements that have come before them, at times in
spite of articulated concerns.
Sentences issued by the ICTY following plea bargains have little retribu-
tive rhyme or reason. On the one hand, Biljana Plavˇi´ , who was responsible
166 Atrocity, Punishment, and International Law

for planning some of the gravest atrocities in Bosnia (forced expulsion of hun-
dreds of thousands of nonSerbs, destruction of 850 nonSerb villages, killings
of many thousands of individuals, widespread sexual assault, and inhumane
destruction),69 was sentenced to eleven years. (This fell below the Prosecutor™s
recommended range of ¬fteen to twenty-¬ve years). On the other hand, rebel
Croatian Serb leader Milan Babi´ , further down on the leadership hierarchy “
and who, unlike Plavˇi´ , testi¬ed against Slobodan Miloˇevi´ “ received a sen-
sc sc
tence of thirteen years for his role in a campaign to expel non-Serbs.70 Whereas
Plavˇi´ was given signi¬cant discount for her post hoc efforts at peace negotia-
tion, Babi´ was not given such discount despite the ICTY™s recognition of the
salience of his efforts.
To be sure, Plavˇi´ was in her seventies at the time of sentencing and the ICTY
took account of her age as a mitigating factor. That said, Plavˇi´ ™s sentence is
only four years longer than that imposed on Miodrag Joki´ , who pled guilty to a
number of war crime charges related to the shelling of Dubrovnik (the charges
against Joki´ related to the destruction of cultural property and the deaths of two
civilians and the wounding of three others).71 In 2005, an ICTY Trial Chamber
sentenced Lieutenant-General Strugar, who was initially charged together with
Joki´ with regard to the 1991 Dubrovnik attacks, to eight years™ imprisonment
for attacks on civilians and destruction/willful damage to heritage/charitable
institutions.72 Strugar is roughly the same age as Plavˇi´ and was sentenced to
only three years less imprisonment than her despite the yawning gap between
the two in terms of the gravity of their impugned conduct.73
Whereas Darko Mrdja was sentenced to 17 years for pleading guilty to direct
involvement in the shooting of 200 persons (only 12 of whom survived),74 Ranko
ˇ sc
Ceˇi´ , a Bosnian Serb police reservist, was sentenced to 18 years for pleading
ˇ sc
guilty to beating to death 10 prisoners and sexually assaulting 2 others.75 Ceˇi´ was
a de facto subordinate to Goran Jelisi´ , who had previously pled guilty to thirteen
ˇ sc
murders at the same camp where Ceˇi´ had committed his crimes. Jelisi´ , c
however, had received a sentence of forty years: by any measure, an incredible
ˇ sc
disparity in sentencing.76 The Ceˇi´ sentence also should be juxtaposed against
the ten-year sentence meted out to Miroslav Deronji´ , an in¬‚uential civilian
leader who substantially participated in a joint criminal enterprise that ordered
the razing of the village of Glogova, in which sixty-four Bosnian Muslim civilians
were killed and many more forcibly displaced.77 In Deronji´ ™s case, the Trial
Chamber ruminated about the suitability of plea bargains for situations of mass
atrocity, but ultimately af¬rmed the plea bargain as well as the light sentence
recommended by the Prosecutor.78 On appeal, the sentence was upheld.79
While there are inherent dif¬culties in comparing the speci¬c situation of
individual defendants, such comparisons are possible, valuable, and “ above
all “ necessary if the sentencing regime is to have predictability or consistency.
These comparisons suggest that, by injecting considerable indeterminacy into
the allocation of punishment, plea bargains undermine its retributive value.80
To be sure, nothing is per se improper about basing punishment on manage-
rial economy, pragmatics, incentivizing rewards, and prosecutorial strategizing.
Quest for Purpose 167

When taken together, these factors can constitute a compelling basis on which
to justify the allocation of punishment. However, this basis never has been
articulated as an explicit goal or objective of punishment by the international
tribunals. To this end, basing punishment thereupon only can be assessed in
relation to how it plays with the avowed goals of sentencing, in this case retri-
bution. And it does not seem to play well. Plea bargaining may mix more easily
with reconciliatory, restorative, or reintegrative aspirations of punishment but,
insofar as these are only distantly conceptualized or operationalized by the ad
hoc tribunals, this relationship is more one of conjecture than actual practice.
Perhaps responding to the ICTY™s lead, national courts in the states emerging
from the former Yugoslavia also have implemented plea bargaining (for exam-
ple, Bosnia and Herzegovina™s criminal procedure saw the introduction of plea
bargaining in 2003). Many judges in domestic courts award huge discounts for
guilty pleas in the name of administrative economy. These discounts make it
dif¬cult for punishment to retain retributive value.
Many (at least half) of the extraordinary international criminals prosecuted
by the East Timor Special Panels pled guilty.81 Resultant plea agreements often
were encouragingly af¬rmed by the Special Panels, who extolled how plea agree-
ments aided in the administration of justice.82 In terms of quanti¬cation of
sentence, those who pled guilty in East Timor received a signi¬cant discount.
Overall, the Special Panels had shown a “markedly lenient approach” to those
who pled guilty, cutting around half of the sentence that would otherwise have
been imposed.83
Many East Timorese defendants pled guilty, however, not because of sen-
tencing concessions (many may not even have understood the plea-bargaining
process) but, rather, because of a “cultural[] commit[ment] to a world view that
places tremendous value on confession, apology, and reconciliation.”84 This sug-
gests that the motivation for guilty pleas as gleaned from domestic common law
legal systems, namely the desire for a reduced sentence among ordinary common
criminals, does not ipso facto apply to all contexts of extraordinary international
criminality. The incorporation of plea bargaining from its entrenched status in
these systems to the very different context of the punishment of international
crimes represents yet another example of borrowing from the familiar to ground
the extraordinary. However, “the wholesale transplant of plea bargaining prac-
tices that successfully procure guilty pleas in the context of domestic crimes is
likely to prove inef¬cient and ineffective in the context of many international
crimes.”85 The East Timorese situation thereby demonstrates the value of an
accountability paradigm that is implemented through differentiated kinds of
procedures keyed to the sociolegal particularities of the af¬‚icted society, instead
of a simple transplant.
The ICTR also has incorporated plea agreements, including charge
bargaining.86 For example, Paul Bisengimana, a former mayor, was sentenced
by an ICTR Trial Chamber in April 2006 to ¬fteen years™ imprisonment for his
role in the murder of one thousand Tutsi who had sought refuge in a church.87
He had pled guilty to two charges of murder and extermination. In return, the
168 Atrocity, Punishment, and International Law

Prosecutor dropped eight other counts, including genocide, complicity in geno-
cide, and rape. Overall, though, plea bargaining has not been particularly preva-
lent at the ICTR. This is so for a number of reasons. One reason is that the ICTR™s
¬rst guilty plea, which involved Jean Kambanda (the Rwandan Prime Minister
during the genocide), led to a life sentence. This sentence was imposed over
Kambanda™s objections and, arguably, his understanding during the plea discus-
sions. Kambanda certainly did not intend a life sentence to issue from the guilty
plea and, unsurprisingly, this outcome placed the plea-bargaining process into
a certain level of disrepute. As an aside, it is noteworthy that Kambanda was
given absolutely no sentencing discount for the reconciliatory and truth-telling
value of his timely guilty plea, unlike the ICTY™s treatment of Plavˇi´ ™s plea.88
This constitutes yet another example of the evident “ and poorly justi¬ed “
sentencing disparities between the ICTY and ICTR.
Another reason why there has been infrequent plea bargaining at the ICTR
is because, according to Combs, many ICTR defendants are dissuaded from
pleading guilty by ideological factors.89 This behavior also would appear to erode
international criminal law™s deterrent capacity as well, insofar as this capacity
is predicated on the assumption that individuals shall modify their behavior in
light of the threat of punishment. The fact that defendants are insuf¬ciently
moved by the prospect of reduced punishment to alter their behavior ex post
belies the possibility such defendants would be willing to alter their behavior ex
Guilty pleas are more frequent in the domestic Rwandan legal system than
at the ICTR. This is the case despite the fact that the requirements of pleading
guilty are more onerous within Rwanda than they are at the ICTR. What might
account for the greater frequency? One possible explanatory factor is that, within
the Rwandan system, the sentencing discounts for those who confess and plead
guilty are somewhat predictable (at least in terms of ranges). Another possible
explanation is that nearly all defendants at the ICTR are high-level accused
who, by virtue of their status, are less inclined to plead guilty owing to a deeper
ideological commitment. A third is that Rwandan defendants “ particularly
lower-level accused “ are more willing to plead guilty within the neotraditional
gacaca process than within an adversarial criminal trial.
In Rwanda “ at the levels of the conventional courts as well as gacaca “ the
confession and guilty plea process is not as much of a contractual exchange
as it is at the ad hoc tribunals. The in situ process (albeit inescapably affected
by downward pressures from the international) involves greater adherence to
standards of truthfulness and apology, although it is very far from a guarantor
thereof. Plea negotiations do not appear to be the norm. Rather, there is a
tendency toward pleading guilty to the charges as stated. Partial or dubious
guilty pleas are rejected, although judges can (and do) count such pleas outside
of the statutory scheme in mitigation of sentence. In the case of gacaca, the
General Assembly (namely, the population of the af¬‚icted community) is able
to reject confessions considered to be incomplete or insincere. This provides an
important democratizing element that contrasts with the exclusion of af¬‚icted
Quest for Purpose 169

populations from plea-bargain negotiations undertaken at the ad hoc tribunals.
If the plea is accepted, the offender becomes subject to a somewhat standardized
regime of discount in which there is discretion within established ranges (this
discretion is contoured by some determined mitigating and aggravating factors).
Overall, I believe that the domestic Rwandan legal system (both the national
courts and gacaca) demonstrates greater predictability, rhyme, and reason in
handling plea discounts than is the case with the ad hoc tribunals.

(ii) deterrence
Deterrence theory justi¬es punishment not because it is deserved, but rather
because punishment consequentially builds a safer world.90 Insofar as deter-
rence assumes that individuals will be dissuaded from offending (or reoffending)
because they fear getting punished, it posits that law is capable of ful¬lling a social
engineering function. Deterrence can be speci¬c to individual offenders or gen-
eral to the community of potential offenders. There is some scattered reference
to the merits of speci¬c deterrence in the jurisprudence of institutions that pun-
ish extraordinary international criminals.91 However, the focus overwhelmingly
is on general deterrence, namely the notion that if one person is punished, this
will reduce the likelihood that another person in that same place or somewhere
else will offend in the future.92 As an ICTR Trial Chamber intoned, punishment
“dissuade[s] for ever[] others who may be tempted in the future to perpetrate
such atrocities [ . . . ].”93 The UN Secretary-General has explicitly endorsed the
value of the international criminal tribunals in “deter[ring] further horrors.”94
Can criminal law deter atrocity? Although there are scattered anecdotal
reports that suggest that potential extraordinary international criminals are
deterred by the punishment of others following criminal trials,95 no system-
atized or conclusive evidence has been proffered.96 In any event, any anecdotal
research must absorb the reality that at times atrocity has continued to occur in
places following the creation of criminal tribunals to punish perpetrators. The
ICTY stands out as an example. It was created in 1993. However, some of the
gravest atrocities in the former Yugoslavia, including the Srebrenica massacre
(1995) and Kosovo ethnic cleansing (1998), occurred while the ICTY was in
full operation. Assuredly, it is somewhat facile to conclude that deterrence may
not be actualized just because atrocity continues after the establishment of a
punishing institution. After all, we can never know how much worse atrocity
might have been if no institution ever had been created. That said, all things
considered, just because we may have some cause to think that some deterrence
has been achieved does not mean that the extant paradigm effectively deters.
Other approaches to sanctioning universally repugnant crimes might be more
adept in attaining deterrent aspirations.
One reality that deterrence theory must contend with is the very low chance
that offenders ever are accused or, if accused, that they ever are taken into
the custody of criminal justice institutions. Selectivity is especially corrosive to
the deterrent value of prosecution and punishment. Criminologists long have
170 Atrocity, Punishment, and International Law

posited that it is the chance of getting caught and the promptness of punish-
ment, and not the severity of punishment, that affects behavior.97 International
tribunals are particularly vexed by the dif¬culties they experience in capturing
indictees. Insofar as international tribunals lack their own police force or agents
of enforcement, they can become dependent on the cooperation of the same
national authorities whose jurisdiction they may have ousted. In its early years,
the ICTY was stymied by the dif¬culty it experienced in capturing indictees.
ICTY of¬cials tenaciously persevered, however, and, as of December 2005, only
6 out of a total of 161 indictees remain at large (although, for the moment, this
group includes high-pro¬le suspects such as Mladi´ and Karadzi´ ).98 Eighteen
c c
of the ICTR indictees remain at large (the ICTR has arrested seventy-two indi-
viduals). Before closing up shop, the East Timor Special Panels were able to
prosecute only 87 of 370 indicted individuals. Many indictees roam around free
in Indonesia.
Moreover, being brought into custody to face trial is one thing; actually
being convicted is another. International criminal law™s focus on individual cul-
pability provable beyond a reasonable doubt “ a hallmark of liberal legalism “
sharply reduces the number of people who can plausibly be brought into the
dock because there always is a risk that insuf¬ciently compelling evidence will
lead to an acquittal. This risk is cited as one of the reasons in favor of introduc-
ing vicarious liability theories into international criminal law, such as JCE, to
which the ¬eld exhibits considerable skittishness. Although JCE may promote
deterrence by increasing the number of potential convicts, any such increase
is a minor one at the margins. There are only a small number of defendants
for whom JCE has played a material difference in terms of the prospect of
In some cases, national institutions are more successful in obtaining custody
over accused offenders. In Rwanda, well over one hundred thousand accused
have been taken into custody. However, in other contexts few (and some-
times no) suspects are indicted or taken into custody by national authorities.
National institutions often are crimped in the exercise of criminal punishment
by amnesties that, in certain cases, may be implemented for eminently laudable
goals of political transition or peace. In other cases, national authorities simply
elect to forget the past.
In sum, the chances of getting caught for committing egregious violations
of human rights “ certainly for heads of state and superior of¬cers “ are higher
today than they were prior to the establishment of institutions at the international
level. That said, notwithstanding the fact that the prospect of getting caught is
greater than it once was, it still remains tiny.
At this juncture, an interlocutor committed to deterrence theory might
respond: if the problem is limited to a lack of institutions, constabulary, and
¬nances, that problem is easy to rectify. Just create more institutions! Provide
more money! And, thereby, increase the likelihood of getting caught. Accord-
ingly, so goes the argument, shortcomings with deterrence are not intrinsic to
the theory. Instead, they derive from the functionally inadequate way in which
Quest for Purpose 171

the theory currently is implemented: the deterrence objective is attainable, but
remains underachieved by virtue of administrative limitations.
At ¬rst blush, it seems plausible that creating new institutions might go some
way to augment deterrence. However, I remain unconvinced that, fundamen-
tally, the existence of more liberal legalist punishing institutions would effec-
tively deter committed extraordinary international criminals. This is because
deterrence™s assumption of a certain degree of perpetrator rationality, which is
grounded in liberalism™s treatment of the ordinary common criminal, seems
particularly ill ¬tting for those who perpetrate atrocity. This assumption already
is hotly debated within the context of isolated common crime. However, its
viability is even more problematic in the context of the chaos of massive vio-
lence, incendiary propaganda, and upended social order that contours atrocity.
Do genocidal fanatics, industrialized into well-oiled machineries of death, make
cost-bene¬t analyses prior to beginning work? In the speci¬c case of terrorism,
will a suicide bomber be deterred by fear of punishment in the event of capture?
Although certain people may be deterred from killing or raping in pursuit of
eliminationist goals by a fear of imminent retaliation (i.e., an enemy army com-
ing around the corner), there is little to suggest that the threat of punishment
by a distant international court would deter. I am not alone in my skepticism.99
M´ gret opines that “[i]t beggars belief to suggest that the average crazed nation-
alist puri¬er or abused child soldier . . . will be deterred by the prospect of facing
trial.”100 He adds that this assumption is “a typical case of liberalism™s hegemo-
nious tendency of constructing the other in its own self-image, preferably along
the lines of some reductionist form of economic rational choice theory.”101
Let us examine two painful realities that jeopardize the assumption of per-
petrator rationality amid cataclysmic events. These are: ¬rst, grati¬cation; and,
second, survival.
First, many perpetrators want to belong to violent groups.102 They ¬nd comfort
and solidarity in these groups. For many participants, violence has meaning and
is compelling. Although certain group organizers may be coldly motivated by
bureaucratic ambitions (such as Adolf Eichmann™s goal of advancing his career)
that might be deterred by the threat of eventual punishment or demotion, many
individuals organized as foot soldiers of evil share an affective motivation for
discriminatory killing. They are captured by angry social norms or, at least, are
captivated by them. As Jaime Malamud-Goti observes, many participants believe
that they are acting for the bene¬t of the collective, not their own personal
gain.103 It is simply not evident that the risk of punishment will deter people
from engaging in violent behavior that they, at the time, believe is morally
justi¬able and perhaps even necessary “ if not downright gratifying.104
Even assuming arguendo that rational choice were possible in the cataclysm
of mass violence, for some people the value of killing or dying for a cause exceeds
the value of living peacefully without the prospect of punishment. Participants
often are motivated by immediate approval from their peers. Cravings for such
approval easily can outweigh the dissuasive effect of distant, and often hypo-
thetical, punishment by an alien international criminal tribunal. Why incur
172 Atrocity, Punishment, and International Law

immediate ostracism in situations where, as perpetrators themselves note,
one person™s insubordination would have made no difference anyway? Alette
Smeulers reports:

Many perpetrators [ . . . ] convince themselves that they do not really have any
control and that it would not have made a difference if they had stood up and
refused to carry out the order. Stangl, commander of Treblinka, said: If I had
sacri¬ced myself, if I had made public what I felt and had died . . . it would
have made no difference. Not an iota. It would all have gone on just the same,
as if it and I had never happened.105

Second, amid the social disintegration and group-based reconstitution that
usually precedes mass violence, individuals often end up joining a marauding
group because to do so is the only viable survival strategy. Anthropologists have
documented such motivations in a variety of contexts, including among child
soldiers in Sierra Leone.106 After all, if one is not part of the group, one is
alone. Being alone makes it all the easier to become victimized or perceived as
belonging with or sympathetic to the “other.” Fears of aloneness are particularly
pronounced among many militia recruits “ orphaned children, adolescents, and
young men without families: in many cases poor and without occupational skills.
Even those individuals for whom violence is not gratifying may willingly join,
insofar as participating in massacre can guarantee survival to the next morning.
There is something luxurious, if not utopian, in the notion that individuals in
such desperate circumstances are amenable to being deterred by the prospect
of some distant international or domestic institution that might punish them
several years after their side might lose the con¬‚ict they currently are embroiled
in. This requires a heavy burden of proof on the part of deterrence theorists. This
burden has not been satisfactorily discharged. Although individuals who join a
marauding group for petty material gain might be deterred by the criminal law,
the same cannot be said for those who join to survive. And those who join for
survival purposes become much more committed and rigorous in their killing
than those who join merely to acquire incidental material trinkets.
Accordingly, criminal trials face signi¬cant obstacles in achieving their goal
of deterring killers. Criminal trials face even greater dif¬culty in reaching bene-
¬ting bystanders, another key group identi¬ed in Chapter 2. Essentially, liberal
criminal law leaves the masses unaccountable: its narrow focus persists despite
the fact that support and acquiescence of the masses is the singular prerequi-
site for atrocity truly to become epidemic. Violence becomes normalized when
neighbors avert their gaze, draw the blinds, and excitedly move into a suddenly
available apartment. This broad public participation, despite its catalytic role, is
overlooked by criminal law, thereby perpetuating a myth and a deception. The
myth is that a handful of people are responsible for endemic levels of violence.
The deception, which inures to the bene¬t of powerful states and organizations,
involves hiding the myriad political, economic, historical, and colonial factors
that create conditions precedent for violence.107
Quest for Purpose 173

Because the silence of the majority, the acquiescence of the bystander, the
enrichment of neighbors, and the nonfeasance of international organizations
never is implicated by a system based on criminalization, any such system does
little to deter these essential prerequisites to mass violence. Although the trial rep-
resents closure, this closure may be chimeric; and, more ominously, prematurely
might divert attention from more expansive reconstruction efforts or dull our sen-
sibilities regarding the inadequacies of criminal trials in unearthing many of the
root causes of systemic violence. On the other hand, a broader-based approach
that contemplates diverse, including collectively based, sanctions might reduce
the appeal of passively acquiescing and, thereby, turn some erstwhile bystanders


. 6
( 11)