. 7
( 11)


into gatekeepers who shutter out and shut down con¬‚ict entrepreneurs before it
becomes too late. I contend that the passive support of the public that bene¬ts
from eliminationism but is not intoxicated by it might, to some degree at least,
be dissipated by regulatory structures that sanction passive support. I develop this
idea in Chapter 7. Although it is not evident that collective sanctions actually
will dissuade public acquiescence (perhaps the passive public also lies beyond
deterrence?), what is evident is that a regulatory system based on select crimi-
nalization, which never even reaches the key constituency of the passive public,
forecloses this possibility and with it a valuable line of research and inquiry.
International criminal law is deeply paradoxical: it courageously operates in
opposition to state interests while stubbornly protecting state interests.108 To the
extent that international criminal law pins blame for atrocity on a small number
of horrible individuals, who generally control a state apparatus, it achieves some
justice and curbs atrocity as a tool of a state™s foreign or domestic policy. However,
if in the process of attributing guilt it pulls our gaze away from the many other
actors involved in the tapestry of atrocity “ including malfeasant, complicit, or
distracted states and their of¬cials, along with decisionmakers in international
organizations “ then it will do little to root out atrocity™s multicausal origins. A
fuller picture of responsibility for wrongdoing will emerge only to the extent
that we resist simple, and comforting, criminal explanations and reach deeper
to a more embarassing place. The institutionalization of some accountability
through criminal trials “ and the conversations these trials produce “ must not
lull us into thinking we have attained justice, but should prod us to go much

(iii) expressivism
Expressivists contend that trial, conviction, and punishment appreciate pub-
lic respect for law. The expressivist punishes to strengthen faith in rule of law
among the general public, as opposed to punishing simply because the perpe-
trator deserves it or because potential perpetrators will be deterred by it. Expres-
sivism also transcends retribution and deterrence in claiming as a central goal
the crafting of historical narratives, their authentication as truths, and their ped-
agogical dissemination to the public. Overall, expressive objectives receive less
attention than retribution or deterrence in the jurisprudence of institutions that
174 Atrocity, Punishment, and International Law

pursue extraordinary international criminals, although they are reliably invoked
as justi¬cations for imposing sanction.109
Much of expressive theory relates to trial and conviction. For example, Judge
Patricia Wald observes that taking indictees into custody and prosecuting them
“put[s] the ¬‚esh of situational application on the bareboned de¬nitions of war
crimes, crimes against humanity, and genocide [ . . . ].”110 It is tempting for the
expressivist who extols the norm-generating and dramaturgical function of law to
focus on trial and conviction. However, punishment, too, has signi¬cant messag-
ing value “ both as an end in and of itself and, also, as contributing to the force
of prosecution and conviction. David Garland posits that punishment “com-
municates meaning . . . about power, authority, legitimacy, normality, morality,
personhood, social relations, and a host of other tangential matters.”111 The
fact that consequences follow a guilty verdict makes law all the more real to
the community.112 This sends a message that the law is to be taken seriously.
Emile Durkheim observed that by expressing condemnation, punishment in
fact could strengthen social solidarity.113 Punishment internalizes “ and even
reinforces “ social norms among the public and, thereby, from the expressivist
perspective proactively promotes law-abiding behavior. Moreover, punishment
can serve a prophylactic purpose “ carrying with it signi¬cant therapeutic value
for victims.
If punishment signals the absolute immutability of core values “ for exam-
ple, the universal repugnance of discriminatory group-based killings “ then
initial plans by con¬‚ict entrepreneurs to inveigle and habituate killers may stall.
Punishment can thereby impede the early indoctrination phases in which aver-
age citizens become assimilated into the machinery of mass violence. This
objective of punishment differs from deterring individuals from killing after
they have become habituated into killing by desire or desperation. Whereas it
seems problematic to deter “ through fear of distant and deferred punishment “
violence once it is imminent or has already begun, it seems somewhat more
plausible to inhibit the mainstreaming of hatemongering as politics owing to
the consolidation, through law and punishment, of a social consensus regarding
the moral unacceptability of such politics. Law and punishment may be able to
decelerate indoctrination because potential indoctrinees to the inchoate stages
where hate is normalized have come to see discrimination-based massacre as
manifestly illegal. Assuredly, it is dif¬cult to combat the dizzying effects of propa-
ganda. But if punishment can create principled citizens who value a normative
structure that repudiates group-based eliminationism, then the size and atten-
tiveness of the propagandists™ audience would drop. In this vein, punishment
operates as moral educator.114
Legal process can narrate history and thereby express shared understand-
ings of the provenance, particulars, and effects of mass violence; punishing
the offender contributes yet another layer of authenticity to this narration.
Truthtelling (or, more colloquially, “discovering the truth”) has been acknowl-
edged by international criminal tribunals and is itself tied to a number of other
Quest for Purpose 175

goals, including the consequentialist goal of national reconciliation.115 Discov-
ering the truth also is frequently evoked by atrocity victims as an important
objective of retrospective legal interventions. Trials create archives of infor-
mation: either through documents, as at Nuremberg, or through testimony,
as at Eichmann™s trial in Jerusalem. The ICTY™s dogged prosecution of the
Srebrenica massacre led to “an archive of eyewitness accounts and often grue-
some photographs and videos.”116 These materials can turn tragedy into a teach-
ing moment. Trials can educate the public through the spectacle of theater “
there is, after all, pedagogical value to performance and communicative value
to dramaturgy.117 This performance is made all the more weighty by the reality
that, coincident with the closing act, comes the in¬‚iction of shame, sanction,
and stigma upon the antagonists. Prosecution and punishment in response to
extraordinary crimes can thereby serve a broader didactic purpose that meets
the interests of history and memory.118
The ICTR™s judicial characterization of the massacre that took place in
Rwanda in 1994 as genocide serves the purpose of indelibly memorializing the
violence; the ICTY Appeals Chamber also very consciously used its judgment
in Krsti´ as a vehicle to pursue declaratory objectives so as to of¬cialize the
Srebrenica tragedy as genocide.119 Prosecution and punishment can manufac-
ture an authoritative version of the truth and, thereby, narrate a story that later
becomes history. The IMT at Nuremberg put a repertoire of Nazi barbarities
on display and condemned “ before the international community “ those of its
architects who had survived so as, in the words of Robert Jackson, to “establish
incredible events by credible evidence.”120 Now, sixty years later, the Nuremberg
judgment remains a ¬xed anchor of our children™s education.
There is good reason to believe that the punishment in¬‚icted by an inter-
national tribunal operating prominently on the global agenda at the cusp of
history has enhanced expressive value in asserting the importance of law, the
stigmatization of the offender who transgresses that law, and the authenticity of
the historical narrative that ensues. International trials have a better chance of
becoming the kinds of “popular trials” that de¬ne a debate, remind us of the
content and value of law, or serve as intergenerational “signposts” in history.121
This is in part because international trials reach a global audience.122 Their lib-
eral legalist modalities are intelligible to communities in the epicenters of global
power. Their reliance on due process may help justice to be seen to be done. On
the other hand, too much due process may give rise to technical proceedings
seen to be overly tilted in favor of iniquitous defendants, who become able to
grandstand and humiliate witnesses.
The didactic value of international proceedings is not preordained. The Tokyo
Tribunal has not become a pedagogical anchor in a manner comparable to the
Nuremberg Tribunal. Contemporary international institutions must be careful
not to overlook the audience that matters more than any other “ namely, directly
af¬‚icted populations. Perceptions among such populations that contemporary
institutions lack clean hands will not be dissipated by fastidious adherence to
176 Atrocity, Punishment, and International Law

due process alone. In determining a process to be just, audiences will assess
much more that simply whether it accords with liberal legalism.
Other than Eichmann, national trials of Nazi atrocity “ whether conducted
by civilian or military instrumentalities “ have not reached Nuremberg™s expres-
sivist level.123 That said, national proceedings regarding Nazi atrocity did pro-
duce salient expressive content, even when it came to the implication of non-
Germans. The Barbie, Touvier, and Papon trials were, at least for the French
nation, didactically valuable popular events. Other proceedings, despite result-
ing in lenient sentences completely disproportionate to the gravity of the under-
lying offenses, narrated the horrors of the Nazi concentration camps to a bewil-
dered public. These proceedings “ many of which were undertaken by West
German courts “ ¬lled a critical gap in the historical tapestry insofar as the
Nuremberg prosecutions were directed toward Nazi aggressive war, not crimes
against humanity or the Holocaust.124 Furthermore, one of the strengths of cer-
tain national institutions is the diversity of mechanisms they rely upon to didacti-
cally weave narratives. In Rwanda, for example, the partie civile lawsuits adduce
and personalize stories of suffering and loss in a victim-centered manner. Mato
oput in Uganda relies on ritual to reintegrate offenders while respecting their
own suffering, which seems particularly apt in the case of child soldiers.
Assuredly, whether liberal criminal trials narrate historical truths that, in turn,
have expressive legitimacy remains a contested question.125 I believe they are
capable of such a function, although I certainly recognize that alternate forms
of accountability may have equivalent or even enhanced truth-telling capacity. I
also recognize that criminal prosecution, followed by incarceration, is limited in
its truth-telling function. In particular, four speci¬c aspects of criminal process
and sanction challenge the quality of the narrative output. These aspects are:
(1) selective truths; (2) interrupted performances; (3) management strategies;
and (4) plea bargains. I consider each of these in turn.

a. Selective Truths
Criminal trials are deliberately selective in terms of the truths they produce. The
application of modern rules of evidence and procedure frames this selectivity.
These rules favor the production of logical and microscopic truths over the dia-
logic and experiential truths that emerge phenomenologically from restorative
justice initiatives.126 For Miriam Aukerman, the formalism and rigidity of trials
make them at times “excruciatingly boring.”127
The rules may create more than just tedium. Although bolstering the authen-
ticity of the narrative, these rules paradoxically also may crimp it. For example,
Martti Koskenniemi writes that evidentiary rules and due process may under-
mine memory by allowing the accused to belittle accusers in cross-examination
and reduce their accusations to “panicky ˜I don™t know™ statements.”128 Rules
may truncate victim storytelling, thereby sowing disappointment;129 but, on the
other hand, may control the extent to which victim storytelling serves ulterior
political purposes unrelated to the guilt or innocence of the accused. Rules also
Quest for Purpose 177

exclude as nonprobative certain facts that local audiences might ¬nd deeply rel-
evant and, in this regard, distort the historical narrative. The situation of Belgian
courts adjudging Rwandan g´ nocidaires constitutes an example. Although the
Belgian prosecutions should be lauded for bringing systematic human rights
abusers to justice, they also rewrite the historical record by presenting Belgium
as a font of justice, instead of weaving into the judicial narrative the much more
complicated role Belgian colonial interventions played in exacerbating ethnic
divisions in Rwanda that laid the groundwork for eventual genocide.
Expressive value is further threatened by the reality that this value often is
externalized from af¬‚icted local communities owing to the distance and mis-
trust evident between such communities and international criminal tribunals.
Procedural differences between liberal criminal trials and expectations among
local populations, in particular non-Western populations, also diminish the pro-
phylactic value of verdict and punishment.

b. Interrupted Performances
The death of Slobodan Miloˇevi´ in the midst of his trial (which, at the time
of his death, had gone on for four years) illustrates the frailties of criminal
process. To be sure, a trial that stops short of verdict and punishment is not
denuded of all expressive value. Prosecuting Miloˇevi´ allowed a worldwide
public to learn in dribs and drabs of the charges against him and the details
of the atrocities he allegedly coordinated. Instrumentally speaking, some of the
testimonial and documentary evidence introduced during the Miloˇevi´ pro- sc
ceedings will be used against other defendants. But Miloˇevi´ ™s death denied
the possibility of a ¬nal sentence: infallible and authoritative. The curtain fell
before the closing act. When the antagonist dies before the protagonist™s pur-
suit is complete, the script becomes frustrated. The performance reaches an
end, but it is an anticlimax. A formal adversarial trial cannot continue posthu-
mously, at least not under current understandings of internationalized due
Miloˇevi´ ™s premature death is an obstacle to the ICTY™s narration of an
overarching story of death and destruction in the Balkans. The ICTY has mit-
igated the impact of this obstacle by indicting 161 individuals in total; and,
quickly following Miloˇevi´ ™s death, by moving ahead with other high-pro¬le
trials, including regarding atrocity at Srebrenica and in Kosovo. That said, the
ICTY had plea-bargained with other defendants, giving up reduced sentences in
exchange for the promise of prized testimony against Miloˇevi´ . These bargains
crimped the expressive value of punishing those defendants in the hopes of a
blockbuster impact in ringingly convicting Miloˇevi´ .
The expressive vulnerabilities of criminal trials, and the impact of an inter-
rupted performance, can be minimized to the extent that the net of account-
ability is broadened. In particular, if accountability ranges beyond high-pro¬le
criminal trials, the resultant greater methodological diversi¬cation diminishes
the risk that an interrupted performance scuttles the overall truth-telling process.
178 Atrocity, Punishment, and International Law

The prosecution of leaders rendered frail through the passage of time neces-
sarily involves a race against time. The sooner justice is delivered the better. Wily
defendants can dither, piddle, and delay. Popular trials create a platform that
places the defendant onto the world™s center stage. If the defendant can make
the trial all about himself, and sel¬shly control the stage though grandstanding,
histrionics, and manipulation, then the proceedings drift away from the victims
and their terrible losses.

c. Management Strategies
The Miloˇevi´ trial™s performativity was susceptible to interruption in part
because the trial had dragged on for so long. The Iraqi High Tribunal (IHT)
applied some lessons learned from the languidness of the Miloˇevi´ proceedings
to its prosecution of Ba™ath Party leaders, including Saddam Hussein.
First, IHT judges exhibited greater vigilance than their ICTY counterparts in
controlling the courtroom and the content of the discussions. On the one hand,
tight control secures managerial and bureaucratic goals, streamlines process,
dissipates in¬‚ammatory controversy, and preserves judicial authority. On the
other hand, though, as levels of control become too tight, they may strangle the
judicial record and thereby in¬‚ict credibility costs. Flattening the narratives to
protect power drains some of their transformative content.
Second, IHT prosecutors elected to proceed through a series of minitrials
instead of, as had been the case with Miloˇevi´ , one overwhelming omnibus
sixty-six count proceeding. The ¬rst minitrial, which led to convictions for crimes
against humanity and war crimes against seven defendants (and a variety of sen-
tences, including a death sentence for Saddam Hussein), involved the killings “
at the hands of the Iraqi state “ of 148 residents of the Shiite village of Dujail.130 In
1982, Dujail had been the site of a failed assassination attempt against Hussein.
In response, Iraqi security forces detained suspects. The Iraqi Revolutionary
Court subsequently sentenced these villagers to death. Executions were carried
out. Hussein™s signature was on the orders.
Subsequent IHT minitrials do involve a higher-stakes context: for example,
proceedings related to the Anfal (Arabic for “spoils of war”) campaign, which had
resulted in the allegedly genocidal massacre of at least (a conservative estimate)
50,000 Kurdish civilians in 1988, and the crushing of the 1991 Shiite upris-
ing in the south. By proceeding sequentially, IHT prosecutors ensure cyclical
episodes of grati¬cation and closure, thereby reducing the risks that long-term
proceedings lead to a deferred all or nothing outcome. They allow different
victim groups, for example Kurds and Shias, to express outrage at the travesties
in¬‚icted upon them through context-speci¬c proceedings. This is a prudential
move. However, it is not without its own drawbacks. It results in a dramaturgi-
cal methodology in which the narrative is told through iterated vignettes. IHT
of¬cials need to be diligent that the digestible parts add up to a compelling,
overarching whole. If discontinuous lower-stakes convictions remain narratively
Quest for Purpose 179

fragmented, then the IHT may, in the name of prudence, have forsaken the
opportunity to leave a hardier historical footprint. Moreover, hanging Hussein
for the Dujail conviction before the remaining minitrials took place induced
an interrupted performance detrimental to the expressive value of these other

d. Pleading Out
Can plea bargains attain truth-telling objectives? Indeed, offenders who plead
guilty may admit wrongdoing, apologize, express remorse, dignify victims, and
provide details regarding the crimes. Self-convicting offenders may even impli-
cate others, although this is not always the case (nor is there any guarantee of
the veracity of the evidence subsequently proffered).131 With regard to high-level
accused, where the exacting nature of the criminal law requires the leader to
be traced to the bodies interred in the mass grave, plea bargains can offer a
partial print of the truth whose value exceeds that of the acquittal that might
result should the prosecution be unable to meet the high threshold of proof
demanded in the pursuit of microscopic and logical truths.
Although the ad hoc tribunals af¬rm that plea bargains contribute to truth-
telling objectives,132 certain institutionalized aspects of plea bargaining at the
ad hoc tribunals whittle down the narrative value of plea-bargained convictions
and sentences. Although some agreements contain a detailed factual basis, in
other cases the offender pleads guilty to fairly bare allegations. In the latter case,
the offender avoids contending with the gruesome, detailed evidence that would
be admitted at trial. Deronji´ ™s plea agreement, which was judicially af¬rmed,
cursorily established the truth only regarding the tragedy that encompassed one
village on one particular day, thereby burying several other potential truths “
namely, accusations involving other spaces and places in Bosnia.133
Charge bargaining, in particular, jeopardizes expressive storytelling. Plavˇi´ ,
in an agreement af¬rmed by the ICTY, pled guilty to one umbrella count of per-
secution as a crime against humanity and the Prosecutor dropped the remaining
seven charges, including two counts of genocide and complicity in genocide.134
In Milan Simi´ ™s case, the ICTY Prosecutor “agreed to withdraw several counts,
including the most serious “ persecution as a crime against humanity relat-
ing to Simi´ ™s mayor-like role [ . . . ].”135 Simi´ , a paraplegic, was sentenced to
c c
¬ve years™ imprisonment. Combs notes that “such a sentence would have been
unthinkable had the factual basis for Simi´ ™s conviction encompassed all the
conduct for which he was initially charged.” As discussed earlier in the context
of retribution, charge bargains push certain allegations off the agenda, thereby
precluding the truth of those allegations from being of¬cially unearthed. It is true
that pleading guilty to an umbrella charge of persecution, a result that obtains
in certain plea bargains, permits a broad array of facts, which may well support
the substance of all of the original charges, to be included in the judicial record.
However, the practice of the ad hoc tribunals has been spotty in this regard.
180 Atrocity, Punishment, and International Law

(iv) conclusion
The preference for incarceration following what liberal international lawyers
deem to be an acceptable criminal trial on the whole falls short of its penological
objectives, in particular retribution and deterrence. This may be because those
objectives are too ambitious. It may also be because the criminal law, standing
alone, simply is not enough nor can ever be enough.
chapter 7

From Law to Justice

In this chapter, I outline two proposals. These proposals begin a conversation.
They do not seal a conclusion. The proposals respond to two major, and intercon-
nected, shortcomings of dominant modalities of prosecution and incarceration:
(1) the democratic de¬cits and externalization of justice that they trigger; and
(2) the dif¬culties they experience in attaining stated penological objectives.
The two proposals, which I present as reforms, are: (1) vertical: to rescript the
division of labor toward greater inclusiveness of in situ sociolegal institutions
and bottom-up input; and (2) horizontal: to look beyond criminal process and
welcome the general regulatory power of law, as well as extralegal interventions,
to holistically capture the broad-based complicity that inheres in mass atrocity.
These two reforms are synergistic. When twinned, they promote a pluralistic
understanding of accountability; their goal is to activate a broader sense of
justice in the aftermath of atrocity. These reforms aim to move the agenda from
law to justice: initially, by advancing international criminal law to a capacious
law of atrocity and, ultimately, to an enterprise that constructively incorporates
extrajudicial initiatives.
Of course, international criminal law as currently implemented through pros-
ecution and incarceration goes some way to promote justice. But it is intrinsically
limited. A richly multivalent approach could go further. It could push penologi-
cal objectives, for instance reintegration and restoration, which heretofore have
been given short shrift. Moreover, a richly multivalent approach could consol-
idate, and better attain, expressivism; and, perhaps, offer new ways of thinking
about how deterrence and retribution “ much more dif¬cult objectives “ might
be operationalized. Such an approach also reminds international lawyers of an
important lesson, wonderfully articulated by Paul Roberts, that international
criminal law is only “ and can only be “ part of the justice picture.1
One implication of these reforms is that sanction might look different and
assume different calibrations in each case of atrocity. Sanction might involve
court proceedings and interventions by institutions other than courts. This recog-
nizes the distinctiveness of each individual atrocity. Recognizing this distinctive-
ness is a valuable exercise. After all, although all three are genocides, the atrocities
of the Holocaust, Rwanda, and Srebrenica (Bosnia) are not identical. Variation
182 Atrocity, Punishment, and International Law

in sanction, however, creates some unevenness, perhaps even some messiness.
Clarity, and our expectations for clarity, could cloud. I envision a penology that
gains its independence through its embrace of procedural diversity. How does
this accord with the position, to which I subscribe, that discrimination-based
atrocity constitutes universal great evil? Might pluralist implementation frag-
ment the coherence of the repudiation of the universal nature of the evils at
hand? Before elaborating on the proposed reforms, I sketch a response to these
questions. I draw from cosmopolitan theory to justify a position that holds that,
although genocide and discrimination-based crimes against humanity are uni-
versal evils, they coherently can be sanctioned in diverse manners that may take
different forms in light of the distinctive social geographies of various atrocities.

(i) pluralist process for universal evil?
Martti Koskenniemi engages with the notion that international law fundamen-
tally is a European tradition derived from a desire to rationalize society through
law.2 From this general perch, it is not too far to jump to the speci¬c conclusion
that condemning the repugnance of extreme evil through law re¬‚ects a ratio-
nalization that may not be universally shared. Koskenniemi, however, goes on
to conclude “ in no uncertain terms “ that “[t]he fact that international law is a
European language does not even slightly stand in the way of its being capable
of expressing something universal.”3
And substantive international criminal law does express something universal,
recognized by deep traditions in moral philosophy: namely, the condemnation
of acts of great (or, in Arendt™s phrasing, extreme) evil and wickedness. Stuart
Hampshire sagely advises that “[t]here is nothing . . . culture-bound in the great
evils of human experience, reaf¬rmed in every age and in every written history
and in every tragedy or ¬ction [ . . . ]. That these great evils are to be averted is
the constant presupposition of moral arguments at all times and in all places
[ . . . ].”4 David Luban, whose work bridges law, ethics, and morals, pointedly
adds: “There is no society [ . . . ] in which gratuitous in¬‚iction of the great evils
is tolerable.”5 No human being wishes to be victimized by the wicked or the
evil. The concomitant to this right is a remedy. This remedy re¬‚ects another
universally shared moral value, namely that the victims of great wickedness can
demand that those who in¬‚icted such acts upon them be held accountable.
It is one thing to agree to the universal repudiation of the great evils and to
agree that victims are entitled to accountability. It is another matter to accept
the universality of categorizing the great evils as crimes. This categorization
is widely accepted among international lawmakers. It is one that the architec-
ture of international criminal justice has internalized. Larry May ¬nds deep
justi¬cation for the universal categorization of the great evils as crimes.6 May™s
work bears parallels to the Eichmann judgment, in which the Supreme Court
of Israel af¬rmed the universal criminality of Eichmann™s wrongdoings because
they “constitute[d] acts which damage vital international interests,” “impair[ed]
the foundations and security of the international community,” and violated
From Law to Justice 183

“universal moral values and humanitarian principles.”7 One of the bene¬ts
of criminalizing the great evils is that the proscriptive certainty required by
the criminal law has triggered fairly precise de¬nitions of genocide and crimes
against humanity.8
My argument regarding the universal condemnation of the great evils does
not hinge upon proof of the universality of categorizing the great evils as crimes.
I accept this categorization, at least functionally, insofar as I routinely invoke the
settled category of extraordinary international criminality and contrast it to ordi-
nary domestic criminality. Nonetheless, law can condemn behavior and repair
victims in many ways. Behavior also can be condemned through sources exoge-
nous to the law entirely. Accordingly, it could be argued that casting the great
evils as crimes represents another imposition of Western juridical categories to a
diverse world order. In the West, the most odious social transgressions are viewed
as crimes sanctionable by the state through imprisonment or, exceptionally (and
decreasingly), execution. This may not be the case everywhere, however. In other
places, the most odious social transgressions may be cast as delicts sanctionable
by society through restitution, reparation, or countermeasures. However, for the
purposes of my argument, what matters is the universality of the condemnability
of the underlying substantive harm, this being the in¬‚iction of great evil, and
the universality of the notion that victimizers are to be held to account. In fact,
as I unpack later in my discussion of horizontal pluralism, there is considerable
value in classifying the great evils as something more than just crimes. Doing so
expands the lexicon of international justice and, thereby, permits prudent appli-
cation of the broader panoply of accountability modalities and methods that are
available under law generally. Outreach to private law “ such as obligation, tort,
contract, and restitution “ and to quasilegal initiatives “ such as public inquiries
and truth commissions “ enhances the quality of accountability.
It is crucial to separate the substantive goals at hand, namely the condemna-
tion of great evil and the promotion of accountability in its aftermath, from the
process regarding how these goals are to be operationalized and the institutions
where this process is to occur. All national legal systems know a distinction, albeit
often blurred,9 between process and substance. Supranational legal orders also
know such a distinction.
It is at the level of the procedural, broadly de¬ned, and the institutional,
broadly understood, that international criminal law as technique is most suscep-
tible to the claim of pyrrhic universalism, deeply rooted in Western visions of
what process should look like. As May intuits: “Those of us raised in the Western
legal tradition often have a visceral reaction to attempts to sidestep legal trials.”10
International criminal law does not exclude individuals outside this tradition
through its condemnation of great evil but, rather, through the ascension and
now expatriation of a particular mode of reporting evil and punishing perpe-
trators. It is at the procedural level that the contingency of international crimi-
nal law™s universalism seems starkest. Assuredly, certain aspects of international
criminal process tap into something genuinely universal in the human expe-
rience. For example, all legal systems envision the need for some relationship
184 Atrocity, Punishment, and International Law

between wrongdoing and sanction. However, international criminal process has
other modalities whose cultural contingency does not resonate universally, in
particular with regard to sanction. One-size-¬ts-all application of these modali-
ties creates dissonance, which is evident in places whose atrocities currently are
being judicialized.
The modalities of international criminal law, in particular those related to
punishment and sentence, tend to universalize through ideological preference
instead of through an independent assessment of the social psychology of the vio-
lence, comparative re¬‚ection about how diverse justice traditions might punish,
and development of multilateral interinstitutional conversations. The choices
are not binary: namely, either to accept the received wisdom of extant interna-
tionalized institutions, on the one hand, or the void of impunity, on the other. It
is disappointing that so much of the brilliant work of international criminal law,
namely the de¬nition of the substantive crimes and the march past impunity,
has been accompanied by a seemingly casual path dependence in delineating
institutions and methodologies capable of putting that work into action.11
That said, we must ask: are some legal processes simply better suited than
others to denounce extreme evil, condemn it, and prevent its reemergence?
If so, are Western methodologies of adjudication and punishment the best we
have? Indeed, were this to be the case, then there would be some experiential or
empirical justi¬cation for their in¬‚uence over international discourse and their
transplant back to national legal orders, even to those orders where they may
not be indigenous. Clearly, there is much good in replacing the many things
that do not work, or that work poorly, with the one that does work, or at least
works better. Despite the monochrome that may result, the question at hand is
not one of aesthetics.
However, in the case of internationalized modalities of punishment, this
argument is not persuasive. The prevailing way of doing things fails to meet the
principal goals that it places upon itself: namely, retribution and deterrence. I am
not convinced that individualized criminal trials and incarceration self-evidently
are the best that we can do. In fact, diversifying processual and penological
methodologies could be a step toward augmenting our collective experiences,
and empirical knowledge, with regard to the role and effectiveness of law in the
aftermath of atrocity.
May notes “a strong Western bias” in the ICC and he provides starkly proce-
dural examples of this, namely the adversarial method and cross-examination.12
He then implies that some sort of procedural uniformity might be inevitable,
“since courts must follow some model, whether Western or non-Western.”13
Assuredly, institutions must abide by some procedure, and May is wise to insist
on this. But that procedure need not be the same everywhere. Some model
does not ineluctably lead to the same model. Multiculturalist theorists such as
Charles Taylor and Radhika Coomaraswamy teach us that the enforcement of
positive human rights need not be static and ¬‚at. Coomaraswamy notes that
“internationally accepted standards and norms do exist, in de¬ance of post-
modernist tendencies. [ . . . ] What must be seen as negotiable are the strategies
From Law to Justice 185

of enforcement and implementation [ . . . ].”14 In a similar vein, the enforce-
ment of humanity™s right to hold its enemies accountable can proceed through
diverse methodologies and integrated sanctions. In the end, humanity can coher-
ently strive to implement universal values through pluralistic procedures and

(ii) cosmopolitan theory
The supple model of accountability that I propose is compatible with cosmopoli-
tan visions of law and authority in the international context. Essentially, cos-
mopolitanism is a tradition in sociopolitical and legal philosophy according to
which all human beings belong to a single community. This community derives
from, and in turn fosters, certain shared values. Cosmopolitans differ as to the
number, nature, and depth of the values that are shared.
Cosmopolitans hold in common a belief that we all owe some duties to each
other. We have obligations and entitlements based on our status, in Diogenes
the Cynic™s phrasing, as “citizens of the world” rather than as citizens of a
particular state.15 Accordingly, we cannot think of distributive justice obligations
as stopping at our national borders. Our responsibilities do not end at the customs
and immigration desk. Instead, they seep beyond, to strangers in faraway lands.
Cosmopolitanism, therefore, offers a philosophical basis for global governance
over a global public.
As with all theoretical perspectives, cosmopolitanism is diverse. A wide vari-
ety of views can be called cosmopolitan. Although cosmopolitans often are
depicted as avid universalists who shutter out local or national identities, this
is a caricature. A more textured reading of cosmopolitan theory reveals much
greater nuance. Even ancient cosmopolitans such as the Stoics recognized that
the citizen of the world need not give up his or her local identi¬cations. These
identi¬cations, in fact, enrich the citizen™s life. Accordingly, the Stoics under-
stood an individual™s identity as ranging through concentric circles that extend
outward from the personal, to the local, and, ultimately, to humanity as a whole:
each circle constitutes an important element of overarching identity.16
Many contemporary cosmopolitans also engage with the diversity of the famil-
ial, local, and national. They accept that human beings have conservative sensi-
bilities that ¬nd comfort and meaning in inherited localism. They acknowledge
local identities as part of the complex nature of human agency and thereby rec-
ognize the pull of partiality and pluralism. The work of leading cosmopolitans
such as Martha Nussbaum,17 David Hollinger,18 David Held,19 Kok-Chor Tan,20
Kwame Anthony Appiah,21 and Paul Schiff Berman22 recognizes, and in most
cases welcomes, multiple af¬liations and overlapping associations.
Among contemporary cosmopolitans, Nussbaum is blunt about her belief
that an emphasis on patriotic pride is “morally dangerous.”23 She warns that
“patriotism is very close to jingoism.”24 However, even Nussbaum refuses to
negate the fact that “all profound human matters are differently realized in dif-
ferent societies.”25 She holds to the position that “[n]one of the major thinkers in
186 Atrocity, Punishment, and International Law

the cosmopolitan tradition denied that we can and should give special attention
to [ . . . ] our own ties of religious and national belonging.”26 Appiah is bolder in
arguing that it is possible for a person to be a cosmopolitan and also a patriot.27
Humanity can commit to universal standards while celebrating difference. For
Appiah, cosmopolitanism in fact delights in the diversity of human cultures.
Institutions can have cosmopolitan obligations (as envisioned by Nussbaum)
that, I would add, they can articulate through local conduits, even those that may
strike the sophisticated global citizen as parochial. Institutions, for instance the
ICC, could host the kind of conversations that Appiah believes bridge encoun-
ters among “people from different ways of life.”28
Tan, while recognizing certain universal moral concerns, takes national and
patriotic attachments seriously. For Tan, these attachments have a de¬ned place
within a cosmopolitan theory of justice. In Democracy and the Global Order,
Held explores how power could be exercised beyond the jurisdiction of states at
the global level for a global public, but he also recognizes that such an exercise of
power could promote a democratic de¬cit. Berman offers a particularly interpen-
etrative and ¬‚exible approach to cosmopolitanism. He notes: “A cosmopolitan
conception of law [ . . . ] aims to capture a middle ground between strict territori-
alism on the one hand and expansive universalism on the other.”29 Attachments
can be ¬‚uid, multilayered, and malleable. Consequently, cosmopolitanism is
not inhospitable to pluralism. Berman concludes:
A cosmopolitan conception [ . . . ] makes no attempt to deny the multirooted
nature of individuals within a variety of communities, both territorial and non-
territorial. [ . . . ] [C]osmopolitanism offers a promising rubric for analyzing law
in a world of diverse normative voices. [ . . . ] [I]t celebrates diverse normative
orders in multiple communities and need not insist on homogenizing that
diversity into one global culture of one international legal framework.30

The model I propose recognizes the universality of our shared membership in
a moral community that condemns great evil and entitles victims thereof, in par-
ticular those most directly affected, to accountability. This aligns the model with
the basic precept of cosmopolitanism. Moreover, the model adopts cosmopoli-
tanism™s acceptance of the richness of local identi¬cations, particularly when
this richness helps promote justice and the ability for people to lead (and live)
a good life. The notion of diverse procedure for universal wrongdoing thereby
¬ts within a cosmopolitan theory of law, tending toward the pluralist end of the
continuum. I avail myself of the term “cosmopolitan pluralism”31 to describe
this position. Cosmopolitan pluralism would support substantive censure at the
global level, but endeavor to allay democratic de¬cit concerns through opti-
mistic incorporation of local control, process, and sanction. Another advantage
to reforms that promote cosmopolitan pluralism horizontally and vertically is
that these reforms recognize that, although instances of genocide and crimes
against humanity are universally condemnable, each particular instance is origi-
nal in its own regard and thereby retains a level of distinctiveness. Cosmopolitan
pluralist reform would preserve the expressive value of punishing the universal
while respecting the importance of not ¬‚attening the particular by permitting
From Law to Justice 187

accountability modalities to vary in each individual case.32 It is through this rela-
tional interplay between universal accountability and pluralistic enforcement
that an independent criminology and penology for mass atrocity can emerge.
A premise of this book is that one of the reasons international criminal law
falls short is because it treats the extraordinary international criminal like the
ordinary common criminal. One extrapolation that could be made from this
premise is the need to develop a cosmopolitan vision of punishment the content
of which, different than that applied to ordinary criminals, becomes universally
applicable to all extraordinary international criminals everywhere. This is not
the direction I take. Such a direction, in fact, runs counter to a pluralized vision
of punishment in which local attachments are to be welcomed and in which
applications of authority ought to begin with the bottom-up and not push down
from the transnational top. Within this pluralized vision, some societies might
decide that it is in their best interests to subject the extraordinary international
criminal to common processes of adversarial prosecution and isolated incar-
ceration. Obviously some polities “ for instance, in the West “ already have
done so. It is possible that many polities freely come to the same conclusion.33
It is probable that some will not, while others will come to a more nuanced
position. Regardless of the direction that actual pluralism takes us, the current
internationalization of the preference for prosecution and incarceration is not
the result of bottom-up consensus but, rather, the product of powerful state
and political interests. This leads to disquieting results insofar as the application
of international criminal law overwhelmingly occurs in non-Western localities
where formal state institutions that propound criminal trial modalities may do
so for ulterior motivations. Western prosecution and punishment for deviant
criminals has become transplanted to (and for) places where it may not re¬‚ect
what af¬‚icted populations would in good faith come up with on their own for
group-based tragedies.
Until the voices of af¬‚icted populations are more clearly heard, and loosened
from the primacy or complementarity of internationalist visions, we simply do
not know exactly what values pertaining to the punishment of the enemies of
humankind truly are shared among us all. Herein lie the beginnings of a sui
generis penology for mass violence. Creating a safer space than what presently
exists for the articulation of such voices is an important step. Insofar as the
evidence on the ground is that af¬‚icted communities seek a diverse array of
legal and extralegal initiatives to respond to atrocity, the emergence of liberal
legalism as the preferential and dominant response seems more pragmatic than
genuine. It seems more political than re¬‚ective. In the next section, I consider
how a somewhat safer discursive space can be created.

(iii) vertical authority allocations: a case
for quali¬ed deference
I propose a test of quali¬ed deference “ in contrast to primacy or complementar-
ity “ in the vertical allocation of institutional authority, which currently radiates
downward from the international, over extraordinary international criminals.
188 Atrocity, Punishment, and International Law

I draw some guidance from European Union experiments with subsidiarity as
an ordering principle. Subsidiarity, a lynchpin of European constitutionalism,
requires “any infringements of the autonomy of the local level by means of pre-
emptive norms enacted on the higher level to be justi¬ed by good reasons.”34
But there are limits to the range of lessons that can be learned from experiments
with subsidiarity. It is dif¬cult to transpose the value of subsidiarity to the context
of punishing extraordinary international criminals. After all, subsidiarity aims to
harmonize local law with the supranational in stable periods. What is more,
subsidiarity does not involve the kinds of widescale collective action problems
that are intrinsic to mass atrocity. National and local entities ought to be entitled
to deference in times of postcon¬‚ict justice, but subsidiarity could prove to be
too generous.
Quali¬ed deference does not involve a blind retreat to national or local
institutions. Such a retreat would be problematic. In some postcon¬‚ict soci-
eties, juridical institutions are devastated, illegitimate, corrupt, manipulable,
complicit in violence, or in the service of repressive social control; not all post-
con¬‚ict societies move toward democracy or peace, some trend in the direction
of authoritarianism; some postcon¬‚ict societies look more like societies between
con¬‚icts. Complementarity, however, is too controlling “ whether intentionally
or unintentionally “ given the incentives it creates for local institutions. Quali¬ed
deference strikes a middle ground between subsidiarity and complementarity.
It creates a rebuttable presumption in favor of local or national institutions that,
unlike complementarity, does not search for procedural compatibility between
their process and liberal criminal law and, unlike primacy, does not explicitly
impose liberal criminal procedure.
I also draw some guidance from the margin of appreciation doctrine, a rule
of judicial interpretation most famously applied in the case law of the Euro-
pean Court of Human Rights. Margin of appreciation doctrine “encourages
international courts to exercise restraint and ¬‚exibility when reviewing the deci-
sions of national authorities.”35 Shany identi¬es two major characteristics of
margin of appreciation: (1) a certain degree of judicial deference with regard
to the execution of international law obligations that (2) becomes applicable
to situations of normative ¬‚exibility (i.e., to international norms that are open-
ended, unsettled, “intrinsically uncertain,” and that preserve a signi¬cant zone
of legality).36 Although margin of appreciation is a more expressly legal doctrine
than subsidiarity, and animates much human rights jurisprudence, it “ too “ has
limitations as precedent. The substantive nature of the normative prohibition of
genocide, crimes against humanity, and war crimes is not open-ended, “intrinsi-
cally uncertain,” or unsettled. It is very important for procedural diversity not to
cloud the shared nature of the condemnation of great evil. Fears of such a mud-
died outcome are one reason why margin of appreciation has not been actively
contemplated by international criminal courts. Quali¬ed deference would bet-
ter resist troubling con¬‚ations of the substantive and the procedural.
As an ordering principle, quali¬ed deference meets important utilitarian
objectives in promoting legitimacy, in warding off what I have elsewhere called
From Law to Justice 189

globalitarianism,37 in dissipating the top-down incentives created by comple-
mentarity, and in minimizing unrealistic expectations of local legitimacy upon
which subsidiarity is predicated. Quali¬ed deference does not purport to resolve
all concerns of mediating the particular with universal essentials, but does take
us down a new and, I believe, fruitful path.
There is considerable value to the most traditional bases of jurisdiction “
namely, territoriality and nationality “ that should not be overlooked. Although
institutions of international criminal punishment profess that the legitimacy of
punishment is enhanced by an institution™s international provenance, the expe-
riences of postcon¬‚ict societies reveal a more complex picture. This should not
be surprising, insofar as interpretations of justice are often multilayered and,
for many people, take root in national and local institutional and procedural
contexts. Practically speaking, too, some advantages arise in proceeding locally
in terms of culling and interpreting information, with regard to cultural coding,
and requiring less in the way of immediate translation. For example, much has
been lost in translation from Kinyarwanda into English and French in hearings
at the ICTR; likewise, the East Timor Special Panels have experienced con-
siderable dif¬culty with translation. But the translation issue cuts both ways.
Obligations to translate into French, English, or Portuguese have rendered the
judicial record accessible to a much broader transnational audience, with the
corresponding bene¬t of a wider dissemination of information and denunciation
(thereby augmenting the expressive value of the judicial output). However, noth-
ing precludes a local process under local control from becoming disseminated
to, and subsequently translated for, a global audience.38
If accepted, what would the operation of quali¬ed deference look like? It
falls to those individuals, including members of af¬‚icted communities, who
enforce the universal goal of condemning the great evils at the national and
local levels to ¬ne-tune the interplay and overlap that emerges from the dialogue
between the local and the universal. I propose that the following interpretive
guidelines contour the implementation of quali¬ed deference:

(1) good faith;
(2) the democratic legitimacy of the procedural rules in question;
(3) the speci¬c characteristics of the violence and of the current political
(4) the avoidance of gratuitous or iterated punishment;
(5) the effect of the procedure on the universal substance; and
(6) the preclusion of the in¬‚iction of great evils on others.

These interpretive guidelines would operate disjunctively. In other words, not
all of them must be met in order for the presumption of quali¬ed deference to
a local or national accountability measure to remain satis¬ed. However, a gross
failure on the part of the measure to meet one of the guidelines could suf¬ce
190 Atrocity, Punishment, and International Law

to reverse the presumption in favor of quali¬ed deference. In cases of failure
to meet the guidelines, internationalized interventions should not replace in
situ modalities, but, to the extent possible, work in tandem with local actors to
develop harmonized structures that respond to the shortcomings.
Let us consider these interpretive guidelines in somewhat greater detail,
beginning with good faith. The construction of good faith envisioned by quali¬ed
deference matches that of virtue ethics.39 For the virtue ethicist, character is tied
to actions; and character, in turn, affects whether ethical actions are appropriate.
This gives rise to the proposition that “whether laws are just is a matter of whether
they express or display suf¬ciently good motives on the part of the legislators.”40
Democratic legitimacy is not assured by legislative vote. Measures adopted
through legislative vote tend to have greater legitimacy than those adopted by
executive ¬at. But, as explored in Chapter 4, centralized state institutions (even
putatively representative ones) may not re¬‚ect on-the-ground values in af¬‚icted
communities. To this end, by democratic legitimacy I intend not a formal pos-
itivist process but, rather, a substantive form of social legitimacy.41 In many
places, the state cannot be taken as a proxy for society or for social legitimacy.
Consequently, there is a need to effect a more ¬ne-grained assessment.
The speci¬c characteristics of the violence and of the current political context
consider the degree to which the violence was popular, whether it has ended,
and whether the society has transitioned into peace and relative security. This
interpretive guideline also asks: What are the effects of retrospective accountabil-
ity on prospective stability? Are national or local sociolegal institutions without
capacity? Were they, and their of¬cials, complicit in the violence? Are they cor-
rupt, susceptible to political pressure, or authoritarian? Postcon¬‚ict, is the society
transitioning toward democracy or drifting toward a new totalitarianism? Before
rescinding quali¬ed deference for capacity reasons, decision-makers ought to
consider whether the international community could help build capacity to
effective levels. The level of destitution in the country ought to trigger our
cosmopolitan distributive justice obligations to build up general infrastructure
within and for members of the af¬‚icted community instead of strikingly expen-
sive criminal courts that stand apart from that community.
The avoidance of gratuitous iterated punishment means something slightly
different than typical understandings of ne bis in idem. For example, the Interna-
tional Association of Penal Law understands ne bis in idem as precluding “double
prosecutions and sanctions.”42 It considers that “non-criminal prosecutions and
decisions with an equivalent punitive effect likewise bar a new prosecution.”43
Quali¬ed deference would trend toward a more nuanced analysis. A local initia-
tive that sequentially pursued civil liability after a criminal conviction had been
obtained would not be suspect on that basis alone, even if the civil liability had
a “punitive effect.” Mere differences in the nature or scope of punishment with
regard to what Western legalist institutions would award would not constitute a
basis to rescind deference.
Another guideline for quali¬ed deference is that the procedural methods
not void the substantive content of the shared universal value, this being the
From Law to Justice 191

condemnation of great evil. This guideline prohibits procedural mechanisms
that directly or indirectly rede¬ne the meaning of that evil so as to trivialize it or
render it so elastic that it loses its speci¬city. This does not mean that different
societies are incapable of adopting slightly different substantive understandings
of the meaning of great evil. In the event a society decides to use as a proxy for
the meaning of these great evils the de¬nitions of these evils found in substantive
international criminal law, it should be free to do so; in the event the society
tinkers with these de¬nitions this would not necessarily run afoul of quali¬ed
deference. After all, the development of norms often arises through patterns of
healthy accretion. That said, there is a need for vigilance given the erga omnes
nature of our shared values regarding the moral unacceptability of the great
The ¬nal element of quali¬ed deference is that the local or national modal-
ities not in¬‚ict great evils on other individuals, whether perpetrators or third
parties. This element sets parameters to the kind of punishment that can be
imposed. Punishment cannot take the form of what cosmopolitan values con-
demn as a great evil.
Some real-life examples should help illustrate these guidelines. I take up
three case studies where, at the time of writing, the presumption of quali¬ed
deference would be reversed: Sudan, Afghanistan (with regard to customary law,
the Pashtunwali), and Iraq.
Civil war has raged in the Sudan since the country achieved independence
in 1956. Historically, much of the internecine con¬‚ict has been between the
northern government (Arab Muslim) and rebels (Christian or animist) in the
south, although a peace agreement has been concluded between these parties.
Starting in 2003, however, a new con¬‚ict ¬‚ared up in the western part of the
country, speci¬cally in the province of Darfur, when different rebel groups
(African Muslim) attacked the Arab Muslim government.44 The government of
Sudan retaliated through a campaign directed against civilian targets. It recruited
and equipped members of Darfur™s Arab tribes in this campaign. Most notorious
among the conscripts are the mounted janjaweed militia.45 Janjaweed engaged
(and, as of mid-2006, continue to engage with the not-so-discreet support of
the Sudanese government) in gruesome incidents of pillage, rape, murder, and
the razing of entire villages. It is estimated that at least two hundred thousand
black Africans have been killed and two million more displaced (many into
neighboring Chad). Many of the displaced refugees are ravaged by famine and
An international commission found evidence of crimes against humanity
and war crimes but concluded that the government of Sudan had not pursued
a policy of genocide.46 This commission found that, together with other actors,
the government of the Sudan was responsible for crimes under international law.
The commission report formed the basis of the United Nations Security Council
referral of the Darfur situation to the ICC.47 By virtue of this referral, the ICC is
mandated to investigate and prosecute crimes against humanity and war crimes
in the Sudan. A list of suspects has been compiled. Sudanese President Omar
192 Atrocity, Punishment, and International Law

Hassan Bashir opposes ICC intervention and has stated that he will not hand
over any suspects to the ICC.48 Sudanese of¬cials have obstructed the gathering
of evidence by ICC investigators.49 Notwithstanding, it looks like international
prosecutions are to begin.
The Sudanese government has done nothing to materially combat ethnic
oppression or disarm the janjaweed; in fact, it has exacerbated the violence. Yet
this same military government, in response to the prospect of judicialization
through the ICC, has implemented its own set of legal proceedings through
a court system that it controls. Sudanese courts have convicted and sentenced
some members of the Sudanese army for “waging war” in Darfur, and others
for torturing and killing civilians.50 These Sudanese proceedings would not be
entitled to quali¬ed deference. Atrocities continue in Darfur under the aus-
pices of the very government that conducts the trials. This certainly does not
demonstrate any basis to impute any good faith to the Sudanese trial process.
The government™s good faith is compromised by the reality that it continues to
be involved in the atrocity that it is purportedly punishing.
The Pashtunwali is customary law in the Pashtun region of Afghanistan.51
It presents a second case study of a local legal system that would not meet
the quali¬ed deference guidelines. The Pashtunwali is not formal state law. As
such, it does not by de¬nition apply to all cases of extraordinary international
criminality in Afghanistan. In fact, the preference of the Afghan government
is to process such cases through the formalized state court system.52 However,
the Pashtunwali remains in¬‚uential in the rural areas of Afghanistan. Many of
these areas lie outside of central control and, at the time of writing, are the
dominion of warlords who share an af¬nity with the Pashtunwali. Attempts by
central authorities to supersede the Pashtunwali have been unsuccessful.
The Pashtunwali is a complex code of conduct that regulates diverse areas
of private and public life, including process and remedies for the in¬‚iction of
violence and serious human rights abuses. For these situations, it propounds a
restorative justice approach in which the family of the human rights abuser is
called upon to make restitution to the family of the abused. Restitution takes
the form of, but is not limited to, transfer of money and livestock. So long as one
of the sanctions contemplated by the Pashtunwali (even if only in extremis) “
namely, the transfer of young girls or women from the family of the human
rights abuser to the family of the abused in order to restore the harm “ remained
operative, the Pashtunwali would not be entitled to quali¬ed deference.53 This
is because sanction would impose a new great evil, namely sexual violence
and terror, on uninvolved third-party children. There is a second reason why
the Pashtunwali would lose its presumptive entitlement to quali¬ed deference.
It lacks democratic legitimacy. The Pashtunwali emerges from the diktat of
patriarchal elites who serve as nonrepresentative religious or military leaders. It
is not a consensual project.54 Who exactly gets to participate in the determination
of local sanctions and processes is of great importance to any assessment of the
democratic or social legitimacy thereof.
From Law to Justice 193

The Iraqi High Tribunal (IHT) would not be entitled to quali¬ed deference
because of the speci¬c characteristics of the security situation prevalent in Iraq
at the time of writing.55 The choice to prosecute Saddam Hussein (and other
defendants) and to showcase these trials as instruments for transitional justice
initially was made under the false belief that foreign troops and an Iraqi police
force would be able to maintain order. Tragically, Iraq is wrought with pervasive
insecurity. There are daily reports of bombings and murder. The IHT itself
has been plagued by violence. Since the proceedings opened on October 19,
2005, several individuals associated with the IHT (including, thus far, a judge
and three defense lawyers) have been assassinated; another defense lawyer was
seriously injured in an ambush.56 Personnel have received death threats: some
have ¬‚ed the country.57
Any accountability process must reasonably guarantee the safety of its partic-
ipants and its audience. One that cannot do so cannot pass muster. Although it
may well be de¬‚ating for those, such as I, who prefer in situ accountability meth-
ods, perhaps the time has come for a serious discussion of moving trials out of
Iraq. Although exceptional, there are times where it bene¬ts roiled societies for
accountability to be pursued elsewhere.58 This is the case if in situ trials cannot
proceed securely or if they induce signi¬cant insecurity. That said, sometimes
it makes sense to defer the pursuit of justice, instead of ousting it from local
hands, until a certain level of stability can be established.
In contrast to these three in situ modalities, neotraditional gacaca as imple-
mented for genocide through the 2004 Organic Law in Rwanda would not upend
the quali¬ed deference presumption. It underachieves its restorative, reintegra-
tive, and reconciliatory potential, but this alone is not a basis to vitiate the
deference entitlement. Although I am deeply concerned about the Rwandan
government™s use of gacaca to consolidate power “ and evidence that gacaca is
manipulated to serve political ends “ I do not believe that this evidence yet rises
to the level of lack of good faith, akin to that demonstrated by the Sudanese
government, that would oust the presumption in favor of quali¬ed deference.
That said, once gacaca panels engage their operations in earnest, the extent to
which the process practically serves political ends and shields the RPF from
allegations of its own wrongdoing (for example, war crimes, which currently
are excluded from the panels™ jurisdiction) would chip away at the deference
presumption. Unlike the case with the Pashtunwali, gacaca™s sanctions do not
include the in¬‚iction of great evils on uninvolved third parties. Gacaca sanc-
tions are not gratuitously iterated. Another key distinction from the Pashtunwali
is that gacaca for genocide, although not as democratic as it could be, retains
central markers of public participation. Judges are elected by the public. Judges
and other decision-makers are not barred from their work on the basis of gender
or religion, although they can be barred if they are suspected to be g´ nocidaires.
All community members can speak at sessions. Admittedly, the gacaca process
cannot guarantee the security of all of its participants. There have been murders
of witnesses. Some of these cases have been prosecuted. In Iraq, on the other
194 Atrocity, Punishment, and International Law

hand, there is poor accountability for the egregious sectarian violence commit-
ted today, including that which targets the IHT. With thousands of gacaca panels
set up throughout Rwanda and tens of thousands of defendants implicated in
the process, it may simply be inevitable that some con¬‚ict develops.59 That said,
Rwanda is not in a situation of pervasive instability akin to that in Iraq.

(iv) horizontal dimensions: obligation
in multiple orders
Cosmopolitan pluralism grounds a horizontal outreach beyond the criminal
law to other dimensions of law and to extrajudicial regulation. The goal of
this proposed outreach is to acknowledge the group-based nature of atrocity, a
task for which criminal trials are not well suited. Were the project of interna-
tional justice to horizontally integrate a broader swath of regulatory mechanisms,
it would become more responsive to group dynamics. To the extent that cos-
mopolitan pluralism favors this capaciousness, it has much to offer as an ordering
I begin by proposing an integration of the law of obligation, which includes
areas of law such as tort, contract, and restitution. This integration spawns an
overarching law of atrocity. In some jurisdictions, these dimensions of law remain
judicialized law in that they involve civil awards that are meted out by judicial
actors to private parties. In these cases, obligation tends to incorporate basic
tenets of liberalism in that it is predicated on individual agency, proximate
causation, and adherence to adversarial modalities of proof, including cross-
examination, that formally occur in a courtroom. Judicialized approaches to
responsibility are subject to a similar cultural and ideological contingency for
which I fault international criminal law. These approaches cannot be assumed
to be universal exemplars of legal responsibility and blindly superimposed upon
cultural contexts that may conceptualize responsibility differently. In other
jurisdictions, for example, obligation is not judicialized or privatized but is
articulated through communal social institutions. In my opinion, an indige-
nous liability scheme would retain its entitlement to quali¬ed deference even
if it did not emulate liberal approaches to fault and liability; quali¬ed defer-
ence, however, would operate to diminish the hazards of abusive communitarian
When implemented horizontally, cosmopolitan pluralism encourages res-
ponses to mass atrocity to attach to law as a whole and not just parochially to one
small subset of law, namely international criminal law. However, development
of a law of atrocity that captures both judicialized and nonjudicialized process
is only the ¬rst step in horizontal expansion. The second step takes the form of
outreach to quasilegal or fully extralegal mechanisms such as truth commissions,
legislative reparations, public inquiries, and the politics of commemoration.
This outreach pushes the enterprise of atrocity law toward the holistic promotion
of justice for atrocity. Just as accountability for extraordinary international crimes
From Law to Justice 195

can be enhanced by the richness and connectivity of local process, so, too, can
it be enhanced by the richness and connectivity of alternate disciplines.
Private law and extrajudicial mechanisms already form part of the practice
of states in response to atrocity. At national levels, for example, extrajudicial
mechanisms are quite commonly invoked in the aftermath of atrocity. However,
the internationalized paradigm generally views these mechanisms as separate
from, subaltern to, and in competition with criminal trials. Complementarity
has poorly thought out how the potential of extrajudicial initiatives could be
harnessed. Instead, complementarity creates incentives that may squeeze such
initiatives out of the justice matrix. Those international instruments that rec-
ognize reparations to victims view their importance as adjunct to international
criminal law.60 Tellingly, while universal criminal jurisdiction for genocide is
well established, the notion of universal civil jurisdiction is not.61
Although international law historically viewed responsibility in terms of the
state, the emergence of the paradigm of individual criminal responsibility has
put pressure on the doctrine of state responsibility. The interface of international
criminal courts with international courts having civil liability powers, such as
the International Court of Justice (ICJ), remains uneasy in matters related to
accountability for the great evils.62 The ICJ was established in 1946 as the prin-
cipal judicial organ of the United Nations. It only has jurisdiction over states.
The ICJ™s rapidly expanding docket includes cases regarding boundary disputes,
treaty interpretation, and the responsibility of states for international wrongs.
Victim states have ¬led claims with the ICJ that allege that other states bear
legal responsibility for serious violations of international humanitarian law and,
even, genocide. These claims have led to some unease among international
lawmakers. Although the law of state responsibility traditionally provides that
breaches of international law trigger a duty to provide reparations, applying
this general duty to the speci¬c instance of the crime of genocide is proving
particularly contentious.
The sidelining of obligation63 within the of¬cial project of international jus-
tice reduces the project™s diversity. The project™s ability to coordinate and tap
the full power of law and regulation to respond to atrocity is thereby impeded. A
cosmopolitan pluralist vision would encourage the dominant model, in which
deviance-based criminal law is transplanted to situations of collective cataclysm,
to release this preference and welcome the myriad ways in which law and politics
can capture atrocity™s group-based nature.
Tort, contract, and restitution implicate involved masses more effectively
by permitting more carefully calibrated measurements of degrees of responsi-
bility beyond the scarlet letter of guilt. These alternate sources of regulation
offer a more textured understanding of the key roles played by many oth-
erwise neglected actors. Tort, contract, and restitution can promote different
goals “ such as restoration, reconciliation, and reparation “ which may be
more capable of actualization. Restorative justice conceptions of accountability,
downplayed by international criminal law, could serve important purposes of
196 Atrocity, Punishment, and International Law

reintegrative shaming that resonate with the transitional needs of many postcon-
¬‚ict societies.64
Tort permits declaratory or monetary relief for violations of state responsibil-
ity and potentially for group liability outside the con¬nes of the state. This relief
might provide an additional layer of justice insofar as the criminal law does not
reach the state as an actor. Tort also permits strict liability, which, in theory,
could monitor groups that know they are about to do something dangerous and
incentivize them to establish proactive duties to intervene. Consideration could
be given to how contract law might impose ex ante conditions on in¬‚uential
of¬ces in politics, media, and the clergy in which the occupancy of such of¬ces
hinged on an of¬ceholder™s refusal to disseminate hate and, in a situation where
others preach hate, a commitment to engage in countervailing action. Con-
tract law also could mandate action on the part of international organizations,
peacekeepers, and private transnational actors. The law of restitution could inte-
grate private reparations well suited for situations where much of the violence
is committed locally by perpetrators known to victims and by neighbors upon
neighbors. Looking beyond, a pluralized law of restitution that avoids liberalist
standing requirements could facilitate group recovery and oblige the disgorging
of the bene¬ts of group violence. In the past, international law has experienced
tensions between liberal approaches to standing and recovery, on the one hand,
and the expectations of victims seeking relief, on the other.65
Going further, I propose a broader integration of extrajudicial and extralegal
modalities such as truth commissions, legislated reparations, public inquiries,
lustration, the politics of commemoration, redistributing wealth, and fostering
constitutional guarantees that structurally curb the concentration of power. The
impulse to broaden the response to mass atrocity must extend beyond legal
proceedings. This impulse should welcome communal sociolegal institutions,
in particular indigenous institutions, and thereby expand the template of pol-
icy options. At present, the interface between these institutions and criminal
tribunals is poorly thought out. It is marked by squabbling and controversy.
Institutions feel pressure to judicialize in order to be deemed complementary
or to receive funding. Instead, a more nurturing interface could be developed.
The de jure or de facto primacy of criminal courts over reconstructive efforts
may not re¬‚ect what the societies under reconstruction actually want. Victims
seek diverse remedies.66 For victims, justice does not singularly involve the
incarceration of a handful of offenders in distant, and often comfortable, pris-
ons. A cosmopolitan pluralist paradigm might better coordinate victim prefer-
ences interinstitutionally. Contemporary international criminal tribunals exer-
cise jurisdiction without being comprehensively linked to other entities central to
postcon¬‚ict reconstruction, in particular institutions and interventions that pro-
mote economic development and constitutional stability. After all, what would
the Nuremberg prosecutions have amounted to without the Marshall Plan and
the Basic Law?
Extant international criminal law institutions resist horizontal expansion for
a number of reasons. One is territorial. The ¬eld, and the sophisticated experts
From Law to Justice 197

well versed in it, has acquired of¬cial leadership as the preferred and pragmatic
response to atrocity. It is asking much of international criminal tribunals to
cede (or even share) that in¬‚uence. Another reason is ideological. International
criminal lawyers often equate civil and political remedies with collective forms
of accountability. For many international criminal lawyers, collective forms of
accountability raise the specter of the collective guilt of an entire nation or eth-
nicity. I believe that it is unfair to raise this specter. Although pursuing obligation
through civil and political remedies does capture a broader number of individ-
uals involved in atrocity, and hence moves toward collectivization, it does not
invariably lead to collective guilt. Guilt, among many other things, is a liberal
criminal law concept. It is what is imputed to those who are convicted pursuant
to a criminal trial and subsequently incarcerated (or, in outlier cases, executed).
Equating civil liability or restorative community service with collective guilt is
overreach. In theory, it could be possible to determine an entire nation crimi-
nally guilty and incarcerate each member of the nation.67 This is decidedly not
what my proposal for horizontal diversi¬cation contemplates. Instead, it aims to
move us away from the current dependence on criminal law.
Rather, my proposal turns to the fundamentally different notion of collec-
tive responsibility. There is a yawning gap between guilt and responsibility.68
Whereas many individuals are responsible for atrocity, a much smaller number
are criminally guilty. A much larger number of individuals are responsible than
can (and deserve to) be captured by criminal trials. Civil liability implicates those
individuals and institutions found to bear some responsibility for discrimination-
based mass atrocity. This can be a large group, hence the recourse to the phrase
collective responsibility. We would do the project of international criminal jus-
tice a disservice if, in implementing international criminal law, out of unfounded
fear of imposing collective guilt we marginalized or sneered down modalities of
accountability that promoted the collective responsibility of groups.
Collective responsibility understandably makes many observers nervous.
After all, civil liability, community service, and, especially, public reintegrative
shaming are powerful measures. Their use as collective sanctions for collective
responsibility merits extremely careful analysis. Probably the most evident start-
ing point for this analysis is to explore how, exactly, to de¬ne the responsible
group. I propose that the responsible group can be de¬ned either crudely or
The crude way structures the responsible group along its most evident charac-
teristics or combinations thereof: for example, nationality, ethnicity, inhabited
territory, or religion. So, for example, it renders all Germans as blameworthy for
the Holocaust; all Arab Sudanese for Darfur. The crude way assigns responsibil-
ity to the group in whose name atrocity was undertaken independently from the
actions of its individual members. The crude way therefore can include individu-
als who are not personally responsible. It can even include individuals who were
incompetent, or unable to do anything, so long as the atrocity was committed
in their collective name. In cases where atrocity is committed at the behest of a
state, the crude way includes within the group all those individuals living within
198 Atrocity, Punishment, and International Law

the jurisdiction of that state. This can have particularly harsh consequences.
Levying sanctions against a collectivity when that collectivity contains both per-
petrators and victims would hinder the victims™ recovery efforts. In sum, the
crude way does not limit the group to the aggregate of those individuals whose
action or inaction culminated in atrocity. Damages ultimately pass through to
and are borne by all group members, regardless of how bravely they resisted, how
servilely they complied, how eagerly they killed, or how much hurt they suffered.
At ¬rst blush, the crude way of group designation is anathema to liberalists.69
The careful way, on the other hand, pays attention to individual agency. It
limits the group to those individuals who, by virtue of their action or inaction,
are demonstrably responsible for atrocity. The careful way requires a more ¬ne-
grained analysis. The careful way thereby abides by Western legalist assumptions
of causation and individual agency. The careful way can be so careful that it can
appear in criminal prosecutions, albeit controversially: for example, doctrines
such as joint criminal enterprise that implicate very small groups acting in
concert.70 In cases of civil responsibility, however, the careful way still conditions
group membership on some sort of demonstrable linkage between action (or
nonfeasance) and the great evil. Individuals or entities for whom no connection
can af¬rmatively be delineated would avoid membership in the sanctioned
group. This renders the careful way dependent on similar modalities that limit
the effectiveness and universality of criminal trials.
The claim ¬led in 1993 by Bosnia and Herzegovina against Serbia and
Montenegro at the ICJ presents a case study of collective responsibility, and
a template for the directions in which collective responsibility could hypo-
thetically head.71 In its claim, Bosnia and Herzegovina asserts that Serbia and
Montenegro, the state into which the FRY was transformed in 2003,72 violated its
obligations to prevent and punish genocide under the Genocide Convention.
Bosnia and Herzegovina asserts that these violations constitute wrongful acts
attributable to Serbia and Montenegro that entail its international responsibil-
ity. Remedies sought include the payment of compensation for damages and
losses. Oral arguments closed in May 2006. The ICJ has reserved judgment.
Regardless of how the ICJ eventually disposes of this matter (a decision may well
be handed down while this book is in production), Bosnia and Herzegovina™s
genocide claim can stimulate a discussion, which transcends the actual jurisdic-
tion of the ICJ, regarding what crude or careful group designation might look
like in a cosmopolitan pluralist accountability framework.
Bosnia and Herzegovina™s claim “ along with a similar one brought against
Serbia and Montenegro by Croatia73 “ has caused jitters in the international law
community. These jitters are understandable. After all, a damages award against
the state of Serbia and Montenegro could run in the billions of dollars. Such
an award might be paid for by the tax contributions of all individual citizens of
Serbia and Montenegro. Alternately, it might be paid for by withholding foreign
aid, which means that projects intended for the bene¬t of all individual citizens
become halted or decelerated. A stinging declaratory award against the state
would trickle down and, by association, mark all state citizens. These citizens
From Law to Justice 199

could become international pariahs. Take Serbia and Montenegro™s World Cup
soccer team. No longer could the players say, “You™ve got those responsible:
they™re the guilty ones in the ICTY dock,” and then keep on playing with the
expectation of full reciprocal respect from the rest of humanity. Instead, the
players, too, would be responsible, in their own small way, and have to pay for
it in cash, reputation, or both.
The practical effects of collective responsibility frameworks are unsettling.
Signs of discomfort with collective responsibility had already surfaced in 1996,
when the ICJ initially found jurisdiction over Bosnia and Herzegovina™s claim
and dismissed preliminary objections thereto.74 One of the preliminary objec-
tions was that the allegations of state responsibility brought by Bosnia and
Herzegovina simply fell outside the scope of the Genocide Convention. The
majority of ICJ judges dismissed this preliminary objection. This group tersely
found that the plain language of the Convention, in particular the compro-
missory clause, did not exclude any form of state responsibility.75 Four judges
disagreed. Two of these four judges “ Judges Shi and Vereshchetin “ ultimately
ruled that the ICJ had jurisdiction over the Bosnian claim. However, they
appended a separate declaration to the ICJ™s judgment. In this declaration, they
expressed their “disquiet” with the holding that the Genocide Convention does
not exclude state responsibility. They reasoned that the Genocide Convention:
is essentially and primarily directed towards the punishment of persons com-
mitting genocide or genocidal acts and the prevention of the commission of
such crimes by individuals. [ . . . ] In substance, the Convention remains an
instrument relating to the criminal responsibility of individuals. The deter-
mination of the international community to bring individual perpetrators of
genocidal acts to justice [ . . . ] points to the most appropriate course of action.76

Judges Shi and Vereshchetin referenced the perniciousness of collective guilt
as a basis for the preference for individualized criminal prosecutions. They
explicitly cited the ICTY as having been created for the prosecution of persons
responsible for serious violations of humanitarian law committed in the terri-
tory of the former Yugoslavia. The preference for the ICTY as the appropriate
forum for judicialization works to the detriment of other fora. In a statement
whose pronounced hedging belies its normativity, Judges Shi and Vereshchetin
concluded that “it might be argued that [the ICJ] is perhaps not the proper
venue for the adjudication of the complaints [ . . . ] raised.”77 So, although both
judges ultimately concluded that the compromissory clause of the Genocide
Convention afforded a jurisdictional basis for Bosnia and Herzegovina™s claim,
they felt “obliged to express [their] concern” over this aspect of the case and,
thereby, sent a strong signal regarding how the merits of this aspect of the case
ultimately ought to be decided. These concerns, although not derailing Bosnia
and Herzegovina™s case at the preliminary stage, inevitably shall weave into the
resolution of the substantive claim and the kind of state responsibility the ICJ
believes the Genocide Convention actually creates. These concerns therefore
form a leitmotiv. Obliquely, they may even have informed the reasons why a
200 Atrocity, Punishment, and International Law

majority of the ICJ dismissed (upon preliminary objection and for jurisdictional
grounds) Serbia and Montenegro™s claim that, when NATO countries bombed
the FRY in 1999, they violated their international obligations.
It is obvious that a larger number of citizens of Serbia and Montenegro than
those actually indicted by the ICTY, or by national prosecutors, were responsible
for genocide at Srebrenica. An even greater number acquiesced in the carrying
out of atrocity generally in the name of Serb nationalism. All such complicit
and bene¬ting bystanders bear some responsibility. So long as those responsible
foresee that they shall never face any sanction, the law does nothing to dissuade
them from acquiescing in atrocity in its inchoate stages, when the enterprise
of atrocity is most vulnerable. Accordingly, I do not believe it is effective for
international criminal responsibility to entirely displace state responsibility in
matters of the great evils.
On the other hand, not all citizens of Serbia and Montenegro were respon-
sible for genocide at Srebrenica or atrocity in general. Moreover, equating the
group with the state does a particularly great disservice to those citizens of Serbia
and Montenegro who resisted genocide, spoke out against it, or were themselves
harmed by it. To this end, a judgment in favor of Bosnia and Herzegovina
would punish those who individually bear no responsibility. Looking at it from
the Bosnian perspective, a successful claim would inure to the bene¬t of all
its citizens. Included among the bene¬ciaries are individuals who were vic-
timized, those who were not, and those who were complicit in the violence or
actually committed it. Similar tracking problems abound. Although Srebrenica™s
Bosnian Muslims were the target of genocide, not each of Srebrenica™s Bosnian
Muslims equally was a victim. European Jewry was the target of the Holocaust,
but some European Jews served as kapos (guards) and in¬‚icted great brutal-
ity upon Jewish detainees. In the end, if Bosnia and Herzegovina™s claim were
successful, it would restitute some individuals while unjustly enriching others.
One poignant paradox is that “[w]ithin Bosnia, the Republika Srpska is actively
opposed to the Bosnian lawsuit and does everything in its power to obstruct it.”78
The ICJ, hampered by its own jurisdictional limits, therefore faces a dif¬cult
choice. If it ¬nds in favor of Bosnia and Herzegovina, it avoids impunity for the
many Serbs responsible and ensures some reparation for the many Bosnian vic-
tims; but it also sanctions nonresponsible Serbs and enriches Bosnian Muslims
with dirty hands.
One response is to alter the incentives to reduce the starkness of the trade-off.
This re¬‚ects implementation of the careful way. Currently, implementation of
the careful way is manifestly problematic for the ICJ insofar as its jurisdiction
is limited to states, not individuals, associations, or governmental departments
or subunits. That said, assuming hypothetically (1) that the ICJ were able to
engage broader forms of civil liability and jurisdiction and (2) that Bosnia and
Herzegovina were successful in establishing the substantive elements of its claim
(namely, infringement of the Genocide Convention),79 then the ICJ™s conjec-
tural pursuit of the careful way would, were it so empowered, allow it to make
¬ne-grained assessments of the group that is actually responsible (and, separately,
of the group entitled to damages). In such a scenario, civil damages or declaratory
From Law to Justice 201

denunciation would perhaps only be awarded against individual members of
the state government. Or against members of associations with control over the
apparatus of the state. Or against corporate entities who funded and equipped
genocide. But the careful way would obstruct liability from trickling down to all
citizens of the state.
A second response is to defend the collective responsibility of the entire popu-
lation of Serbia and Montenegro. This is the crude way. Assuming once again that
Bosnia and Herzegovina were successful on its substantive claim, an ICJ adher-
ing to the crude way would enter damages and denunciatory declarations against
the state, as it is empowered to do, and no eyebrows would be raised regarding
a trickle-down effect to every individual citizen of Serbia and Montenegro. The
details of who did what, and which Bosnian receives what, would become irrel-
evant in the name of a greater good. Sanctioners minded toward the crude
way also might invoke proxies to associate the actions of the Serbian state with
the Serbian people. One possible proxy is the electoral/political behavior of the
majority, or a plurality, of the population. Thomas Franck proposes that: “Where
the people whose leaders are committing international crimes have the option
to vote them out, or to stage mass protests against those acts, and fail to do so
it is not only morally appropriate but good social policy that they be made to
assume a degree of responsibility for the things done in their name.”80
Whereas the careful response may soothe some liberal sensibilities, the crude
position as applied to atrocity is bound to be controversial “ at least to a Western
audience. Of course, some of this controversy amounts to posturing. After all,
collective sanctions that capture the nonresponsible are found with some fre-
quency within the domestic law of Western states.81 Collective sanctions that
capture the innocent exist, often to a greater degree, in other sociolegal orders
as well.82
Bosnia and Herzegovina™s ICJ claim re¬‚ects the inevitability that interna-
tional criminal lawyers will have to juristically deal with collective responsibility.
Even if Bosnia™s claim were to be dismissed on all fronts, it will not be the last
kind of collective responsibility claim ever brought. In fact, the ICJ issued a col-
lective responsibility award in another case, DRC v. Uganda, which primarily
involved use of force matters, although also implicated violations of interna-
tional criminal and humanitarian law (albeit not genocide).83 The ICJ ordered
Uganda to pay reparations in light of its responsibility for armed activity, plun-
dering, and massive human rights violations in the DRC. Speci¬cally, the ICJ
located Uganda™s responsibility for inter alia killing, torture, training of child sol-
diers, incitement of ethnic con¬‚ict, and other forms of inhumane treatment.84
The pursuit of collective responsibility is a strong preference among victims,
who exhibit great cleverness in attempting to articulate this pursuit through
juristic channels. The inevitability of assessing the place of collective responsi-
bility within the project of international justice, however, should be a cause for
contemplation and optimism, not embarrassment or annoyance.
By turning collective responsibility into a bˆ te noire, skeptics properly remind
us of dark days: for example, the results of the Treaty of Versailles, which imposed
“victor™s justice” reparations on an entire nation and arguably promoted cycles
202 Atrocity, Punishment, and International Law

of further violence. On the other hand, as I have argued elsewhere, I believe
that international criminal lawyers™ fears of collective responsibility have inhib-
ited dispassionate conversations about its potential in thwarting atrocity and ret-
rospectively promoting justice.85 Any realistic analysis of discrimination-based
mass atrocity teaches us that the violence is deeply collective in nature and,
what is more, that its collective nature surpasses the aggregate of all individual
action. Just as it is counterproductive to downplay the role of the collective as
a factor that induces the impulse toward atrocity, it also is counterproductive to
eschew considering how collective remedies can confront collective action. I
believe there is value in debating collective responsibility for reasons rooted in
communitarian moral theory, as developed by George Fletcher,86 and also, as
noted by Larry May,87 for reconciliation. Let me advance a third reason, which
I ¬nd more compelling: utilitarianism.
Many atrocities begin with the devious kindling of con¬‚ict entrepreneurs,
who seek to in¬‚ame and exacerbate communal tensions. Community responses
to this kindling are not predestined. How the community responds is the central
determinant regarding whether violence subsequently erupts and, if so, of its
amplitude. If community members ignore these ¬‚ames, and look past attempts to
habituate them into violence and hatred, then the con¬‚ict entrepreneur remains
marginal. If community members are attracted to the ¬‚ames, and identify with
violence and hatred, then the wheels of atrocity are set in motion. And once set
in motion, these wheels quickly become unstoppable by anything other than
the use of countervailing force.
I argued in Chapter 6 that criminal punishment goes some way to developing
expressive values that edify a moral consensus regarding the manifest illegality of
discrimination-based violence. This consensus might serve as a bulwark against
exhortations by con¬‚ict entrepreneurs in favor of such violence. I also argued
in Chapter 6 that the threat of criminal punishment will not deter committed
individuals acculturated into hatred from implementing their own ¬nal solu-
tions. Criminal law does little to deter eliminationist killers. However, let me
suggest here that collective responsibility might go some way ex ante, in a util-
itarian sense, to diminish the mainstreaming of con¬‚ict entrepreneurship and
the festering of cultures of hatred. If normalized, collective responsibility could
augment the likelihood of sanction for a much broader number of individuals.
Collective responsibility could reach the catalytic group of bene¬ting bystanders
I identi¬ed in Chapter 2.
Group members are in an advantageous position to identify, monitor, and
quash the behavior of con¬‚ict entrepreneurs before it metastasizes. Because
the criminal law paradigm does not reach group members, it provides them no
incentive to cabin or control the behavior of con¬‚ict entrepreneurs. Collective
responsibility might do more to encourage group members to control con¬‚ict
entrepreneurs early on, and hence serve as gatekeepers, because they would
be called to task afterwards. A collective responsibility paradigm could thereby
serve a monitoring function. Group members would, as Mark Osiel suggests,
begin to police each other™s activities and responses.88 The threat of collective
From Law to Justice 203

sanctions may activate group members to marginalize the conduct of con¬‚ict
entrepreneurs or, in the best-case scenario, snuff it out. Cosmopolitan pluralism
would encourage an interface with collective responsibility mechanisms that,
in turn, could go some way to plugging an important gap left by criminal trials.
Given that passive acquiescence rarely “ if ever “ is implicated by a system
based on individualized criminal law, it is unclear how this system can deter this
fundamental prerequisite to mass atrocity.
States have duties to their citizens; to which I would add that citizens also
have certain duties to the state. One of these is a duty to prevent the state from
actualizing extraordinary international crimes. This duty becomes all the more
onerous to the extent that citizens have input into political decisionmaking.
Citizens should be put on notice that they cannot stand by while hatemongering
becomes normalized.
Collective responsibility frameworks can implicate bene¬ting bystanders.
These frameworks can thereby af¬x a cost to an individual™s drawing the blinds,
receiving a promotion at work because the “other” got ¬red, moving into a
suddenly vacated apartment, and acquiescing in the hijacking of the state by
extremists. It is well-nigh impossible to deter a suicide bomber or crazed ideo-
logue. Once an individual has passed a threshold of habituation in or affection
for violence, has deeply imbibed hatred, or needs to kill to survive, the law can
offer little in deterrence. However, the law may more plausibly reach the much
larger group of people that passively allow the con¬‚ict entrepreneur to assume
of¬ce, procure weapons, and build a power base of habituated killers. Any struc-
ture that incentivizes the masses to root out the con¬‚ict entrepreneur before that
individual can indoctrinate and brainwash will diminish the depth of perpetra-
tor moral disengagement that is a condition precedent to mass atrocity. Such a
structure thereby inhibits early on, when inhibition still remains possible, the
“escalating commitments” that psychologist James Waller believes demarcate
the “road to extraordinary evil.”89 The social death of the victims “ a precondi-
tion to their actual deaths “ may thereby be impeded. Capturing all individuals


. 7
( 11)