. 3
( 11)


oner™s demonstration of rehabilitation, as well as any substantial cooperation
by the prisoner with the ICTY Prosecutor. The ICTY Practice Direction also
mentions as factors the behavior of the convict during the period of incarcera-
tion, conditions of incarceration, and the results of psychiatric or psychological
The ICTR has not yet granted early release. It may begin to do so should
ICTR convicts serve sentence in states whose ordinary domestic criminal law
contemplates early release (Mali™s does not guarantee such a bene¬t).56 The
ICTY has granted early release quite actively, although it has not granted every
application.57 Anto Furundˇ ija is one convict for whom early release has been
granted. Furundˇ ija had been charged with war crimes arising out of his interro-
gation of a civilian and a soldier, and his presence while both were being beaten
and the civilian was raped. He was convicted in 1998 of co-perpetrating torture
and of aiding and abetting outrages upon personal dignity, including rape, and
sentenced to ten years™ imprisonment. President Judge Meron based his deci-
sion to release Furundˇ ija early, in August 2004, on a number of key pieces of
evidence. These included: (1) the contents of a con¬dential memorandum from
the ICTY Registry; (2) “a letter from the Minister of Justice of Finland and [ . . . ]
report from the Chief Of¬cer of the Kylm¨ koski prison, where Mr. Furundˇ ija
a z
served the majority of his sentence, [ . . . informing . . . ] that Mr. Furundˇ ija has
behaved impeccably during his imprisonment; has been employed both inside
and outside the prison; has been very cooperative in his relationship with the
prison staff; and has maintained exceptional discipline and correct behaviour”;
(3) a psychological assessment, prepared by the Finnish prison authorities, noting
no impediment to Furundˇ ija™s release; and (4) an internal memorandum con-
cerning Furundˇ ija™s cooperation with the Of¬ce of the Prosecutor.58 Applying
this evidence to the law, Judge Meron granted release based on his ¬ndings that
“as reported by the Finnish authorities, Mr. Furundˇ ija has accepted the judge-
ment he received as fair and has expressed remorse for the suffering of victims”;
Punishment in International Tribunals 55

that he “is resolved to be reintegrated into society, exhibited good behaviour
in detention, and has a strong attachment to his family”; that the evidence
“establishes the strong likelihood that Mr. Furundˇ ija will successfully reinte-
grate himself into the community upon release”; and that “Mr. Furundˇ ija™s z
case is no less appropriate for a grant of early release than that of other pris-
oners previously granted early release.”59 Early release was granted in this case
notwithstanding the report of the Of¬ce of the Prosecutor that Furundˇ ija has
not cooperated with it.
Although the prospect of early release may be anticipated at the time the
sentence initially is ¬xed, ICTY judges have ruled that this prospect should not
factor into the determination of the length of the sentence.60 In other words, it
is improper to increase the length of sentence to absorb the possibility of early

(ii) sentencing practice
As of May 2006, the ICTR and ICTY, when taken together, have issued nearly
eighty convictions. Some of these remain subject to appeal. In the discussion
that follows, I refer to sentences that remain subject to appeal as “un¬nalized
sentences” and those sentences upon which the Appeals Chamber has ren-
dered judgment or which the convict elected not to appeal as “¬nalized sent-
The East Timor Special Panels had convicted eighty-four individuals (arising
out of ¬fty-¬ve trials) before ceasing operations (after funding ran out) on May
20, 2005.61 This total represents only one-quarter of all individuals indicted for
serious crimes pertaining to atrocity in East Timor in 1999. Those who bear pri-
mary responsibility for this violence have yet to be held to account. Throughout
their short-lived operation, the Special Panels were hobbled by weak resources,
including a lack of proper translation/interpretation and research expertise; per-
sonal Internet access for the judges only became available in late 2001.62 At the
time of writing in 2006, over a year after the Special Panels had shut down,
con¬‚ict “ reportedly fueled by animosity between those who sympathized with
Indonesian rule and those who did not “ persists in East Timor. To this end,
the work of the Special Panels has not lanced the boil of violence “ if such an
ambitious goal ever were possible through recourse to the criminal law. The
East Timorese government has expressed reluctance to empanel new courts or
tribunals, although it has been more supportive of truth commissions and public
Jurisprudentially, the work of the ad hocs has been in¬‚uential. The Spe-
cial Panels incorporated ICTY and ICTR jurisprudence.63 It is quite likely that
the jurisprudence of the ad hocs shall guide the ICC and other institutions,
such as the Special Court for Sierra Leone and the Extraordinary Chambers in
the Courts of Cambodia, when these institutions begin to issue sentences. The
Kosovo hybrid panels also have issued a number of sentences: there is, how-
ever, considerable reticence on the part of the judges “ even the international
56 Atrocity, Punishment, and International Law

judges “ in Kosovo to refer to the work of other international criminal justice
At the ICTR, of the twenty-four individuals who have been convicted at
the time of writing (un¬nalized sentences), eleven have been sentenced to
life imprisonment (in certain cases, to multiple life sentences). The remaining
thirteen individuals have been sentenced to the following ¬xed terms: one to
forty-¬ve years, two to thirty-¬ve years, one to thirty years, one to twenty-seven
years, three to twenty-¬ve years, two to ¬fteen years, and the remaining three
to terms ranging from six to twelve years.65 The ICTR Appeals Chamber had
reduced one life sentence to a ¬xed term of forty-¬ve years because of, inter
alia, its proprio motu ¬nding of serious violations of the defendant™s fundamen-
tal rights during his arrest and detention.66 In another case, that of Laurent
Semanza, the Appeals Chamber increased a sentence from twenty-¬ve to thirty-
¬ve years.67 Nearly all of the ICTR™s convictions are for genocide or crimes
against humanity. There have been very few convictions for war crimes. Of
those ICTR defendants who have received ¬xed term sentences, including those
¬nalized on appeal and those not yet ¬nalized, the mean sentence is 23.5 years
and the median sentence is 25 years. However, as of May 2006, a number of
heavy term sentences remain subject to appeal, along with four life sentences.
This is why the mean and median ¬nalized term sentences are lower. The
ICTR has acquitted three defendants. One problem that has arisen is that some
acquitted individuals, who are Rwandan citizens, remain in detention insofar
as no country is willing to admit them into their territory. A similar situation
may arise for convicts who eventually are released once they have served their
As of May 2006, the ICTY has issued forty-nine ¬nal sentences; an additional
¬ve sentences remain in the appeals process.68 As of this juncture, all have been
term sentences (the one life sentence that had been issued was reduced on
appeal to a forty-year term).69 The ICTY™s ¬nalized sentences range from 2.5 to
40 years. Among the ¬nalized sentences, the mean sentence is 14.3 years and the
median sentence is 12 years. Among all sentences, the mean is 14.75 years and the
median is 13 years. The average length of sentences is slightly lower in 2006 than
it had been in 2002, when the mean ICTY term sentence was ¬fteen years and
the median term sentence sixteen years.70 A number of lengthy sentences issued
by the ICTY Trial Chambers have been reduced on appeal: the life sentence to
Dr. Staki´ ; and forty-six years to General Krsti´ and forty-¬ve years to General
c c
Blaˇki´ , which were reduced to thirty-¬ve and nine years, respectively.71 In 2006,
the Appeals Chamber reduced another heavy term sentence “ twenty-seven years
to Momir Nikoli´ following a plea bargain “ to twenty years when it found that
the Trial Chamber had committed a number of errors. On the other hand, the
Appeals Chamber af¬rmed Dario Kordi´ ™s sentence of twenty-¬ve years despite
allowing certain grounds of his appeal. As of the time of writing, the heavi-

est sentence that remains under appeal is Radoslav Brdanin™s thirty-two-year
Punishment in International Tribunals 57

table 3.1. Ad hoc tribunals “ all sentences (including those subject to appeal
as of May 2006)

No. of Life imp. Term Mean Median
sentences (%) sentence (%) term term
ICTR 24 11 (45.8%) 13 (54.2%) 23.5 yrs. 25 yrs.
ICTY 54 0 54 14.75 yrs. 13 yrs.

As Tables 3.1 and 3.2 indicate, there is a considerable disparity between ICTY
and ICTR sentencing practices. Even if all of the ICTR™s life sentences are
taken out, and comparison is made only between term sentences, the ICTR
sentences considerably more harshly. In Chapter 6, I explore whether the differ-
ence between ICTY and ICTR sentencing can be accounted for on a principled
basis. I consider three potential, and mutually inclusive, explanations for this
differential judicial behavior: incorporation of national law, the sheer gravity of
atrocity in Rwanda, and that a much larger proportion of ICTR convictions have
been for genocide.
Disparity between ICTY and ICTR sentences grows when account is taken
of one important limitation to the data as reported earlier. The data do not
re¬‚ect the ICTY™s practice of early release. Given that approximately 15 percent
of ICTY convicts have thus far bene¬ted from early release,72 and that the law-
in-practice of the ICTY is to grant such release after service of two-thirds of the
sentence, the data summarizing length of incarceration at the ICTY could be
reduced accordingly.
Table 3.3 presents the sentencing practice of the East Timor Special Panels.73
Although I report eighty-four individuals convicted, the Special Panels actually
issued eighty-¬ve convictions. One individual, Gilberto Fernandes, was con-
victed on two separate occasions for two different crimes.74 Because the Spe-
cial Panels had jurisdiction over serious ordinary crimes as well as extraordi-
nary international crimes, Table 3.3 separates the punishments reported for
serious ordinary crimes (i.e., murder under the Indonesian Penal Code) from
crimes explicitly identi¬ed in the case reports as extraordinary international
crimes (mostly crimes against humanity). Sixty of the eighty-¬ve convictions
were for extraordinary international crimes. In some cases, particularly judg-
ments stemming from indictments issued in the ¬rst year of the Special Panels™
operation, convictions were entered for serious ordinary murder in situations

table 3.2. Ad hoc tribunals “ ¬nalized sentences (as of May 2006)

No. of Life imp. Term Mean Median
sentences (%) sentence (%) term term
ICTR 16 7 (43.75%) 9 (56.25%) 20.9 yrs. 15 yrs.
ICTY 49 0 49 14.3 yrs. 12 yrs.
58 Atrocity, Punishment, and International Law

table 3.3. East Timor Special Panels

No. of Mean Median
sentences term term
Ordinary Serious Crimes 25 6.3 yrs. 5 yrs.
Extraordinary International Crimes 60 9.9 yrs. 8 yrs.

where the factual context involved the kind of violence (i.e., apparently com-
mitted as part of a widespread or systematic attack on a civilian popula-
tion) that could ostensibly qualify as extraordinary international criminality.
In subsequent years, certain other indictments were amended to proceed
on the basis of crimes against humanity. Convictions stemming from indict-
ments issued as of 2003 overwhelmingly were for extraordinary international
The Special Panels issued a broad range of terms of imprisonment: from 11
months to 15 years for ordinary crimes and from 2 to 33 1/3 years for extraor-
dinary international crimes. The Special Panels™ enabling instruments pre-
cluded them from awarding a life sentence. Mean and median sentences issued
by the Special Panels for international crimes are 9.9 and 8 years, respec-
tively. The sentences of the Special Panels, which in early years of opera-
tion were compatible with those of the ICTY (if not slightly longer), progres-
sively dipped well below ICTY levels. Paradoxically, as indictments increasingly
began to charge extraordinary international crimes, sentences grew shorter.
This decrease in length of sentence was in large part due to the frequent
plea bargaining of extraordinary international crimes, even though it is unclear
whether defendants who self-convicted actually were motivated by the prospect
of reduced terms of imprisonment. Also, sentences tended to be reduced on
The length of Special Panel sentences becomes even shorter when the effects
of conditional release and Presidential Decrees are considered. I did not include
these in the data set. As with the ICTY, the Special Panels permit a convict
to be conditionally released following service of two-thirds of the sentence and
satisfaction of other criteria. Approximately 10 percent of all convicts have, at the
time of writing, bene¬ted from conditional release. Unlike with the ICTY, the
Special Panels also permit sentence reduction by virtue of Presidential Decree.
Approximately 10 percent of convicts bene¬ted from Presidential Decrees issued
on May 20, 2005 (the ¬nal day of the Special Panels™ operation). Reductions
ranged from 9 months to over 8 years (the three longest cumulative sentences
issued by the Special Panels, namely 33 1/3 years, were reduced by Presidential
Decree to 25 years).
Judges, in particular those sitting on the ICTY and ICTR, have, as the ICTR
Trial Chamber held in Prosecutor v. Rutaganda, “unfettered discretion”75 to sen-
tence. Although they are limited by the positive law instruments with regard to
Punishment in International Tribunals 59

the type of punishment they can issue (imprisonment and restitution, although
the latter does not ¬gure at all in the law-in-practice of the ICTY or ICTR),
judges have been willing to utilize their “unfettered discretion to go beyond
the circumstances stated in the Statute and Rules to ensure justice in matters of
sentencing.”76 This self-delegation of authority has led, perhaps inexorably, to an
erratic quanti¬cation of sentence. There are no formalized sentencing guide-
lines, whether mandatory or advisory, for international judges who sentence
extraordinary international criminals. In fact, the ICTY Appeals Chamber has
emphasized the inappropriateness of setting down a de¬nitive list of sentencing
guidelines.77 Furthermore, the practice of fairly active appellate intervention
leads to additional discretion and unpredictability in the operation of the ICTY,
ICTR, and Special Panels. Trial judges have initial discretion to ¬x sentences
and, then, their appellate counterparts often revisit determinations made at trial
in a manner that is not clearly cabined or explicated. Insofar as there is no reg-
ular practice of count-by-count sentencing (instead, an overall sentence often
is given), in the event an Appeals Chamber overturns certain convictions the
process of determining exactly what the effect of those quashed convictions is
on the revised sentence becomes nebulous.
International judges are comfortable with their discretionary powers to ¬x
sentence within the traditional mode of incarceration notwithstanding the

concomitant lack of consistency in sentencing. In Delali´ (Celebi´ i), thec
ICTY Appeals Chamber nodded approvingly to the “considerable amount
of discretion” to fashion a sentence, commenting that this discretion stems
from the “over-riding obligation to individualise a penalty to ¬t the indi-
vidual circumstances of the accused and the gravity of the crime.”78 Inso-
far as genocide, crimes against humanity, and war crimes can be committed
“in a multitude of ways,” another advantage to this “almost limitless” dis-
cretion is that punishment can be individualized to “vastly differing levels
of culpability.”79 Recognition of judicial discretion in the ¬xing of sentences
remains a ¬rm point of reference in the jurisprudence of both the ICTY and
In the end, although individualizing the penalty certainly is desirable, the
bene¬ts thereof dissipate when there is no coherent framework in which to
predictably consider the factors germane to, or the goals of, sentencing.

(iii) penological justi¬cation and implementation:
the jurisprudence
Although they are not formally bound by stare decisis, judges “ in particular at the
ICTY and ICTR “ do refer to prior judgments (of their tribunal as well as others,
including national courts). For the most part, these references pertain to points
of law and factors to consider in sentencing. The references do not pertain to the
determination of the actual sentence. In fact, precedential guidance that may
¬‚ow from previous sentences issued by the ICTY and ICTR is “very limited” and
60 Atrocity, Punishment, and International Law

not a “proper avenue to challenge a Trial Chambers™ ¬nding in exercising its dis-
cretion to impose a sentence.”81 The ICTR Appeals Chamber held “ in its 2005
judgment in Prosecutor v. Semanza “ that “comparisons [to other cases] may be
of limited value given that each case has its own particular circumstances [ . . . ][.]
Ultimately, the decision as to the length of sentence is a discretionary one, turn-
ing on the circumstances of the case.”82 Judges demonstrate little willingness
to engage in meaningful comparative analysis even of similarly situated extraor-
dinary international criminals in determining the length of imprisonment. For
example, the ICTY Appeals Chamber intoned in Prosecutor v. Babi´ that, even
if a comparison were feasible owing to the substantial similarity between two
cases, there only would be grounds to intervene if the two sentences were so out
of reasonable proportion “so as to suggest capriciousness or excessiveness.”83 In
Prosecutor v. Staki´ , the ICTY Appeals Chamber went so far as so hold that com-
parisons with other cases “were inappropriate as the . . . case was of a ˜unique™
nature.”84 Although the upshot of this legal methodology is suppleness, which
may inure to the bene¬t of a relatively new area of law, it also risks a slapdash
approach geared to obtaining a desired result in each individual case instead of
a predictable and independent sentencing heuristic.
Despite their considerable discretion, international criminal judges do refer
to important theoretical principles in meting out sentences. These points of ref-
erence emanate largely from ordinary criminal law and include garden-variety
rationales for punishment such as deterrence and retribution; but also expres-
sivism, which can take on somewhat innovative meaning.85 The structure pro-
vided by theoretical principles helps explain why international criminal tribunals
punish more severely in some cases and less so in others. Insofar as the positive
law documents essentially are silent as to the penological purpose of the sen-
tences imposed, much of this structure has emerged from the jurisprudence of
the sentencing institutions. While the jurisprudence provides some direction,
however, it also is internally contradictory in terms of the goals of sentencing,
leading to a lack of predictability or coherence regarding the actual quantum
of sentence imposed in individual cases. Moreover, as I explore in Chapter
6, regardless of operational incoherence, retribution and deterrence are very
dif¬cult to operationalize in the context of mass atrocity through the tools of
punishment currently available.

a. Why Punish?
Retribution and general deterrence are the two most prominent punishment
rationales in international criminal law.86 Whereas retribution had been a
major motivating factor at Nuremberg,87 the general deterrence motivation
has acquired some traction in contemporary institutions.88 However, consider-
able indeterminacy and confusion persist. The ad hoc tribunals vacillate when
it comes to prioritizing the weight to accord to retribution and deterrence in
sentencing. For example, over the past ¬ve years the ICTY has issued judg-
ments that cite retribution and general deterrence as “equally important,”89
Punishment in International Tribunals 61

judgments that cite retribution as the “primary objective” and deterrence as a
“further hope,” warning deterrence “should not be given undue prominence,”90
and judgments that ¬‚atly state “deterrence is probably the most important factor
in the assessment of appropriate sentences.”91 A survey of all the cases of the ad
hoc tribunals over time, though, reveals a preference for retributive motivations,
especially when it comes to the aggravating and mitigating factors the tribunals
consider in ¬xing sentence.92 The case law of the East Timor Special Panels
demonstrates a similar stated preference.93
Although there are many divergent schools of retributivism, what all retribu-
tivists generally share is the understanding that the in¬‚iction of punishment rec-
ti¬es the moral balance insofar as punishment is what the perpetrator deserves.
Punishment, therefore, is to be proportionate to the nature and extent of the
crime.94 Although retribution is the most prominently cited rationale, it is also
one with which the international tribunals express the most nervousness. These
jitters come from concerns that punishment may be perceived as equating
revenge. Accordingly, the ICTY Appeals Chamber has emphasized that “retribu-
tion should not be misunderstood as a way of expressing revenge or vengeance.”95
Judges assume the undesirability of revenge as a response to extraordinary inter-
national criminality and predicate this assumption on the belief that quashing
revenge is a step in breaking the cycle of violence, maintaining the dignity
of those who in¬‚ict punishment, and the civilized nature of the punishing
The ICTY™s fear of looking vengeful has induced it, on occasion, to push
retribution in a new and contemplative direction, in which retribution is con-
structed as the “expression of condemnation and outrage of the international
community.”96 This understanding of retribution, which remains an outlier posi-
tion, diverges from the dominant narrative of retribution at the international
tribunals. Interestingly, this understanding moves retribution in the direction
of expressivism, which is a third, and currently subordinate, justi¬cation for
punishment. The expressivist punishes to strengthen faith in rule of law among
the general public, as opposed to punishing simply because the perpetrator
deserves it or will be deterred by it. From an expressivist perspective, punish-
ment proactively embeds the normative value of law within the community.97
Expressivism also transcends retribution and deterrence in claiming as a cen-
tral goal to build historical narratives and educate the public about these
General deterrence considers that the purpose of prosecuting and punishing
those who commit mass atrocity is to dissuade others from doing so in the future.
Speci¬c deterrence implies that punishing the offender will deter that offender
from reoffending in the future. Initially, international criminal tribunals ascribed
scarce importance to speci¬c deterrence.98 This skepticism has thawed in more
recent jurisprudence, particularly at the ICTY.99 However, when the activity
of international criminal justice institutions is taken as a whole, the focus of
deterrence remains oriented to general deterrence. In the case of East Timor,
the general deterrence concerns are vivid. In Prosecutor v. Beno, judges noted
62 Atrocity, Punishment, and International Law

that “there is an additional requirement for deterrence because just across a hard-
to-guard border live hundreds of recalcitrant ex-militia men with the capability
of once again destabilizing this country by means of murder.”100
From a deterrence perspective, punishment is in¬‚icted not because the
offender deserves it, but because of the consequentialist effect of punishment in
reducing recidivism. There are other consequentialist rationales. These include
rehabilitation, whose place within the practice of international sentencing
remains marginal (although, again, subject to palpably inconsistent treatment
among judgments and even within the same judgment).101 Insofar as child sol-
diers are concerned, rehabilitation is given considerable currency in the positive
law of the Special Court for Sierra Leone, although, gauging by the indictments
thus far issued, it does not appear that it will ¬gure much in the jurisprudence
of the institution when it begins to hand out sentences. Consequentialist ratio-
nales also include incapacitation.102 Although it is self-evident that, by isolating
hatemongers and inhibiting their dissemination of vitriol, international criminal
law may minimize con¬‚ict, incapacitation is generally not proffered as a central
goal of punishment.
Reconciliation and peace were identi¬ed by the Security Council as major
purposes of the ICTY and ICTR. Judges, however, have not ascribed them much
in¬‚uence. Assuredly, reconciliation and peace as consequential aspirations do
surface as penological goals in some of the judgments of the ad hoc tribunals,
in particular more recent judgments, but efforts to operationalize them in the
allocation of sentence remain incoherent. The 2005 Babi´ decision by the ICTY
Appeals Chamber is illustrative. Babi´ was the former Prime Minister in the
breakaway Krajina Serb republic after Croatia had declared independence in
1991. He pled guilty to a single count of persecutions as a crime against human-
ity. The Trial Chamber sentenced him to thirteen years™ imprisonment. Babi´ c
appealed his sentence. The Appeals Chamber found the Trial Chamber had
erred by giving insuf¬cient weight to his efforts in post hoc peace negotiations.
However, citing retributive concerns, the Appeals Chamber then refused to
reduce the sentence issued by the Trial Chamber in spite of the error.104 Babi´ c
committed suicide in 2006. In dissent, Judge Mumba would have reduced the
sentence.105 She noted that overturning the weight the Trial Chamber gave to
Babi´ ™s contributions to the restoration of peace but then refusing to opera-
tionalize these contributions in the actual quanti¬cation of sentence implies “
seemingly incorrectly “ their negligible value. I would extend the analysis beyond
this particular defendant. The in¬‚uence of peace and reconciliation as goals
of punishing extraordinary international criminals is unpredictable, perhaps
because it is such an ambitious goal. I say unpredictable insofar as another
high-pro¬le plea-bargained defendant had received operational discount in the
quantum of sentence for her post hoc peace-making efforts.106
Although their constitutive instruments mention restorative objectives, and
the ICC has taken af¬rmative steps in this regard, in their practice international
criminal tribunals still remain distant from victim-centered restorative modali-
ties that may correspond more closely to the expectations of local populations in
Punishment in International Tribunals 63

the places where atrocity has been and incipiently is being judicialized, including
by the ICC in Uganda and the Democratic Republic of Congo.107 The rationales
on which international criminal tribunals primarily ground their punishment,
namely retribution and deterrence, resonate more deeply within dominant crim-
inal justice systems than in systems, in particular local systems, in which atrocity
is becoming internationally judicialized. To be sure, nuggets of retributivism
and deterrence exist in virtually all criminal justice systems (whether secular
or religious; national or local; formal or informal). And nuggets of restorative
objectives are found in virtually all criminal justice systems. But the role that
restoration plays in dominant justice systems is less than what it plays in the jus-
tice systems of weaker states, where there remain vibrantly powerful local and
customary methods of dispute resolution in which restorative goals and methods
often occupy a primary place. I examine these methods in Chapters 4 and 5.
Many bottom-up transitional justice movements invoke sanctions such as
apologies, shaming, sharing the truth, lustration, and reparations; and often are
willing to procure these by offering amnesties to perpetrators.108 This is the
case even though such modalities are often at odds with, and largely squeezed
out by, the operation of the international criminal law paradigm. International
criminal law responds poorly to the preferences of local populations when such
preferences con¬‚ict with its normative worldview. This leaves local populations
with little recourse but to articulate these preferences outside of and at times in
resistance to top-down internationalist pressures and, thereby, expend consid-
erable resources and effort at a particularly vulnerable time in mediating with
international legal regimes. To be sure, given the frequency of truth commis-
sions and nonjudicialized approaches to postcon¬‚ict justice, local populations
do experience some success in this process of mediation, but this also invites
a much deeper inquiry whether and how more inclusive internationalist struc-
tures can be edi¬ed to channel these energies more constructively. Although
there are times when the international community declaratively can recognize
the role such initiatives might play, for example in its 2005 referral of the Darfur
situation to the ICC, these initiatives are at most given a role of adjunct or
additional complement to the ¬xture of liberal procedural legalism.109

b. What Factors to Consider in Punishing?
At the ad hoc tribunals, determinations of what can constitute aggravating or
mitigating factors “ as well as the weight to attach to these “ lie within the
discretion of the Trial Chamber.110 Retributive concerns dominate the factors
international criminal law institutions view as aggravating or mitigating in the
imposition of sentence. This is particularly the case with aggravating factors.
These factors mostly attach to the extent of the wrongdoer™s culpability, blame-
worthiness, immorality, and desert. In fact, when counsel for one defendant
urged the ICTY Appeals Chamber to reconsider a Trial Chamber sentence
based on a “trend in international law” away from retribution, the Appeals
Chamber sharply disagreed.111 The Appeals Chamber found this “alleged” trend
64 Atrocity, Punishment, and International Law

to be unsubstantiated and instead underscored the importance of retribution as
a general sentencing factor.112
Although the positive law of the ad hoc tribunals provides only one illustration
of a mitigating or aggravating circumstance,113 the jurisprudence develops many
more. The following aggravating circumstances arise in the jurisprudence:
r the gravity and egregiousness of the crimes, identi¬ed as the primary consid-
eration in imposing sentence;114
r the breadth of the crimes (e.g., numbers of victims)115 and the suffering
r the youth of the victims117 or their general vulnerability;118
r the nature of the perpetrator™s involvement (active role, principal perpetrator,
or secondary/indirect involvement);119
r premeditation and discriminatory intent;120
r position as a superior, in particular abuse of that position;121 and
r behavior of the accused during trial.122
In order to affect sentence, aggravating factors must be proven beyond a
reasonable doubt.123 The ICTY has stated that an aggravating factor only can
increase the sentence if that factor did not form an element of the actual
offense.124 For example, when discriminatory intent forms part of the requisite
elements for proof of the crimes charged, it will not be considered separately
as an aggravating factor in sentencing. The ICTY has taken a similar approach
to command responsibility, holding that a defendant convicted based on com-
mand responsibility cannot receive aggravated punishment merely because he
held a superior position, but only if he abused the superior position.125
Mitigating factors, which require proof only on a balance of probabilities,126
r whether and when the accused pled guilty;127
r substantial cooperation on the part of the offender;128
r remorse;129
r the youth,130 advanced age,131 and other personal circumstances of the
offender (including whether married and with children);132
r the extent to which the offender was subject to duress, orders, or coercion;133
r the “good character” of the offender;134
r the chaos of constant armed con¬‚ict;135
r that the offender did not have a previous criminal record for ordinary common
crimes;136 and
r human rights violations suffered by the offender during pre-trial or trial pro-
On the subject of plea bargains, the overall practice of international institu-
tions is to sentence defendants who plead guilty to a shorter term of imprison-
ment than they would have received were they to be convicted following a trial.
That said, as I explore further in Chapter 6, the actual discount rate that attaches
to plea bargains is dif¬cult to measure and, in fact, ¬‚uctuates markedly among
Punishment in International Tribunals 65

defendants even when sentenced by the same institution. This creates consid-
erable indeterminacy. In the aggregate, however, pleading guilty is a relevant,
albeit controversial, mitigating factor.
Although the ad hoc tribunals began their operations by viewing plea bar-
gains with disfavor, this approach has changed over time.138 Rule amendments
eventually were adopted that permitted plea bargaining.139 Plea bargaining ¬rst
proliferated at the ICTY, in part due to the willingness of ICTY defendants to
barter for a reduced sentence.140 In response to this heady recourse to plea bar-
gaining, however, ICTY judges began to express the need for some caution in
approving plea bargains and, in some cases, exercised a greater level of indepen-
dent curial review over bargains concluded between defendants and the Of¬ce
of the Prosecution.141 Rulings by the ICTY Trial Chambers (some of which have
been af¬rmed on appeal) in a small number of cases to impose a sentence longer
than the range contemplated in the plea agreements have had somewhat of a
chilling effect on plea bargaining practice.
There have been fewer plea agreements at the ICTR. The ICC permits
proceedings on an admission of guilt.142 It is reasonable to expect that ICC judges
will treat an admission of guilt as a mitigating factor. The Special Panels af¬rmed
a large number of plea bargains, particularly as the institution became defunded
and its mandate wound down. In some cases the Special Panels sentenced
perpetrators who self-convicted to very modest terms, including under ¬ve years
for crimes against humanity.
In addition to pleading guilty,143 the Special Panels claimed similar aggravat-
ing and mitigating factors than the ICTR and ICTY in the exercise of their
discretion to punish. A review of the Special Panels™ jurisprudence reveals
considerable attention paid to gravity,144 vulnerability of victims,145 superior
responsibility,146 and political context147 as aggravating factors; and, as mitigat-
ing factors, remorse,148 personal/family circumstance,149 and position as a sub-
ordinate/coercive environment.150 In the case of the Special Panels, sentencing
guidelines from more than one national justice system in¬‚uenced, but certainly
did not structure, the work of judges in punishing international crimes.151 The
Special Panels refer to traditional indigenous principles in sentencing, such as
adat (taking responsibility/paying respect to authority) and, in this sense, take
important steps toward the development of more autonomous, and theoretically
composite, approaches to punishment pertinent for East Timor.152
By and large, the aggravating and mitigating factors considered by inter-
national tribunals in punishing international crimes resemble those used by
domestic courts of dominant states when they sentence perpetrators of ordinary
common crime. Many of these factors emerge in the international jurisprudence
because international judges engage in comparative legal analysis of these dom-
inant systems whose tenets they then incorporate. The only factor that stands
out in exception is the discounting of a sentence owing to the chaos that may
ensue from endemic armed con¬‚ict or coercive environments.153 For the ad hoc
tribunals, however, this is “not a decisive factor”154 and was in fact explicitly con-
demned by the ICTY Appeals Chamber in the Blaˇki´ decision.155 The Special
66 Atrocity, Punishment, and International Law

Panels have been somewhat more re¬‚ective regarding the limitations of human
agency in contexts of collective violence.156 However, this factor remains an
outlier in the actual quanti¬cation of sentence. The Special Panels, which had
dual jurisdiction, did not differentiate among the criteria used or the theory of
punishment espoused when it came to sentencing ordinary crimes or sentenc-
ing international crimes. They turned to the same aggravating and mitigating
factors for both sets of crimes.

(iv) conclusion
Positive law instruments permit incarceration and restitution as punishments
for extraordinary international criminals, but thus far the law-in-practice of con-
temporary international institutions has been limited to incarcerating along
the ordinary lines of the penitentiary model. Although the practice of the ICTR,
ICTY, and East Timor Special Panels suggests that retributive motivations retain
the greatest currency, a palpable level of indeterminacy remains with regard to
why international criminal institutions punish individual offenders. Judges still
remain unsure, and often divided, about the purpose of the punishments they
mete out.
The vagueness of the positive law frameworks enables judges to access a
wide range of evidence in determining sentence. Judges have injected some
order into this process by developing a typology of aggravating and mitigating
factors as variables. Although international judges have come a long way since
Nuremberg and Tokyo, they have not developed a cogent framework or heuristic
to standardize the weight to attribute to each of the many pieces of evidence
available for consideration in the typology of aggravating or mitigating factors.
International criminal sentencing practice remains “open-ended.”157 Recourse
to aggravating and mitigating factors, and the weight to attribute thereto, is
avowedly discretionary. This leads to indeterminacy at a second level, namely
how much imprisonment is levied out to individual convicts. This indeterminacy
endures notwithstanding the emergence of a ¬‚edgling jurisprudence that might
help systematize sentencing. Although individualized sentencing has many ben-
e¬ts, these become jeopardized when no rubric exists to ensure consistent and
proportionate application of standard criteria among individual defendants. The
erratic sentencing practice could also affect the coherence158 and legitimacy of
the punishing institutions, which, in turn, may undermine con¬dence in their
rationality and even, as H.L.A. Hart warned, bring the law into contempt.159
Although different societies may sentence differently “ and this diversity is to be
welcomed “ once a punishing regime has been established for an atrocity, it is
important that, regardless of its theory or modality, it works in a principled and
predictable manner in how it treats individual defendants.
Punishment serves a very important role in providing subtle and ¬ne-grained
assessments of individual responsibility. Criminal liability as delineated by a
forced choice between acquittal or conviction offers little more than crude
binary reductionism. Sentence, however, can serve to refract that reductionism.
Punishment in International Tribunals 67

Therefore, it becomes all the more important that sentence be effected coher-
ently as well as thoughtfully. In the case of extant international criminal tri-
bunals, a gap emerges between the avowed goals of sentencing and the actual
outputs of the sentencing process. In the end, the abundance of discretion feeds
this gap.
chapter 4

Punishment of International Crimes in National
and Local Criminal Justice Institutions

National and local criminal justice institutions play a key role in sanctioning
extraordinary international criminals. These institutions in fact undertake the
bulk of the work. International institutions are designed to prosecute individuals
alleged to bear the greatest responsibility for atrocity and, therefore, are intended
to focus on leaders and organizers although, in practice, they do prosecute lower-
level offenders (as was routinely the case in East Timor). For the most part,
though, lower-level offenders “ many of whom, like the Rwandan bulldozer
driver Anastase Nkinamubanzi, killed many innocents in grisly fashion “ remain
in the hands of national and local institutions.
The scholarly literature on how domestic courts punish international crimes
when they exercise national, territorial, or universal jurisdiction is limited. This is
a notable lacuna insofar as the ICC formally defers to national courts as the front
line of prosecution and punishment through the doctrine of complementarity.
Moreover, the completion strategies of the ICTY and ICTR activate the referral
of cases to national institutions (in the states of the former Yugoslavia or Rwanda,
or the courts of any state). These referrals preserve the primacy of the ad hoc
tribunals over national institutions, but lead to the reality that cases will be (and
already are) processed at the national level.
Overall, and to varying degrees inter se, national and local criminal jus-
tice institutions tend to gloss over the conceptual differences between ordinary
domestic crimes, on the one hand, and extraordinary international crimes, on
the other. Overwhelmingly, national frameworks in many states punish extraor-
dinary international criminals through the same methods as ordinary common
criminals “ principally imprisonment “ within a system designed for ordinary
common criminals. They punish largely, though not entirely, for the same rea-
sons; aggravating and mitigating factors tend to track those of select ordinary
criminal law.
Survey research demonstrates that certain domestic frameworks, particularly
those in European and common law countries, do punish extraordinary inter-
national criminals more harshly than ordinary domestic criminals insofar as
they contemplate an increased term of imprisonment for extraordinary interna-
tional crimes.1 This phenomenon could suggest that the gravity of extraordinary
Punishment of International Crimes Nationally and Locally 69

international crimes is viewed as greater and that what is required to articulate
this enhanced gravity “ and, thereby, retributive goals “ is a formulaic adjust-
ment upward in terms of the number of years a convict serves. That said, actually
adding to sentence and requiring the convict to serve these additional years is
not possible in many places, insofar as, in the words of an ICTY Trial Chamber,
“in most countries a single act of murder attracts life imprisonment or the death
penalty, as either an optional or mandatory sanction.”2 It therefore becomes dif-
¬cult to make the extraordinary international criminal, who may be responsible
for the deaths of hundreds while pursuing eliminationist ends, actually spend
more time in jail than the ordinary criminal who murders one person for pro¬t,
out of anger, or inadvertently in the course of committing a felony.
In any event, statutory treatment that permits longer sentences for extraordi-
nary international crimes often emerges in countries that never have prosecuted
a single individual for such crimes. When the practice of states that actually have
prosecuted atrocity crimes is taken into account, the picture becomes consider-
ably more nuanced and kaleidoscopic. A deep review of the jurisprudence from
such states reveals a textured composition: there are several stated penological
goals that, in addition to retribution, include deterrence, reconciliation, and
restoration. Principled attainment of these goals is obscured by virtue of a pro-
nounced level of discretion in sentencing. For example, in terms of underscoring
the gravity of the offense, there is no predictable pattern within these jurisdic-
tions of punishing a similar physical act (e.g., murder or rape) more severely
when committed in situations of con¬‚ict or genocide than when committed in
ordinary times. In Rwanda, for example, certain punishments for extraordinary
international criminals are in fact lower than what would attach to offenders in
ordinary times; furthermore, confessing and pleading guilty to an extraordinary
international crime will trigger a signi¬cant sentencing discount that is unavail-
able for ordinary crimes. In the states of the former Yugoslavia, judges do not
consistently sentence more severely for wrongdoing committed as war crimes
than committed ordinarily. Most jurisdictions prosecuting World War II atrocity
simply transplanted the punishments ordinarily available for common criminals
to perpetrators of atrocity, although there are examples where punishments for
atrocity crimes were explicitly made harsher than those available for common
Looking beyond, national prosecutors steering political transition may face
a particularly unique set of circumstances in determining whether or not to
exercise their discretion to prosecute, thereby leaving offenders unpunished for
what might be perceived as a greater overall good. Many states have awarded
amnesties to extraordinary international criminals that they would never award
to ordinary common criminals.3 In South Africa, for instance, political crimes
were open to amnesty, whereas ordinary crimes were not.4 Thus, individuals
animated by political motives were treated more leniently than those inspired
by private motives. In prosecutions following World War II, many Holocaust
perpetrators were treated lightly by the courts and pardoned as early as several
years after conviction, while others were quickly executed.
70 Atrocity, Punishment, and International Law

Although national institutions have not developed a special penology for
extraordinary international crime, and have not fared much better in theoriz-
ing the criminality of mass atrocity, they do engage in some methodological
creativity and, in some instances, turn to a broader diversity of sanction than
international tribunals. On occasion, national and local justice institutions cul-
tivate approaches to punishment that diverge from international norms. These
approaches may more accurately re¬‚ect sociolegal norms of the places most
immediately af¬‚icted by mass atrocity. They also may prompt concerns regard-
ing communitarian punishment and the quality of justice.
However, the range of domestic initiatives that diverge from international
norms is circumscribed by the pressures that international institutions, even
though they focus on a narrow band of perpetrators, exert over their national and
local counterparts. This gives rise, as I explore in greater detail in this and the fol-
lowing chapter, to legal transplants from the international to the national, many
of which are welcomed by state actors to manage the in¬‚uence of local com-
munities and curtail the diffusion of authority. These transplants have a homog-
enizing effect on the kind of sanction visited upon atrocity perpetrators. In the
end, local communities, often deeply af¬‚icted by atrocity, have been hemmed
in by these exogenous pressures when they endeavor to develop approaches to
punishing perpetrators that depart from liberal international modalities.
Regarding research methodology: it would be overwhelming to review every
case by a national court or local institution that concerned conduct that might
be classi¬ed as an extraordinary international crime. Domestic institutions have
been called upon in many different contexts to retrospectively redress civil vio-
lence or sanction abuses by military personnel. Examples include, but are far
from limited to: Greece, Argentina, Bolivia, Chile, Peru, Guatemala, Haiti,
United States,5 Germany (following reuni¬cation), many Eastern European
countries (with regard to crimes committed under Stalin [for example, in Latvian
courts] or more recently under Communist rule), Burundi, Ethiopia, Indonesia,
Afghanistan, and Sri Lanka.6
For the purpose of the analysis at hand, I propose that a fruitful avenue of
research is to explore the activities of national and local institutions in punish-
ing extraordinary international crimes that these institutions themselves de¬ne
as such and that also have been or are being prosecuted at the international
level. I select in this regard three atrocities: the 1994 Rwandan genocide, eth-
nic cleansing and genocide in the Balkans throughout the 1990s, and the Nazi
Holocaust (although I also include some discussion of Japanese and German
war crimes against combatants and civilians). In this chapter, I comparatively
review the activities of national and local legal institutions sharing contacts
with these three atrocities. Each of these three case studies evidences the kind
of discriminatorily motivated violence that runs to the heart of international
criminal law™s proscriptions.7 Therefore, centering the analysis on these three
case studies harmonizes the discussion with my focus, as set out in Chapter 1,
on ideologically and politically motivated violence; it also permits comparative
assessments between the international sphere and national/local spheres. The
purpose of this analysis is to qualitatively document sentencing practices, see
Punishment of International Crimes Nationally and Locally 71

tendencies, and sketch rationales. In some places assailed by atrocity “ for exam-
ple, Afghanistan, Kosovo, and Rwanda “ law at the state level differs from law as
practiced traditionally at the local level, especially in matters of procedure. The
study of local process is a complex undertaking. Given the paucity of research
on local modalities of punishment for perpetrators of great evil, however, I am
hopeful that even the cursory overview this chapter provides will advance the
discussion and signal other important work that remains to be done.
Among these three case studies, I devote the most attention to Rwanda. I am
motivated in this regard by my own experiences in the country as well as the large
number of sentences that Rwandan institutions have issued. Although the quality
of scholarly analysis of accountability initiatives in Rwanda has grown, much
work remains to be done insofar as these initiatives still remain understudied. By
way of example, Mark Osiel “ one of the most insightful and in¬‚uential authors
on issues of international criminal justice “ downplays the Rwandan experience,
which does not mesh with his theoretical modeling of postcon¬‚ict prosecutorial

(i) rwanda
Between April and July 1994, anywhere from 500,000 to 800,000 people were
massacred in genocidal pogroms in Rwanda.9 This is a staggering amount of
death in a country with a total population of about eight million. Many of the
killings were unspeakably brutal. They were in no way depersonalized through
technology: a study conducted by the Rwandan government concluded that
nearly 38 percent of victims were killed by machete, 16.8 percent by club, and
14.8 percent by ¬rearm; other means of murder included grenades, swords,
knives, drowning, sticks, rocks, and bare-handed assault.10
The perpetrators of the violence were members of the majority Hutu eth-
nic group, radicalized by an extremist Hutu government. The overwhelming
majority of victims were members of the minority Tutsi group. The Hutu com-
prise approximately 85 percent of Rwanda™s population, the Tutsi 14 percent.
The genocide was quelled when a Tutsi army (the Rwandan Patriotic Army,
RPA),11 based in neighboring Uganda, ousted the genocidal Hutu government
and seized power. The political wing of this Tutsi group, the Rwandan Patriotic
Front (RPF), currently retains a ¬rm grip on power in Rwanda.
The judicialization of atrocity in Rwanda proceeds through three sets of

(1) the ICTR, established by the Security Council and sited in Arusha,
(2) domestic courts, overwhelmingly in Rwanda but also in a handful of
foreign jurisdictions, including Belgium; and
(3) a modi¬ed form of gacaca (traditional dispute resolution), adapted
for genocide-related crimes and standardized through centralized national
72 Atrocity, Punishment, and International Law

The Organic Law on Gacaca Jurisdictions, which ¬rst took effect in 2001 and
has been subsequently amended (including important amendments in 2004),
creates gacaca courts to hear genocide-related charges. In Rwanda, the term
“Organic Law” refers to laws that rank higher in normativity than ordinary laws,
and are secondary only to the Constitution. Another Organic Law, from 1996,
organizes criminal proceedings for genocide or crimes against humanity and
offenses committed in connection thereto, initially in Specialized Chambers
within the conventional national and military courts.12 These proceedings also
invoke Rwandan general criminal and constitutional law, as well as substantive
international criminal law as codi¬ed in treaties, and thereby re¬‚ect an inter-
penetration of various sources of law, both general and speci¬c. The Specialized
Chambers were formally abolished by the 2001 Organic Law, which formally
repealed the 1996 Organic Law, but speci¬ed that the 1996 Organic Law remains
applicable to all cases forwarded to the Specialized Chambers that now are to be
handled by the national courts. Genocide trials have continued in the national
court system, although their number has tapered off.13 By mid-2002, 7,181 pros-
ecutions had occurred in the Specialized Chambers; by 2003 the overall ¬gure
in the national courts rose to “slightly more than 8,000”;14 and, by 2005, approxi-
mately 10,000 prosecutions had occurred. Although this is a substantial number
of trials “ “better than the record of many European countries following the
Second World War”15 “ it only involves a small portion of the total detainee
Thus far, the ICTR has arrested seventy-two individuals. The Rwandan gov-
ernment had initially requested the creation of an international tribunal, but
then cast its Security Council vote against the ICTR. The Rwandan govern-
ment objected inter alia to the siting of the ICTR outside Rwanda, its limited
temporal jurisdiction, the absence of Rwandans on its staff, and its inability to
issue a death sentence. That said, in practice the Rwandan government gener-
ally, though certainly not routinely, cooperates with the ICTR. Attitudes of the
Rwandan population toward the ICTR range from disinterested to skeptical.
The International Committee of the Red Cross estimates that 89,000 indi-
viduals remain detained in Rwanda on genocide-related charges.16 The ¬gure
formerly was higher insofar as over 36,000 other individuals had been slated
for parole (provisional release) in recent years owing to lack of evidence, age,
in¬rmity, or illness; the majority of these individuals, however, were paroled
because they had confessed to involvement in the genocide.17 Many of these
parolees have undertaken to participate in gacaca, an undertaking on which
their parole remains contingent. Many parolees have committed to attending
reeducation camps where they receive government-sponsored instruction on
justice and reconciliation.
Eventually, the Rwandan government intends for all but those who remain
accused of the most serious offenses to be prosecuted through gacaca. In late
2005, the gacaca Secretariat announced its intention to establish a new national
court to try individuals accused of the most serious offenses (estimated at up to
ten thousand persons). So, once again, the Rwandan justice system is subject to
Punishment of International Crimes Nationally and Locally 73

profound structural reform and reconstitution. The jurisprudence of the Spe-
cialized Chambers and ongoing verdicts of conventional courts would provide
some guidance to gacaca judges regarding ¬ner points of liability and sanction
and, thereby, likely would inform the work of this new national court. In my
estimation, gacaca, which portended to be a signi¬cant departure from retribu-
tivism and a turn towards restoration, has been underemployed in this regard
owing to pressures it has faced to re¬‚ect the ideal-type adversarial criminal trial
and to serve state, as opposed to local, interests.

a. National Courts in Rwanda, Including Specialized Chambers
The 1996 Organic Law creates four categories of culpability. These are: (a) Cate-
gory 1 (planners, organizers, those in positions of authority, notorious murderers
[with zeal or excessive malice], and sexual torturers); (b) Category 2 (perpetrators
of intentional homicide or serious bodily assault causing death); (c) Category 3
(perpetrators of other serious assaults); and (d) Category 4 (perpetrators of prop-
erty offenses).
Article 14 of the Organic Law deals with punishment. It links the severity
of punishment, as well as its form, to the gravity of the offense as represented
ordinally by the category of culpability.18 In some cases, the linkage is to a ¬xed
sentence, while in others it is to a permissible range of sentence. The discretion
of judges in ¬xing sentence is thereby fettered, unlike at the ICTR where judges
are accorded broad discretion regarding the length of sentence to be imposed
(although the nature of punishment at the ICTR is limited to imprisonment
and restitution).19 The Organic Law does not contemplate early release. The
Organic Law™s explicit linkage of penalty to type of offense provides a level of
predictability in sentence, which is important to the credibility of the punishing
framework although, as is the case with ¬xed sentences in any jurisdiction, might
lead to inequities in individual cases. Article 14 stipulates that punishments are
those listed in the Rwandan Penal Code, which applies to ordinary criminal
offenses, except for:
r Category 1 offenses, which are punishable by death;
r Category 2 offenses, for which death is replaced with life imprisonment; and
r Category 4 offenses shall only give rise to civil damages determined by agree-
ment between the parties, failing which rules related to criminal proceedings
shall apply although any sentence issued is to be suspended.
The case law provides examples of Category 4 offenders sentenced to jail
time.20 Given that many Category 4 defendants have been detained for several
years by the time they eventually face trial, even if they were sentenced to jail
time their likely fate would be release for time already served.
The language of the Organic Law leaves Category 3 offenders subject to
the ordinary sentences provided by ordinary Rwandan criminal law for serious
assault. Judges have considerable scope for discretion in sentencing Category 3
offenders insofar as the sentence ranges are relatively broad. Punishments
74 Atrocity, Punishment, and International Law

normally imposed by the Rwandan Penal Code for conduct that could con-
stitute a Category 3 offense include, but are not limited to:
r Penal Code article 318: violent attacks (one month to one year);
r Penal Code article 319: violent attacks causing an illness or inability to work
(two months to two years; six months to three years if committed with pre-
r Penal Code article 320: violent attacks causing, inter alia, serious mutilation
or incurable illness (two to ¬ve years; ¬ve to ten years if premeditation is
Although death sentences can be awarded to Category 1 offenders “ and
courts continue to issue them “ no executions have occurred since 1998. Why
have Rwandan authorities apparently abandoned the death penalty in prac-
tice? Arguably, Rwanda has decided that this particularly retributive sanction
no longer promotes the goals of genocide prosecutions. Other reasons, which I
also believe to be in¬‚uential, include the international community™s condem-
nation of the death penalty, the skepticism of transnational nongovernmental
and donor organizations, and the ICTR™s position that it will not transfer any
cases to the domestic Rwandan authorities where the accused could face the
death penalty. On this latter note, a similar dynamic emerges with regard to
Sierra Leone: “[i]f the death penalty were not prohibited at the [Sierra Leone]
Special Court, it would be next to impossible to secure funding from European
and certain other major donors.”22
The Organic Law encourages defendants to confess their guilt. It creates a
scheme that incentivizes confessions and guilty pleas. Many defendants have
availed themselves of this scheme, although not as many as authorities initially
had hoped. In order for a confession and guilty plea to be valid, the Organic Law
(articles 5 and 6) requires that it be made before trial, describe in detail all the
offenses and victims, provide information regarding other involved individuals,
include an apology, and contain an explicit plea offer. A plea that fails to comport
with these requirements, or which the prosecutor deems is inaccurate, will be
rejected.23 Insofar as there is no explicit process of negotiation, the practice
of plea bargaining in Rwanda differs from that at the ICTY or ICTR, where
defendants may plead guilty to a single umbrella charge or to a subset of charges
and international prosecutors, in turn, may drop other charges. Also, unlike at
the ad hoc international tribunals, the statutory framework heavily regulates the
effects of the confession and guilty plea on sentence.
If the court accepts a guilty plea, then Organic Law articles 15 and 16 govern.
These base the extent of the sentence reduction on when the accused confesses
and for which crime the accused pleads guilty. For example: under article 15, if
the guilty plea is entered before charges are ¬led, then for a Category 2 offender
the punishment “ ordinarily life imprisonment “ is reduced to a sentence of
between seven and eleven years, to be ¬xed by the court. This is a major reduc-
tion. A Category 3 offender who properly pleads guilty before charges are ¬led
is sentenced to one-third of the jail time that the court normally would impose.
Punishment of International Crimes Nationally and Locally 75

The goal here clearly is to encourage those who believe they have committed
a genocide-related offense to come forth on a voluntary basis before they of¬-
cially become suspects. In one case, Minist` re Public v. Bugirimfura et al., one
defendant pled guilty completely and sincerely to genocide (Category 2) before
charges were brought and received a sentence of ten years; four other defen-
dants went to trial, after which they were found guilty of genocide (Category 2),
and were sentenced to life imprisonment.24 In this case, the plea bargain stated
the facts, the names of the victims, denounced collaborators, announced regrets,
and mentioned that the defendant was sorry. The court dipped below the pros-
ecutor™s recommendation of twelve years in its issuance of a ten-year sentence.
The court is not bound to follow the prosecutor™s recommendation regarding
sentence on a guilty plea and has the power to accept a guilty plea that the
prosecutor has rejected.25
Article 16 establishes punishment for guilty pleas entered after charges have
been ¬led. Here, the sentence for an offender pleading guilty to a Category 2
crime is to be ¬xed by the court within a range of twelve to ¬fteen years; for
a Category 3 offender the sentence is one-half the term that would normally
be imposed. In Minist` re Public v. Bizuru et al., a joint trial involving a similar
set of factual circumstances, those defendants who pled guilty to a Category 2
offense before charges were brought were sentenced to eleven years, whereas
another defendant who pled to a Category 2 offense after charges were brought
was sentenced to ¬fteen years.26
Defendants who confess and plead guilty to a Category 1 offense are ineligible
for the sentence reductions found in articles 15 and 16. There is an exception
for individuals who are not on an of¬cial list of Category 1 suspects maintained
by the prosecution. If individuals come forth, confess, and plead guilty to what
is a Category 1 offense, they shall be reclassi¬ed as Category 2 offenders.27
Although the guilty pleas are heavily regulated by statute, they have spawned
a considerable amount of interpretive jurisprudence, particularly with regard to
factors that courts consider in specifying sentence within the ranges established
by the Organic Law. A skeletal typology of aggravating and mitigating factors
has thereby emerged. Another topic of concern to judges is what to do with the
incomplete, irregular, or unacceptable plea: in other words, a plea that does not
conform to the Organic Law™s requirements regarding form, content, timeliness,
or truth. In some cases, examined next, courts will give these some weight in
mitigating sentence.
Article 17 of the Organic Law permits the court to punish by stripping the
convict of certain civic rights. This can be permanent (d´ gradation civique
perp´ tuelle et totale [sometimes referred to as d´ ¬ntive]) or limited either in scope
e e
or temporal duration (d´ gradation civique limit´ e). The Organic Law again links
e e
the severity of the d´ gradation civique to the offense for which the defendant
is convicted. A review of the jurisprudence reveals that many convictions for
genocide-related offenses are accompanied by an order for d´ gradation civique.
This supplemental sanction can calibrate the retributive value of punishment
by creating proportionality in sentence for more egregious crimes, for example
76 Atrocity, Punishment, and International Law

through the determination whether or not to make the d´ gradation civique tem-
porary or permanent within the same category of offenses. D´ gradation civique
is a form of shunning and stigma insofar as the perpetrator is hindered from
reintegrating back into the community. In this regard, d´ gradation civique is at
odds with rehabilitative, reintegrative, or reconciliatory purposes of punishment.
Examples of those civic rights or privileges stripped through d´ gradation civique
include: the right to vote; other political rights (such as to be a candidate); to
serve as an expert or witness in trials or to be deposed judicially other than
for the giving of simple facts;28 the right to carry arms; to serve in the armed
forces; to be police of¬cers; and to teach in any educational institution.29 The
restrictions on admissibility and weight of a convict™s subsequent testimony do
limit the use of such testimony for the purpose of inculpating others and this
explains, at least in part, why “ unlike the practice at the ICTY “ there is limited
recourse to bargaining away charges in exchange for procuring an individual™s
testimony against others. The Conseil de Guerre, which adjudicates military
of¬cers accused of offenses related to genocide or crimes against humanity, also
can punish by expelling convicts from the armed forces (d´ gradation militaire
or exclusion de l™arm´ e).
In a December 2002 report, Amnesty International compiled statistics regard-
ing a total of 7,181 persons judged for genocide-related crimes in Rwanda since
1997.31 Amnesty International found that 9.5 percent of defendants were sen-
tenced to capital punishment, 27.1 percent to life imprisonment, 40.5 percent
to ¬xed prison terms, and 19.1 percent were acquitted. In a 2000 Report to the
United Nations, Special Representative Michel Moussalli stated that 2,406 per-
sons had been tried by the genocide courts of whom 14.4 percent were sentenced
to death, 30.3 percent to life imprisonment, 34 percent to terms between one
and twenty years, and 19 percent acquitted.32 Longitudinally, the Amnesty Inter-
national study demonstrates the following trends: decline in capital sentences
from 30 percent of perpetrators in 1997 to 3.4 percent in 2002 “ with steady
annual decreases; decline in life imprisonment from 32.4 percent of perpetra-
tors in 1997 to 20.5 percent in 2002; and increase in ¬xed prison terms from
27.7 percent of perpetrators in 1997 to 47.2 percent in 2002.33 These trends
arise from a number of factors, including that the initial trials focused on the
more notorious killers and that, with the passage of time, increased recourse
was made to guilty pleas (including in recent years with a view to entering
the gacaca system). The acquittal rate in 2002 was nearly three times that
in 1997.
The Amnesty International statistics, however comprehensive, do not illus-
trate the factors the domestic genocide courts consider in sentencing that tran-
scend the guidelines provided by the Organic Law. The statistics are silent as to
how the Rwandan genocide courts exercise their limited discretion with regard to
punishing Category 2 and 3 offenders. Nor do they reveal the ways in which the
Rwandan courts at times mold the statutory framework to suit unusual circum-
stances; or how, through the language, tone, and texture of their judgments,
they give voice to certain penological goals in a manner that transcends the
Punishment of International Crimes Nationally and Locally 77

quanti¬cation of sentence or involves legal sanction from outside the criminal
law (e.g., civil damages awards).
Accordingly, I conducted a qualitative review of the published judgments (in
French) of the Rwandan genocide courts. The database of judgments I reviewed
is that compiled and maintained by Avocats sans fronti` res (ASF), which they
publish in bound volumes and maintain online. These judgments comprise
a wide sample from across Rwanda; however, insofar as ASF publishes only the
more sophisticated judgments, this sample is not fully representative of what
transpires in many Rwandan courts but, rather, is representative of those judg-
ments that address questions of fact and law in a manner that carries interpretive
value. In the cases I reviewed, when aggregated, defendants received the fol-
lowing sentences (largely consonant with the Amnesty International ¬ndings):
15 percent death, 30 percent life imprisonment, and 55 percent ¬xed terms.
Among the ¬xed terms, I calculated the median term to be 11 years, and the
mean term 15.25 years. Courts issued orders against nearly all convicted defen-
dants for restitution/compensation based on collateral private lawsuits (parties
civiles) ¬led by the victims and/or surviving family members. In the event of
conviction, courts also issued orders for legal fees to be assessed against the
defendant for the cost of the criminal proceedings in the event of conviction.35
These orders are often made under threat of the forced seizure of all of the
defendants™ real and personal property in the event of inability to pay (which is
commonly the case). In one case, the court issued a ¬ne.36
These judgments are the product of a court system that has grown in sophisti-
cation since it began hearing cases, initially somewhat haphazardly, in the after-
math of genocidal devastation. Moreover, dating from well before the genocide,
the history of Rwandan courts is one of corruption and partiality. To this end, the
current state of the Rwandan judiciary is more reliable and competent than it
likely ever has been. That said, although the judgments in the database go into
great depth regarding the proof of the crimes, they remain very cursory as to sen-
tencing. There are no separate sentencing hearings. The general practice is for
the court to issue a sentence with no explanation.37 This is not unusual, insofar
as those institutions that prosecute extraordinary international criminals gener-
ally accord punishment a markedly lower level of attention than they accord
questions of culpability. Nevertheless, a deep reading of the published Rwan-
dan cases does permit the delineation of certain trends with regard to why the
courts, in particular the Specialized Chambers, punish more or less severely in
individual cases, especially when an exercise of discretion is involved. That said,
there is surprisingly limited discussion of the purposes of punishment and how
sentence can promote these various purposes. Engagement with penological
theory is essentially nonexistent.
The articles of the Organic Law say nothing speci¬c about the goals of pun-
ishment; the preamble provides only vague and generic references. One of the
unof¬cial commentaries to the Organic Law identi¬es the following as peno-
logical purposes: punish the guilty (which, although unarticulated, seems to
be a retributive goal), serve as a dissuasive example, protect the people, and
78 Atrocity, Punishment, and International Law

rehabilitate the accused.38 Reconciliation is a goal of the confession and guilty
plea process, although it is unclear whether the operation of the process is effec-
tive at attaining this goal.39 The ordinary Rwandan Penal Code also is reticent
regarding the principles of punishment.40
Insofar as the sentences established by the 1996 Organic Law are, by statute,
directly calibrated to a hierarchy of offenses ordinally ranked in light of the
gravity of the crime, the primary purpose of punishment arguably is retribution.
For the retributivist, after all, the severity of punishment is to be proportion-
ate to the gravity of the offense. The retributive purpose is protected through
the Organic Law™s explicit removal of judicial discretion to depart from the
statutorily prescribed range of sentence for each offense. Comments by Rwan-
dan government of¬cials and prosecutors “ along with academic observers “
also suggest the importance of retribution.41 There are, however, a number of
interesting wrinkles to the apparent importance of retribution. One wrinkle is
that the Rwandan courts, in the judgments I reviewed, do not explicitly note the
salience of retribution or explicitly apply retributive values to the speci¬c context
of genocide. A second wrinkle is one of internal consistency. The punishment
for certain offenses under article 14 that would fall in Categories 2 and 4 is less
onerous than the punishments ordinarily available under the Rwandan Penal
Code.42 For example, the Rwandan Penal Code provides capital punishment for
certain premeditated murders and felony murders that would, in the language
of the Organic Law, fall within Category 2 offenses for which life imprisonment
is the sentence. Certain property damage offenses, such as arson, are punished
by long terms of imprisonment under the ordinary Penal Code but much more
lightly under the Organic Law. This reality intersects oddly with the retributive
aims of the Organic Law and reveals that punishment in collective violence has
other goals.
The domestic Rwandan courts have identi¬ed factors they consider in
quantifying sentence in individual cases, especially with regard to mitigating
factors. A number of mitigating factors (circonstances att´ nuantes) emerge from
articles 82 and 83 of the Rwandan Penal Code. However, courts make no effort to
explain why these factors, intended for common criminals committing ordinary
deviant crimes, are appropriate for perpetrators of great evil in the context of
collective cataclysm.
Aggravating factors, as had been the case at Nuremberg, often are assumed
from the grisly nature of the conduct. What is more, in the Rwandan context
they already are implied in the severity of sanction insofar as the factors that go to
identifying liability for a Category 1 offense (such as senior position, zeal, organiz-
ing, notoriousness, and particular brutality [m´ chancet´ excessive]) correspond
e e
to those factors to which judges pursuing retribution could be expected to turn
in order to award sentence within an entirely discretionary sentencing structure.
In one case, an aggravating factor (the defendant tore out the eyes of his victims
prior to killing them) was cited to void the defendant™s partial guilty plea of any
effect in mitigation, resulting in a Category 1 conviction (death sentence).43 At
times, the Rwandan courts mention certain factors in aggravation even though
Punishment of International Crimes Nationally and Locally 79

they are not able to increase the level of punishment. Doing so serves multiple
purposes. First, to promote an expressive or declaratory function that further
stigmatizes the convict. Second, to emphasize why the convict was placed into
a particular category in the ¬rst place.44 Third, to explain why the court chose
not to follow defense counsel™s recommendation as to sentencing or chose to
ascribe little weight to circumstances that might otherwise be mitigating.45 The
court also may use the amount of civil damages in a punitive sense (instead of
merely compensatory or restitutionary) to operationalize aggravating factors.
My review of the case law reveals recourse to the following as mitigating
(a) Partial, incomplete, tardy, or irregular guilty pleas. Proper guilty pleas,
namely those that conform to the statutory requirements, carry signi¬cant
weight in reducing sentence (this is quanti¬ed in the Organic Law). A guilty
plea that falls outside of the statutory requirements, although void for the
purposes of formally reducing the sentence, may still be given discretionary
weight as a factor in mitigation.46 A court is especially willing to reduce sen-
tence when the irregular guilty plea is found to facilitate its work, contribute
to the telling of the truth, evidence a request for forgiveness, or is sincere.
The amount given as a discount in mitigation will not be as generous as
the statutorily provided discounts. That said, this reduction may permit the
defendant to receive a sentence below the minimum statutory sentence, for
example, a sentence of twenty years or less for a Category 2 conviction for
which the sanction is life imprisonment.47
(b) Minor status. Offenders under the age of fourteen cannot incur penal
responsibility in Rwanda. Offenders between the ages of fourteen and eigh-
teen at the time of committing the offense can incur penal responsibility, but
are entitled to raise their status as minors as a mitigating factor in sentencing.
The Rwandan courts give this factor considerable weight in mitigation. In fact,
they avail themselves of this factor, enumerated in article 77 of the Rwandan
Penal Code, even in situations where the statutory plea bargaining scheme
applies. For example, in Minist` re Public v. Nzabonimpa, the court cumu-
lated the guilty plea made before the proceedings and the fact the accused
was a minor under eighteen at the time of the offense to sharply reduce his
sentence to ¬ve years™ imprisonment plus legal costs.48 The charges against
this defendant involved his killing ¬ve Tutsi children between the ages of ten
and ¬fteen years with a masu (a club studded with nails) and having informed
others of their hideaway. The court did not justify why such a signi¬cantly
discounted sentence for a perpetrator because of his youth, which might be
apposite in the context of ordinary deviant crime, remains so in situations of
the perpetration of extraordinary evil such as genocide. After all, a signi¬cant
set of the reported cases involve minors, suggesting the complex agency of
minors as both victimizers and victims in Rwanda. Minority also is given
considerable cumulative weight in cases of irregular guilty pleas, moving
defendants in such situations well below the statutory punishment scheme.49
80 Atrocity, Punishment, and International Law

The fact that minors between the ages of fourteen and eighteen can face
criminal prosecution and punishment is another factor that distinguishes the
domestic Rwandan process from the ICTR.
(c) Coercion (“contrainte”). In one case, the court viewed as a mitigating
factor that accomplices obliged the defendant to participate by hitting him
with a machete.50 Elsewhere the courts have called this contrainte, which
loosely translates as coercion.51 Although ill de¬ned, contrainte implies that
the defendant was pressured into participating in the crimes. One compli-
cation here is that the Rwandan courts recognize duress as a substantive
defense. It would be odd to de¬ne contrainte the same way as duress, insofar
as the latter is a complete af¬rmative defense whereas the former only goes to
mitigation of sentence. Therefore, logically contrainte should fall somewhere
below the requirements for duress. This distinction, however, is not always
clear in the jurisprudence. Following orders, a closely related mitigating fac-
tor, also surfaces in the case law.52 The courts do not make an independent
inquiry about coercion in the context of mass atrocity in considering it as a
mitigating factor, once again transplanting it from its place within ordinary
common criminal law despite the different regulatory purposes of criminal-
izing mass atrocity and criminalizing isolated deviance.
(d) Individual characteristics. Rwandan courts have considered the following
in mitigation: that the defendant sheltered Tutsi during the genocide,53 the
ethnic status (Twa) of the defendants,54 lack of education (cited both as a factor
in mitigation55 and as a factor reducing the defendant™s criminal responsibility
from that of a Category 1 offender to a Category 2 offender),56 the defendant™s
weak physical health,57 and that the defendant did not organize the attacks.58

Individuals convicted of genocide-related criminal offenses also face civil
liability. This liability arises through collateral, private civil claims initiated by
surviving victims and certain surviving relatives of deceased victims; claims also
can be brought by the prosecutor on behalf of private parties. Private claims are
brought by virtue of the partie civile process established under ordinary Rwandan
law (applicable to all crimes) and are folded into the Organic Law.59 Collateral
plaintiff-initiated damages actions are common to civil law systems generally, but
in Rwanda the process also draws from traditions of custom and reparation that
animate gacaca. The criminal and civil avenues are not procedurally separate,
as is the case in common law systems. Partie civile damage awards arise in the
majority of the criminal cases published in the ASF database.
Partie civile lawsuits move in the direction of restitution and compensation
for those victims most immediately aggrieved. However, these lawsuits can also
serve retributive aims insofar as they permit a further marker of differentia-
tion among perpetrators (even within the same category in a ¬xed scheme of
mandatory sentencing) with regard to the gravity of their crimes. Civil liabil-
ity can thereby constitute an additional layer of punishment. To be sure, it
remains doubtful that successful claimants shall have their claims satis¬ed, as
Punishment of International Crimes Nationally and Locally 81

many problems arise with regard to the failure to pay compensation.60 Most
defendants are indigent. That said, those against whom unsatis¬ed claims have
been entered ostensibly would be dogged by that civil liability for life. In some
cases, the Rwandan state also is condemned as being jointly and severally liable
(damages in solidum) because of its incapacity to prevent the massacres; how-
ever, the Rwandan government has eschewed payments and has in fact insu-
lated itself from liability. Other than some modest funds raised through tax
revenues, authorities in Rwanda balk when it comes to providing reparations
to survivors or state compensation (even in the case of successful partie civile
claims). That said, even when an award remains uncollectible, the partie civile
process represents symbolic justice, promotes declaratory purposes, and consti-
tutes another layer in the public narratives regarding victimization during the
genocide. This process also permits victims some direct involvement in the legal
Civil damage awards constitute a signi¬cant departure from the law-in-
practice of the ICTR. One of the reasons why the ICTR has not issued a resti-
tutionary award is that the defendants appearing before it have been declared
indigent. However, failing to exercise the power to issue restitutionary awards for
this reason alone is not terribly compelling, insofar as the domestic Rwandan
experience reveals that ruling on civil claims can, at a minimum, serve important
expressivist, declaratory, and truth-telling purposes.
In Rwanda, if an accused is acquitted of all charges, no civil damages are
possible.61 The criminal conviction is a prerequisite for civil liability. However,
not all forms of individual criminal responsibility under the Organic Law trigger
civil damage consequences. For example, in a 2002 decision by a trial court
in Gikongoro, a defendant convicted of a Category 3 offense on the basis of
associational liability (association de malfaiteurs) and sentenced to ¬ve years™
imprisonment was found not to incur civil liability and the partie civile claim
failed.62 The court held that the defendant was not responsible for the losses
of the claimants™ family members (who were murdered) or possessions (which
were pillaged). The defendant committed neither offense, and was criminally
responsible only on the basis of a generalized associational presence. In any
event, the criminal conviction ultimately was quashed on appeal at the Cour
d™appel de Nyabisindu on December 11, 2002, and the defendant acquitted.
The size of the awards can be substantial. For example, in Minist` re Public v.
Twahirwa, the court awarded 144 million Rwandan francs against a Category 1
offender sentenced to death.63 In Minist` re Public v. Higiro (C´ lestin) et al., a
e e
trial court in Butare sentenced one of the defendants, Basomingera (Category 2)
to life imprisonment, d´ gradation civique, 1.5 million Rwandan francs to two vic-
tims each (for a total of 3 million) for the loss of their parents, plus miscellaneous
expenses such as court costs; in the event of nonpayment the court ordered the
forced sale of his possessions.64 In 2004, the average exchange rate of U.S. $1 was
575 Rwandan francs; in 2000, the exchange rate was 1 to 400. At the 2004 rate,
1 million Rwandan francs equals U.S. $1,739, a sum larger than the per capita
annual GDP in Rwanda.
82 Atrocity, Punishment, and International Law

At times the partie civile action is severed and to be determined later (for
instance, to give the claimants time to collect justi¬catory evidence). But the
action often is ruled upon immediately following the determination of the defen-
dant™s guilt or innocence. Rwandan courts examine the claimants™ losses in
great detail. In Minist` re Public v. Nteziryayo (Emmanuel) et al., the trial court
grouped damages into two categories: moral damages (for pain of losses of cer-
tain relatives) and material damages (loss of goods that had been pillaged, stolen,
or destroyed).65 It prepared a detailed schema, in Rwandan francs, of the value
to attribute to each of the heads of damage:66

Moral damages: 10 million for loss of a mother or a father
8 million for a child
5 million for a sibling
3 million for another close relative (i.e., uncle, aunt,
nephew, niece)
Material damages: 300,000 for a cow
20,000 for a goat
2,000 for a chicken
1,000 for a rabbit
2 million for a house built out of wood and thatch
5 million for a house built out of bricks with metal doors
5 million for household articles
1 million for the harvest

In Minist` re Public v. Rwanteli et al., a trial court in Cyangugu reviewed
damage claims for a broad array of losses, including pigs, goats, coffee, cement
bags, bags of green beans, and also because one of the victims had to spend
much time hiding in the weeds to escape from the killers.67 In this case, a
total of twenty-four million Rwandan francs was ordered, followed by the forced
liquidation of the assets of all of the defendants. The quanti¬cation of the heads
of damage in the Rwanteli case differed from the schema in the Nteziryayo case.
Differences are found among other cases as well. For example, in Auditorat
Miltaire v. Ukurikiyimfura et al., the Conseil de Guerre awarded moral damages
in the amounts of 10 million Rwandan francs for a spouse; 8 million for a child;
5 million for a parent or sibling; 3 million for a grandparent or grandchild;
2 million for an uncle, aunt, nephew, or niece; and 1 million for a brother- or
sister-in-law.68 The case law demonstrates signi¬cant discrepancies among courts
sitting in different regions of the country (and even within the same court) with
regard to the amount of loss awarded per type of damage.69 This creates a certain
level of inconsistency, insofar as plaintiffs are treated differently depending on the
discretionary (“selon sa sagesse”) or equitable exercise of authority by the court
that adjudicates their case and whatever schematic it may generally or speci¬cally
apply. This contradicts the emergence of national standards, although it may be
well tailored to do justice to the individual circumstances of the parties civiles.
These thorough discussions of the depth of the claimants™ suffering allow
their stories to be told in vivid, personal detail. It inks dignity and pain onto the
Punishment of International Crimes Nationally and Locally 83

pages of judicial documents often distinguished by antiseptic, sterile prose. Civil
damages comprise a genre of compensation, storytelling, and sanction that has
not occurred at the ICTR.70

b. Foreign National or Military Courts
A handful of foreign states have prosecuted Rwandans for genocide, crimes
against humanity, or war crimes. There have been very few such trials.
One example is the prosecution of Fulgence Niyonteze in the Swiss mili-
tary court system. Niyonteze had been a mayor in Rwanda, but was living in
Switzerland at the time the prosecution was initiated. He was accused of order-
ing the massacre of Tutsi and moderate Hutu in his village. On April 29, 1999,
the Switzerland Military Court of First Instance convicted him of the murder
of at least three people, attempted murder, and breach of international con-
ventions (war crimes in internal armed con¬‚ict). The initial sentence was life
imprisonment. On appeal, the sentence was reduced to fourteen years insofar
as a number of convictions for common crimes were quashed.71 In sentencing,
the Military Court of Appeal alluded to retribution, noting the intrinsic gravity
of the crimes. It observed that the murders were committed atrociously and that
the corpses were denied a digni¬ed burial in that they were left abandoned in
latrines. The Military Court of Appeal mentioned Niyonteze™s leadership role as
mayor, noted his coldness and hatred (grande froideur, haine), and that he failed
to express any remorse or empathy. The court did raise in mitigation, as the East
Timor Special Panels have done on occasion, that Niyonteze faced a chaotic
situation that left him with only limited room for decisionmaking. Niyonteze
also was found to have saved the lives of some people who were close to him,
for whom he had produced false documents.
Proceedings by foreign courts can obstruct Rwanda™s wish to prosecute alleged
perpetrators at home. For example, Rwanda insists on bringing Wenceslas Mun-
yeshyaka, a Catholic priest, to trial in Rwanda for his alleged involvement in
atrocities in Kigali.72 Yet Munyeshyaka is in France, where authorities are in the
process of prosecuting him. The French prosecutions (which began in 1995)
have been dismissed, appealed, and now restarted. Considerable controversy
has arisen with regard to the question whether the French courts are competent
to try crimes committed by a foreigner against foreigners in a foreign country.
Questions of Munyeshyaka™s responsibility have not yet been addressed. These
complex jurisdictional questions would not arise were the proceedings to take
place in Rwanda. In the interim, the accused continues to perform his pastoral
duties in a parish near Paris. The desire by French courts to prosecute Mun-
yeshyaka has deferred assessment of his actual responsibility and, by triggering
this lengthy delay during which time the defendant exercises his vocation, has
diminished the severity of whatever punishment might ensue.
In 2001, a Belgian jury found four Rwandans resident in Belgium, including
two nuns, guilty of multiple national and international crimes.73 The court
sentenced them to terms of twenty years, ¬fteen years, and two to terms of twelve
years each. Although references were made to Belgian domestic law, no explicit
84 Atrocity, Punishment, and International Law

elucidation was made of penological rationales as grounds for the variations in
the sentences. The case contains one reference to a mitigating circumstance for
each of the defendants: no evidence of prior criminal convictions.74 This seems
a fairly odd factor to consider in mitigation. Doing so effectively implies that
extraordinary international criminals ought to see the severity of their sanction
diminished because they had not been convicted of ordinary common crimes
either before or after their acts of extreme evil. Given the multiple nature of
the crimes, and the absence of any confession, plea, or remorse, the Belgian
sentences are comparatively light in relation to what domestic Rwandan courts
and the ICTR would issue. The defendants also faced civil sanction.


. 3
( 11)