Putting “Purpose” in Context

Dr Sarah Finnin is the author of the book Elements of Accessorial Modes of Liability: Article 25(3)(b) and (e) of the Rome Statute of the International Criminal Court (Martinus Nijhoff, 2012). She is presently an Associate Legal Officer at the ICTY

Nema Milaninia is a Legal Officer, Office of the Prosecutor, International Criminal Tribunal for the former-Yugoslavia.

 The views expressed herein are those of the authors alone and do not necessarily reflect the views of the International Criminal Tribunal for the former Yugoslavia or the United Nations in general.

The terms of Article 25(3)(c) of the Rome Statute diverge from the standard definition of the mental element required for aiding and abetting under customary international law. Article 25(3)(c) requires that an accused act “[f]or the purpose of facilitating the commission” of a crime. In doing so, it provides for a mental element different from, and in addition to, the “knowledge” or “intent” requirements as defined in Article 30 of the Statute. Article 25(3)(c) echoes the approach originally developed by the American Law Institute in its Model Penal Code (“MPC”), which requires that an accomplice act “with the purpose of promoting or facilitating the commission of the offense” [Section 2.06(3)(a)(ii)]. By contrast, customary international law, as reflected in the jurisprudence of the ICTY [Šainović et al AJ, para. 1649], ICTR [Kalimanzira AJ, para. 86], SCSL [Taylor AJ, para. 436] and ECCC [Chea and Samphan TJ, para. 704], requires only that the accused know that his or her acts “assist in the commission of the offense”.

Given that the term “purpose” is not defined in the Rome Statute, the International Criminal Court (“ICC” or “Court”), will have to interpret the term in accordance with Article 21 on applicable law, and with general principles of treaty interpretation as set out in the Vienna Convention on the Law of Treaties (“VCLT”). Under Article 31 of the VCLT, the terms of Article 25(3)(c) must be interpreted “in good faith in accordance with [their] ordinary meaning”, “in their context and in the light of [the Rome Statute’s] object and purpose”. That “object and purpose” will, in our view, include the limitations on the Court’s jurisdiction stemming from the principles of gravity and complementarity [Rome Statute, Articles 1, 17(1)(d), 53(2)(c)].

Article 25(3)(c) represents a political compromise resulting from disagreement amongst common law and civil law representatives at the diplomatic conference for the establishment of the ICC, who had difficulty agreeing on the most appropriate means of limiting the scope of application of Article 25(3)(c) in the specific context of the ICC [Taylor AJ, para. 435; Scheffer Amicus Curiae Brief in John Doe v. Nestle, S.A., No. 10-56739, pp. 11-13]. The ICC was never intended to prosecute the full range of individuals who make a culpable contribution to the commission or attempted commission of a crime within the jurisdiction of the Court [Rome Statute, Article 17(1)(d)]. The Court was designed to function in a manner that is complementary to national criminal jurisdictions, which themselves maintain the primary responsibility for trying the vast majority of perpetrators.

This is not to say that the approach adopted in Article 25(3)(c) was an appropriate means for limiting the scope of cases which come before the ICC. In fact, an additional “purpose” requirement is problematic for a number of reasons. For example, it might have the effect of protecting individuals from liability where they take advantage of situations of armed conflict for financial gain, knowing that their conduct makes a substantial contribution to the commission of international crimes. Though one may be tempted to look to the MPC in an effort to resolve some of these problems, the Court may only have reference to the MPC itself, its commentaries, or the interpretation of MPC under domestic law, in one of two ways: (i) as a supplementary means of interpretation in accordance with Article 32 of the VCLT, to the extent that they informed the “preparatory work of the [Rome Statute] and the circumstances of its conclusion”; or (ii) as just one source, amongst others, from which the Court might derive a general principle of law under Article 21(1)(c) of the Statute.

It may not be necessary, however, to have recourse to the MPC to develop a workable interpretation of “purpose”. There is scope for the Court to interpret the “purpose” requirement broadly, and in a manner that minimises the divergence from customary international law. First, as a matter of evidentiary proof, where there is evidence that an accused had knowledge that his or her conduct would facilitate the commission of a crime, and nevertheless engaged in that conduct, the Court could infer that the accused acted for the purpose of facilitating the commission of that crime. Second, the Court could interpret the terms of Article 25(3)(c) as requiring only that facilitating the commission of a crime be a purpose of the accused’s conduct, but not the sole purpose. For example, evidence of a financial motive would not itself exclude a finding that the accused also acted for the purpose of facilitating the commission of a crime.

Regardless of the interpretation ultimately adopted by the Court, Article 25(3)(c) was not intended to reflect State practice and opinio juris and thus codify customary international law. Article 10 of the Rome Statute itself provides that its provisions should not “be interpreted as limiting or prejudicing in any way existing or developing rules of international law” for purposes other than cases directly before the Court. As David Scheffer, head of the U.S. delegation to the Rome Conference, has recently argued before the U.S. Ninth Circuit Court of Appeals, “[w]hile some other articles of the Rome Statute ended up reflecting customary international law, Article 25(3)(c) is not one of them.” [Scheffer Amicus Curiae Brief in John Doe v. Nestle, S.A., No. 10-56739, p. 9].

Yet ironically, in some instances, Article 25(3)(c) has been treated at the domestic level as if it were a reflection of customary international law. This has occurred in two ways: (i) through the direct incorporation of the Rome Statute (including Article 25) into domestic legislation by States implementing their obligations under the Statute; and (ii) where domestic courts improperly rely on Article 25 as a source of customary international law regarding individual criminal responsibility in their own jurisprudence (for example, in U.S. Alien Tort Statute litigation where reliance on Article 25(3)(c) is having a real and immediate impact on the scope of corporate liability for aiding and abetting international crimes) [Aziz v. Alcolac, Inc., 658 F.3d 388, 399–400 (4th Cir. 2011); Presbyterian Church of Sudan v. Talisman Energy, Inc., 582 F.3d 244, 259 (2d Cir. 2009)].

It would be wrong for States to limit their jurisdiction over international crimes in this way. Eschewing the recognised standard under customary international law in favour of the political compromise contained in Article 25(3)(c) would create an “impunity gap” which the Rome Statute’s principle of complementarity was designed to avoid. The very purpose of complementarity is that States maintain primary responsibility for prosecuting international crimes. It does not mean that the jurisdiction of States needs to mirror that of the ICC. The ICC was envisioned as an additional tool in the fight against impunity that would exercise jurisdiction over only those cases of most serious concern to the international community as a whole. States, however, can and should maintain the customary international law standard for aiding and abetting, so as to ensure that all parties who contribute to the commission of international crimes are held accountable.

Incorrect or Inappropriate: The ICC’s “Purpose” Standard of Complicity

Dr Flavio Noto is the author of the book Secondary Liability in International Criminal Law: A Study on Aiding and Abetting or Otherwise Assisting the Commission of International Crimes (DIKE, 2013)

As James Stewart correctly highlights, most scholars interpret “purpose” as the mental element for aiding and abetting in Article 25(3)(c) as requiring a volitional commitment to the criminal outcome, a mens rea standard also known as dolus directus in the 1st degree. Three lines of argumentation can be discerned:

Some authors claim that the purposive motivation requirement balances the low objective threshold. That reasoning is difficult to uphold given that the Lubanga Trial Chamber interpreted Article 25(3)(c) as requiring substantial effect (even though it did so in an obiter dictum). Other commentators appear to view Article 25(3)(c) in the light of domestic doctrines restricting the scope of aiding and abetting by means of an elevated mens rea requirement. Lastly, a variety of scholars derive a dolus directus in the 1st degree threshold from their reading of Article 2.06 MPC, on which Article 25(3)(c), they claim, was based on.

To interpret Article 25(3)(c) as requiring a volitional commitment to the criminal outcome has drawn a fair share of criticism on policy grounds. Some commentators correctly point out that a dolus directus in the 1st degree threshold renders the Statute’s provision on aiding and abetting virtually futile; in the case of material support it would hardly be possible to prove that the aider and abettor acted for more than just financial profit. Also, it is argued, to privilege cold-blooded profit maximising vis-à-vis ardent facilitation is not only mistaken from a criminological point of view but also contrary to the object and purpose of the Rome Statute. These criticisms mirror the identical discussion on the suitable mens rea standard for aiding and abetting in domestic criminal law.

To give way to these policy considerations may indeed be in accordance with the object and purpose of the Rome Statute. The trouble is, though, that Article 31(1) of the Vienna Convention on the Law of Treaties commands to give precedence to the ordinary meaning of the wording. Purpose may not necessarily be tantamount to a volitional commitment to the result of the crime but it certainly appears semantically closer to that position than to a mere knowledge standard.

The somewhat simplistic lowest common denominator in legal literature is that purpose requires “more than knowledge”. While this truism is unhelpful, there is merit in suggesting that proof of certain knowledge fulfills the mens rea required by Article 25(3)(c):

Dubber claims that the aider and abettor’s commitment to the criminal outcome can be derived from his certain knowledge about the facilitating effect his assistance has on the crime. This reasoning not only follows a similar logic applied in domestic criminal law but also the logic applied in Kvočka to explain the way in which a knowing facilitator turns into a participant in a JCE sharing the intent of his or her co-perpetrator. Due to the fine line between knowledge and shared intent the trouble with this logic is to determine at which point in time the facilitator’s knowledge crosses the threshold.

Taking this observation a step further, Farrell asserts that it is unsettled whether “the purposive element requires the aider and abettor to act with the sole purpose for contributing to the commission of a crime“. Cassel and Vest therefore contend that the aider and abettor aims at facilitating the crime, except that this is not his or her primary but “secondary purpose”, that is, a “prerequisite necessarily linked” to his or her primary purpose – the latter being, for instance, financial gain.

This line of argument mimics the fiction – meticulously examined by Stuckenberg – that certain knowledge about an undesired but anticipated side-effect is tantamount to a volitional commitment to that side-effect (think of the well known textbook example of the assassin who shoots down an airliner with the aim to kill a specific person sitting in the cramped passenger rows). Applying this fiction to aiding and abetting, Vest points out that it makes limited sense “that someone who knows that a specific result is an inevitable consequence of his conduct should be treated in a different manner than someone who consciously desires this result”. In other words, the intent to contribute to a crime must be separated from the underlying motives and desires. British criminal law reflects this when purporting that an act of assistance is intended both “if it is chosen as an end in itself, or as a means to an end.” Thus, the aider and abettor who is aware or foresees that the realization of the success of the crime is a necessary or foreseeable consequence of his act of assistance acts with the purpose of facilitating the crime. After all, as a comparison of the wording of Article 25(3)(d)(i) and Article 25(3)(c) shows, the Rome Statute itself seems to make a difference between contributing “with the aim of furthering” a crime and assisting “for the purpose of facilitating” it.

One point of criticism concerning a certain knowledge-standard can be expected to be that it is not in accordance with the MPC, which is widely regarded to have served as the basis for Article 25(3)(c). Two points should be highlighted:

First, the fact that Article 25(3)(c) reflects the MPC’s provision on aiding and abetting does not imply that it was the intention of the drafters of the Rome Statute to follow the MPC’s interpretation on purposive facilitation.

Second, as James Stewart rightly points out, is questionable whether the MPC’s provision on aiding and abetting requires the facilitator to desire the criminal outcome. In some US jurisdictions the purpose requirement appears to have been interpreted in a way that the accessory must possess the full mens rea required of the perpetrator. However, Mueller, Weisberg and Weiss have adduced authorities showing that in other jurisdictions purpose is fulfilled by proof of mere knowledge, triggering the question as to the circumstances under which knowledge is not tantamount to purpose.

Third, it is questionable why the MPC’s tests should apply within the completely different framework that governs the ICC. Commentators have shown that in some MPC jurisdictions a looser view that in reality made certain knowledge sufficient has been put forth with respect to serious crimes, whereas a higher mens rea standard has primarily been reserved for less serious ones. Furthermore, to require purpose for aiding and abetting appears to have been a means to justify or excuse the aider and abettor due to mistake of fact or to redress over-inclusion. Also, scholars have pointed out that the case for a dolus directus in the 1st degree standard was strong in jurisdictions granting no punishment discount to accessories and even stronger in jurisdictions barring the jury to lower the sentence for minor involvement in a crime. The Rome Statute, unlike the MPC, addresses these issues. That makes a volitional commitment requirement for aiding and abetting redundant and inappropriate.

What is the ICC’s Standard for Complicity Really?

Complicity, or aiding and abetting, has attracted a great deal of attention of late in the practice of international criminal justice, theoretical literature on the topic and within the blogosphere. Unfortunately, several younger scholars who have written excellent books on the topic haven’t had the opportunity to weigh in, so I wanted to give them an opportunity to advertise their great work and take part in the respectfully critical enterprise I am initiating here.

I also wanted to use this opportunity to raise a question that has not attracted much attention elsewhere, namely, how do we interpret the reference to “purpose” in the ICC Statute’s definition of aiding and abetting? In that regard, I recently stumbled over the fact that I and, I suspect, many others in ICL have misinterpreted the provision, but I wanted to put that idea to scholars with real expertise on these issues. Thus, I have invited the following three authors to respond to the question I set out further below:

Dr Flavio Noto – Under the direction of Hans Vest and others, Flavio authored an excellent thesis entitled Secondary Liability in International Criminal Law: A Study on Aiding and Abetting or Otherwise Assisting the Commission of International Crimes (DIKE, 2013).

Screenshot 2014-12-04 15.13.34


Dr Sarah Finnin – Sarah wrote a fantastic book for her doctoral thesis entitled Elements of Accessorial Modes of Liability: Article 25(3)(b) and (e) of the Rome Statute of the International Criminal Court (Martinus Nijhoff, 2012). She is presently an Associate Legal Officer at the ICTY;

Finnin Book Coverpage


Dr Cassandra Steer – Cassandra is about to publish a great book entitled Translating Guilt: Identifying Leadership Liability for Mass Atrocity (T.M.C Asser Press, 2015). She is a Lecturer and Researcher at the Amsterdam Centre for International Law, but about to join McGill.

Steer - Book Cover

The Question

 To recall, Article 25(3)(c) of the ICC Statute states that:

“In accordance with this Statute, a person shall be criminally responsible and liable for punishment for a crime within the jurisdiction of the Court if that person:…  (c) For the purpose of facilitating the commission of such a crime, aids, abets or otherwise assists in its commission or its attempted commission, including providing the means for its commission”

The received wisdom is that the reference to purpose requires a volitional commitment to the consummated offense, which makes the ICC standard higher than the position of knowledge in other international tribunals. In previous work, I have assumed as much, but I recently discovered that this is at least incomplete and potentially incorrect.

In a recent article I authored for Markus Dubber and Tatjana Hörnle’s Oxford Handbook of Criminal Law, which I creatively titled Complicity, I set out in greater detail why I suspect that the received wisdom about purpose in the ICC Statute might be misguided. To summarize (see pages 21 – 25 of the article), these reasons include:

  • The ICC definition follows the US Model Penal Code – The ICC purports to follow the US Model Penal Code (MPC), but the MPC contains a provision immediately after its famous reference to purpose, indicating that “[w]hen causing a particular result is an element of an offense, an accomplice in the conduct causing such result is an accomplice in the commission of that offense if he acts with the kind of culpability, if any, with respect to that result that is sufficient for the commission of the offense.” (see page 22 of the article)
  • Consequently, the ICC should arguably be interpreted in keeping with the US Model Penal Code – Several authors have argued that the ICC Statute should be interpreted like the MPC. If so, would this not entail reading the second aspect of aiding and abetting in the MPC into the ICC Statute? (see footnote 59 of the article). Note, in fairness to Sarah Finnin, her reference to this idea in her book is quoting Elewa Badar, “The Mental Element in the Rome Statute of the International Criminal Court: A Commentary from a Comparative Criminal Law Perspective” (2008) 19 Criminal Law Forum 473, 507.
  • The drafting history of the ICC Statute confirms this reading – David Scheffer and Donald Piragoff, who were involved in negotiating the provision in question for the U.S. and Canada respectively, indicate that, in essence, the reference to purpose was really meant to be synonymous with intention. Piragoff, for instance, has written that “aiding and abetting by an accused requires both knowledge of the crime being committed by the principal and some intentional conduct by the accused that constitutes the participation.” (See page 352 of this article)
  • National understandings of “purpose” for aiding and abetting do not require a volitional commitment to the outcome of the crime – the relatively few national systems that adopt purpose standards for aiding and abetting do not interpret the term as requiring a volitional commitment to the consummated crime. These include New Zealand, Canada, Israel and perhaps most significantly, the United States (See page 25 of the article).

More precisely:

US Federal Criminal Law – In a recent US Supreme Court decision, the majority held that “for purposes of aiding and abetting law, a person who actively partici­pates in a criminal scheme knowing its extent and charac­ter intends that scheme’s commission.” Rosemond v. United States, 134 S. Ct. 1240, 1243 (2014) (emphasis added);

US State Law – In a very helpful recent doctrinal survey of state law in the United States, John Decker reports that only three (3) states within the United States have purpose standards for aiding and abetting that require a volitional commitment to the criminal outcome. See John F. Decker, The Mental State Requirement for Accomplice Liability in American Criminal Law, 60 S. C. L. Rev. 237 (2008).

Israel – also a “purpose” jurisdiction, but the leading case stipulates that “where the aider only foresees the possibility of the commission of the principal offense, the aider may be convicted if it is his or her desire that should the offense actually be committed, his or her act will facilitate its commission.” Itzhak Kugler, Israel, in The Handbook of Comparative Criminal Law 352, 370 (Kevin Jon Heller & Markus Dubber eds.) (citing the Israeli Supreme Court case of CA. 320/99 Plonit v. State of Israel 55(3) PD 22 [1999];

Canada – although the criminal code stipulates purpose as the required mental element for complicity, the leading case on the proposition states that “Parliament’s use of the term ‘purpose’ in s. 21(1)(b) should not be seen as incorporating the notion of ‘desire’ into the mental state for party liability, and that the word should instead be understood as being essentially synonymous with ‘intention’”. R v. Hibbert [1995] 2 S.C.R. 973.

New Zealand – § 66(1) of the Crimes Act 1961 states that “[e]very one is a part to and guilt of an offence who… (b) does or omits an act for the purpose of aiding any person to commit the offence” (emphasis added), but the Supreme Court interprets this in the following terms: “A party will be liable as an aider and abettor only if he or she had knowledge of the essential matters constituting the offence.” Mahana Makarini Edmonds v. R. [2011] NZSC 159, para. 25 (emphasis added)

Given this history, I welcome critical engagement with the idea that the ICC standard for aiding and abetting should not be interpreted as requiring a volitional commitment to the criminal outcome. If so, what does purpose mean?

The Role and Weight of Desert

Sasha Greenawalt, Professor of Law, Pace Law School. He is the author of the excellent new article International Criminal Law for Retributivists.

I am grateful to Adil for his insightful comments.  Adil was quite generous to serve as a discussant when I presented an early draft of my article at the 2013 Law and Society Association Annual Meeting and so am I especially thankful that he has taken the time to participate in yet another exchange.

Adil agrees that retributivism is an incomplete theory of punishment and that non-retributive considerations can outweigh retributive reasons to punish.  Adil’s takes issue, however, with the particular role and weight that I assign to retributive considerations.  In so doing, Adil raises several important points, some of which involve nuances that my article does not expressly address. Nevertheless, I believe that his points are consistent with my argument, and that my analysis can, or already does, incorporate them.

At the outset, it’s important to emphasize an important difference in focus between Adil’s comments and my article.  Unlike Adil, I am not interested in defending the best view of retributivism.  Instead, I am primarily focused on exploring how different understandings of retributivism could approach international criminal law. In particular, what I describe as a “good reason retributivism,” is my attempt to identify how retributivism can operate in the real world as a plausible, affirmative rationale for punishment.  As I explain in my response to Mark Drumbl, the label reflects my attempt to distill something that is already present in retributive theory rather than to propose a novel approach to punishment.   In other words, retributivists might disagree about whether there is a moral obligation to punish the deserving, or about whether deserved punishment is a good, while nevertheless agreeing that, practically speaking, retributive reasons operate along the lines that I have described.

Adil’s principal critique is that “I view moral responsibility for past wrongdoing [as] merely one good reason to punish among others, a reason that competes on equal terms with good reasons not to punish.”   On Adil’s account, by contrast, desert is the only reason to punish.  Utilitarian considerations favoring punishment play a more limited role:  they can defeat utilitarian considerations disfavoring punishment (thereby defeating the defeaters), but they cannot provide an affirmative reason to punish.

This is an elegant and interesting way to put it, but am I having trouble identifying the difference between Adil’s “only reason retributivism” and what he describes as my “one good reason retributivism.”  On my account, desert is both necessary to punishment and provides an affirmative, prima facie reason to punish.  (Hence, I am not sure why Adil believes that I embrace a purely “negative retributivism” that lacks this affirmative function.)  My good reason retributivism does not contemplate that utilitarian considerations could justify punishment all by themselves without support from retributive considerations.  As with Adil’s approach, “retributive reasons . . . serve a unique and indispensible function in the justification of punishment.”

How then does Adil’s account differ from my own?  Perhaps the difference lies in the fact that I, unlike Adil, do not specify that utilitarian considerations favoring punishment can only act as “defeaters of defeaters,” rather than as affirmative reasons to punish.  I’m not sure how much this distinction matters, however.  If desert provides a necessary, prima facie reason to punish, then what role could these other considerations play except to help defeat arguments opposing punishment?  If there is a conceptual difference, I’m not sure it has any practical impact.

Another question concerns the relative weight of retributive reasons.  Adil contrasts his “strong retributivism” with my “weak retributivism.”  My own view is there is room for disagreement regarding how much weight retributive arguments should carry in the face of countervailing non-retributive reasons.  I’m not sure how Adil’s framework provides any greater clarity on this point.  Take, for example, Douglas Husak’s observation that the value of punishing the deserving is arguably, from the start, outweighed by the “inevitab[ility] that the practice of punishment will suffer from (at least) each of the following three deficiencies: It will be tremendously expensive, subject to grave error, and susceptible to enormous abuse.”  This argument would seem to be entirely compatible with Adil’s approach:  In Adil’s terms, it could well be that retributive reasons are never enough to overcome these three defeaters, and that the practice of punishment will always require additional non-retributive defeaters of defeaters to support the retributive reason to punish.

My idea that desert might merely play a tie-breaking role in justifying punishment was inspired by Husak’s example.  My point is that even if one thinks that retributive arguments are, by themselves, readily defeated by the negative consequences of punishment, they can play still a powerful role in situations where the balance of non-retributive considerations both favoring and disfavoring punishment yields no clear answer.  It strikes me that international criminal law often involves uncertainties of this nature.  Of course, Adil is correct that if the retributive reasons to punish are too weak, then they cannot play even this tie-breaking role.

Otherwise, I very much appreciate Adil’s reflections on both consequential retributivism and the distributive component of retributive justice.  These will require further reflection, but I do find them compelling.

The Point of International Criminal Justice

Adil Ahmad Haque, Professor of Law and Judge Jon O. Newman Scholar, Rutgers School of Law, Newark

Alexander Greenawalt (“Sasha” to those lucky enough to know him) begins his article, International Criminal Law for Retributivists, with Mirjan Damaska’s famous question: “What is the point of international criminal justice?”  This general question conceals a number of more specific questions:

  • What is the point of international crimes?  Why not rely on domestic criminal law to prohibit serious violence?
  • What is the point of international criminal tribunals?  Why not leave the prosecution of international crimes to national courts?
  • What is the point of international criminal prosecution?  Why prosecute international crimes committed in some conflicts but not others, some international crimes committed in a given conflict but not others, and some of those responsible for a given international crime but not others?
  • What is the point of international criminal punishment?  Why not deal with international criminals through lustration, compensation, or the exchange of apology and forgiveness?
  • What is the point of international criminal sentencing?  Why punish some international criminals more severely than others?

In principle, these questions might have different answers.  The point of international crimes might be expressive, the point of international tribunals preventive, the point of international sentencing retributive, and so on.

As I read him, Sasha believes that all such questions rest on a false premise.  There is no aspect of international criminal justice that has a single point; all serve multiple purposes and are subject to multiple constraints.  In each domain, the relevant actors must simply weigh the various relevant considerations—retributive, expressive, or consequentialist—on both sides of a given policy issue.

On Sasha’s view, the fact that most international criminals deserve punishment is simply one good reason among others to create new international crimes; establish new tribunals; prosecute particular conflicts, crimes, or defendants; punish particular defendants; or set particular sentences.  In each domain, there may be opposing, non-retributive reasons not to proceed—to conserve resources, to permit conflict resolution, to facilitate post-conflict reconciliation, and so on.  The right course of action therefore depends on the overall balance of competing reasons.  Since the balance of reasons is a contingent matter, Sasha concludes that retributivists are not committed to any particular position on any of the pressing policy questions facing the ICL regime.  Retributivists might even support general amnesty for international crimes if that is what the balance of reasons supports.


In one way, Sasha is clearly right.  The retributive duty to punish wrongdoers is not absolute.  To that extent, the right answer to any ICL policy question indeed depends on the relative strength of relevant non-retributive considerations.

Nevertheless, I reject the view that moral responsibility for past wrongdoing is merely one good reason to punish among others, a reason that competes on equal terms with good reasons not to punish.  On my view, only retributive reasons affirmatively justify punishment.  The expressive value and consequential benefits of punishment contribute to the justification of punishment only indirectly, by offsetting the expressive disvalue and consequential costs of punishment.  The expressive value and consequential benefits of punishment are not reasons to punish (“favorers”) but merely reasons to discount certain reasons not to punish (“defeaters of defeaters”).  Retributive reasons to punish then compete with the remaining, undefeated expressive and consequentialist reasons not to punish.  Put another way, retributive reasons compete, not with the “gross” costs of punishment, but only with the “net” costs of punishment.  On my view, retributive reasons can still be outweighed, but they serve a unique and indispensible function in the justification of punishment.

Why do only retributive reasons directly count in favor of punishment?  Ordinarily, it is morally wrong to intentionally harm other people.  However, people can make themselves morally liable to suffer certain harms for certain purposes.  Importantly, the reasons that make it permissible to harm someone also limit the amount of harm that may be inflicted as well as the purpose for which it may be inflicted.

For example, those who are morally responsible for an unjust threat to others (i) forfeit their right not to be (ii) defensively harmed as a causal means of averting that very unjust threat (iii) so long as the harm they suffer is not substantially greater than the harm they threaten.  Notice that forfeiture, purpose, and proportionality share a common foundation.   Similarly, on my view, those who are morally responsible for an unjust harm (i) forfeit their right not to be (ii) punitively harmed as a constitutive means of punishing that very unjust harm (iii) so long as the harm they suffer is not substantially greater than the harm they inflicted (discounted by their degree of fault).

In contrast, Sasha embraces “negative retributivism”, the view that those who are morally responsible for an unjust harm (i) forfeit their right not to be (ii) instrumentally harmed as a causal means of producing various intrinsic goods (iii) so long as the harm they suffer is not substantially greater than the harm they inflicted discounted by their degree of fault.  This account of moral liability to punishment has always struck me as incoherent.  We should instead accept that moral responsibility for past wrongdoing is the basis, purpose, and limit of moral liability to punishment.


Sasha at times suggests, not only that past wrongdoing is merely one good reason to punish among others, but also that past wrongdoing is not a particularly strong reason to punish.  At one point, Sasha considers the possibility that retributive reasons may play only a “tie-breaking” role.  In this passage, Sasha refers not to (imaginary) situations in which we know that the future costs and benefits of punishment will be on a par, but rather to (common) situations in which we cannot reliably predict the future costs and benefits of punishment.  However, if we cannot reliably predict that the benefits of punishment will outweigh the costs, then we probably also cannot predict that the benefits of punishment plus the (apparently minor) retributive value of punishment will outweigh the costs.  So even this tie-breaking role seems illusory.

ICL detractors like to point out that the Special Court for Sierra Leone (SCSL) cost over $200,000,000 and punished only 9 defendants of international crimes.  Surely, the argument goes, we should not spend over $20,000,000 simply to give one wrongdoer the punishment he or she deserves—not when we could use that money to give many children the education they deserve, many patients the treatment they deserve, and so on.

Such arguments are particularly effective when they take on board the “consequentialist retributivism” with which Sasha sometimes flirts.  On this view, deserved punishment is an intrinsic good that we should try to maximize.  The retributive value of the SCSL is just the amount of deserved punishment it produced.  Since the SCSL only produced around 350 years of deserved incarceration (at a cost of over $500,000 per year), it can seem like a waste of money.

Fortunately, consequentialist retributivism is false. I have argued elsewhere that deserved punishment is not an intrinsic good.  Certainly, it should not be maximized.  Consider the widely accepted principle that international tribunals should prosecute those “most responsible” for international crimes.  Evidently, following this principle will not maximize deserved punishment.  Since the “most responsible” are mortal, there is a physical limit to how many units of deserved punishment we can inflict upon them.  Indeed, generally we can produce more units of deserved punishment by punishing a handful of minor participants than by punishing the mastermind they served.  Yet such a distribution of deserved punishment would be unjust.

The distributive component of retributive justice remains largely unexplored.  I suspect that retributivists generally should first seek to vindicate the rights of as many victims as possible and only then seek to vindicate these rights as fully as possible.  Forced to choose, generally we ought to punish fewer of those responsible for more wrongs rather than punish more of those responsible for fewer wrongs.  In general, it is more just to (partially) vindicate the rights of more victims rather than (more fully) vindicate the rights of fewer victims.  There may be extreme cases in which full vindication of fewer victims can take precedence over partial vindication of more victims.  But it is the vindication of rights, not deserved punishment, that retributivists should seek, not to maximize, but to justly distribute.

It follows that ICL ought to prioritize those “most responsible” for international crimes because generally this will vindicate the most victims of the most serious wrongdoing.  Since the SCSL vindicated the rights of tens of thousands of victims of extreme wrongdoing, it was more than worth the cost.  At a few thousand dollars per victim vindicated, this was justice at a bargain.


Sasha calls his view “good reason retributivism”.  We could instead call it “one good reason retributivism” or “weak reason retributivism”.  We could call my view “only reason retributivism” or “strong reason retributivism.”  On my view, the point of international criminal justice is to ensure just retribution for serious wrongdoing.  Non-retributive considerations will inevitably intrude but they will always remain, in an important sense, beside the point.


Vive La Rétributivisme Timide!

Sasha Greenawalt, Professor of Law, Pace Law School. He is the author of the excellent new article International Criminal Law for Retributivists.

I am grateful to Mark Drumbl for taking the time to provide his generous and insightful comments on my article, and also to James Stewart, both for his kind words and for organizing this discussion.  I’m quite honored that James has selected my article to inaugurate this important and exciting new blog.

In this post, I respond to Mark’s critiques, which have pushed me to think hard and harder, and even to consult Google Translate.  As I understand it, Mark is disappointed by my fainthearted retributivism, which he deems anemic, indeterminate, overly intellectual, and woefully lacking in hot-blooded passions.  He objects that “retributive theory cannot simply become all things to all persons.”

For the most part, I am happy to plead guilty as charged.  After all, the idea that retributive commitments are compatible with a wide range of policy positions about international criminal law is one of the main points of my article.  In defense of this view, I will say the following.  First, I believe that I am speaking the truth about retributivism.  Accordingly, it is not my fault if the theory turns out to be less muscular than one might hope.  Second, I don’t think it’s quite as bad as all that.  There are many things that retributivism cannot do, but some quite important things that it can do.

The central point (which I am hardly the first to make) is that retributivism, by its nature, is an incomplete theory of punishment.  In contrast to utilitarian approaches that appeal to a single overarching vision of human well-being, the retributive claim that punishment is justified intrinsically by the desert of the offender simply cannot provide a comprehensive metric around which to establish, order, and administer a system of criminal justice.  The point is just as true at the domestic level as it is at the international level.  The most obvious example is resource allocation:  a commitment to deserved punishment cannot, by itself, tell a legislature how to distribute funds between health care and the criminal justice system.  Nor can it tell the international community whether, and to what extent, to fund the establishment of an international criminal tribunal.

Instead, I believe that retributivism does two things.  First, by specifying that justified punishment must be deserved, it imposes an important negative constraint:  a prohibition against knowingly punishing the innocent.  This negative constraint is not especially controversial, but even here retributive theory raises an important and problematic question:  how do we know that the guilty actually deserve punishment?  Mark suggests that the retributive appeal to just deserts is at root an emotional response, and I think he may be right.  Does this reliance on bare, moral intuitions—ones that may well be “ignorant, biblical, and simple,”—expose a deep vulnerability in retributive theory?  Quite possibly, but if I may attempt to paraphrase Douglas Husak, the alternative is even more problematic:  How can we ever justify the cruelty of criminal punishment if we do not at some root level believe it is deserved?  Perhaps, the very existence of criminal justice is inevitably grounded on these innate retributive emotions, whether we acknowledge them or not.

The second thing that retributivism does is to supply an affirmative reason to punish the guilty, one that appeals to the culpability of the wrongdoer alone rather than to the expected social benefits of punishment.  But because retributivism is an incomplete theory of punishment, the retributive impulse must coexist with other rationales either favoring or disfavoring punishment.  As a consequence, the retributive impulse to punish a particular offender will sometimes be trumped or compromised.  Possible reasons include, among others, a legislative desire to fund priorities other than criminal justice; a prosecutorial judgment that limited resources are better focused on other, more culpable offenders; the ability to obtain the accused’s cooperation in the prosecution of other, more culpable offenders; an unacceptable risk of wrongful conviction in light of evidentiary shortcomings; a disinclination to reward prosecutorial misconduct that has comprised the accused’s due process rights; or, as sometimes happens in the wake of mass atrocities, a desire to support an amnesty agreement or alternative justice mechanism that may itself be justified by a variety of imperatives including the need to conserve limited resources,  to end bloodshed, and to promote social reconciliation.

This second function of retributivism—as a good, but incomplete reason to punish—is what I describe as “good reason retributivism.”  Although the label is my own, it reflects my attempt to distill something that is already present in retributive theory rather than to propose a novel approach to punishment.  And I will venture to suggest that something of this approach appears in Mark’s own post when he maintains that he “remain[s] unconvinced that retribution can overcome the curse of selectivity in international criminal law.”  In other words (and acknowledging apologetically that I may be doing horrible violence to Mark’s actual views), even if there are powerful retributive reasons to prosecute someone before an international criminal tribunal, Mark maintains that those reasons are inadequate to overcome the more powerful, overriding interest in avoiding a system of highly selective, unequal justice.  (And as Mark sets forth in a terrific book, alternative mechanisms may provide a better way to reconcile these two interests).

As far as I am aware, Mark does not propose a metric for determining how much selectivity is too much selectivity.  By contrast, I might argue on retributive grounds that Sudanese President Omar al-Bashir’s alleged role in Darfur atrocities involves culpability of such great severity that it is worth demanding his prosecution before the International Criminal Court (“ICC”) notwithstanding the highly selective focus of that institution.  Both Mark and I may be confident in our judgments, but I doubt we can prove each other wrong.  Yet these are precisely the kinds of moral judgments that people can and do routinely make based on an indeterminate but non-arbitrary balancing of competing considerations.  And I don’t think that the existence of such indeterminacy is tantamount to an “endless buffet line” (however appetizing that analogy may be!).

As I explain in the article, I believe that the retributive rationale plays a powerful role in driving international criminal prosecutions, one that has repercussions throughout the system.  And I believe that the retributive impulse works pretty much as Mark describes: as an intuitive and outraged call for justice.  But I do not believe that embracing this impulse requires one to turn a blind eye to non-retributive values, or to what Mark aptly terms “the explosive paradoxes and dizzying quandaries of international crimes.”

There is much that I am glossing over, including the differences between various understandings of retributivism, the nuances of the international criminal justice system, and important differences among different types of institutional questions (legislative, prosecutorial, judicial etc.).  I must, however, register my disagreement with Mark’s claim that “[e]ye for an eye is central to [retributivism’s] DNA.”  As far as I am aware, lex talionis has no modern defenders among retributivist scholars, although Jeremy Waldron has brilliantly argued that the idea “is (fortunately) not as silly as it sounds.”  The concept of proportionate sentencing is more widely embraced, but, for reasons I explore in the article, I fear it may fail to provide the degree of practical guidance that Mark desires.

Otherwise, I, for one, am not convinced that Mark and I are divided by any deep disagreement.  Perhaps he will prove me wrong, but until then . . . Vive La Rétributivisme Timide!



Adjectival Retributivism: A Tale of Theory as Empath

Mark A. Drumbl, Class of 1975 Alumni Professor of Law & Director, Transnational Law Institute, Washington & Lee School of Law.

Sasha Greenawalt rehabilitates retribution as a justification for punishing people who violate international criminal law. I say “rehabilitate” because Sasha begins with the observation that an “anti-retributivist strain” has seeped into conversations about how to deal with perpetrators of genocide, crimes against humanity, and systematic war crimes. Sasha wants to stanch this strain by cleansing retribution of its undeserved stain.

Sasha hopes to make retributivism relevant in a context where conversations about the normative basis of punishment for extraordinary crimes are vividly diversifying. Penological rationales such as expressivism, positive general prevention, and restoration have entered the discursive frame. The international tribunals tout their deterrent potential, as well as other utilitarian goals (which border on the utopian) such as reconciliation, peace, and ending impunity.

Retribution has its skeptics. I am one of them. Sasha’s excellent work has pushed me to think twice, think differently, and think better, about things. Fundamentally, however, I remain unconvinced that retribution can overcome the curse of selectivity in international criminal law; I am unsure how it squares with the extraordinary nature of jus cogens violations and the cauldron of collective violence; I am unsettled by analogizing too closely from the regulation of ordinary crime at national levels. But all this is beside the point. These debates will not be resolved here. They may not be resolvable. People have been quarrelling about the purposes of punishment and the role of mercy since time immemorial. These quibbles pop up in all places, whether intellectual, vulgar, aesthetic, popular, or political.

I am more interested here in the dialectics. How does retributive theory, which Sasha enlivens, approach and contribute to the conversation about the normative basis of international punishment? How does it seek to call the question, presumably in its favor? What arguments does it marshal?

Sasha responds to “anti-retributivism” by demonstrating that retributive theory is heterogeneous and supple. He rightly resists those who essentialize retribution as primitive or instinctual or crude. He chafes at the bad rap retribution gets in certain quarters. For Sasha, there’s a lot more to retribution than harshness. He shows how retributive theory is nimble, flexible, erudite, and sophisticated. It is so versatile, in fact, that it can accommodate many of the concerns central to utilitarian, and other, justifications for punishment. Here, Sasha takes the reader on a journey through an alphabet soup of adjectives. We learn of consequential retributivism, minimalist retributivism, political retributivism, threshold retributivism, maximal retributivism, and contractarian retributivism. Sasha delivers good reason retributivism. We learn that retributive theory is so capacious that it can justify general amnesties. It can even justify complete impunity!

It is invigorating to peel back the layers, and embrace the hybridized possibilities, of retributive theory. I’ve toyed with this myself in terms of agentic retributivism. By this I mean measuring the culpability of an accused not by the harm caused or malevolence exercised in absolute terms but, rather, by how much harm the perpetrator caused or how much malevolence the perpetrator exercised in light of the social spaces that he or she occupied at the time of the commission of the offense. Interest in this question stems from my concern about how international criminal law should approach compromised perpetrators: rank-and-file soldiers, low-level militia members, militarized youth, and persons who may not have killed directly but told the killers where the oppressed were hiding.

But retributive theory cannot simply become all things to all persons. At some point, all these clever variants of retribution become more about the prefix adjective than the noun. Consequentialist retribution in particular strikes me as a poignant example. The end-game of these concatenations is that retribution risks becoming gutted, or so abstracted, that its raw value as normative justification evaporates. What is gained by becoming an endless buffet line? An account for all seasons?

Also hobbled is the ability of retribution to serve as a practical principle to apply in establishing the quantum of sentence for an individual defendant. This latter point is very important. We should welcome conversations about the normative basis of a punishing institution. But once that institution exists, and it begins to punish individuals, a need arises for principled, clear, and consistent sentencing. Accessibility matters. When retributive theory has so many competing compounds, what is a judge to do and what is a convict to expect?

If retributive theory becomes an empath that absorbs and accommodates everyone’s wishes and needs – even those that contradict its central ethos – then retributivists allow themselves to be defined by others. General debates about the justification for punishment simply morph into feuds within retributive theory about the justification for punishment.

Over-intellectualizing retribution reminds me of the fate of modes of liability at the international tribunals. James Stewart and Leila Sadat have chided the staggering confusion. Modes of liability – rendered both fulsome and bare by relentlessly inordinate determinations – remain clumsy and incoherent, thereby baffling the laity and frustrating survivors in post-conflict transitions.

Call me ignorant, biblical, and simple. Or homely. But for me, retribution is about the emotions.  It is about just deserts and resetting the moral balance following an act that tears the communal fabric. Retributivism is about voicing outrage; determining the culpability of the perpetrator and responding in kind.  Eye for an eye is central to its DNA.  Retributivism means anger, pain, and rectification channeled through a dignified legal process. It’s about expiation. It’s visceral. Why run from this? Staying the hand of vengeance does not require becoming the Tin Man.

When rules become defined by their exceptions, when principles become overrun by a cacophony of qualifiers, then the core fades. Retributivists might be better off sticking with their core. If their cri de coeur doesn’t fit with the explosive paradoxes and dizzying quandaries of international crimes, then tant pis – so be it.

Symposium: Greenawalt’s International Criminal Law for Retributivists

I’m thrilled to host Sasha Greenawalt’s excellent new article International Criminal Law for Retributivists as the inaugural point of discussion for this blog.

In the early years of international criminal justice, Mark Drumbl and Rob Sloane argued, very eloquently, that retribution was not a viable basis for punishment in international criminal law.

Sasha takes issue with their approach, drawing heavily on a rich literature in the theory of criminal law. His article is brilliantly written, profound in scope, and has important real world implications.

At the same time, it will certainly not convince everyone. For that reason, I have cajoled Mark Drumbl, Adil Haque, Rob Sloane and Meg deGuzman to write blogs criticizing Greenawalt’s great piece. Of course, I’ll also invite Greenawalt to reply.

In terms of timing, I’ll post Drumbl and Haque’s responses in the coming days, then ask Greenawalt to respond to them later this week.

Rob Sloane and Meg deGuzman have agreed to reopen the debate in late December this year, so the discussion will take place in two split tranches. We’ll then give Greenawalt the final word.

So, welcome to all my friends and colleagues! It’s a real privilege to have you launching what I hope will become a central platform for these sorts of scholarly debates.

Why Another Blog?

Greetings! I’m glad to welcome you to this new blog. I wanted to begin by setting out what I think the particular identity of this forum will be and to distinguish it slightly from all the other excellent blogs that already exist on international criminal justice and connected fields. Why, in other words, have another blog?

In short, there are several reasons I wanted to begin this blog, many of which track my blogging manifesto, which is intended to act as a guide to guest bloggers. I elaborate on several of these below:

Respectful critique – I wanted to create a space that enabled respectful critique of new thinking in the various fields that interest me. Often I read excellent new research and want to engage quickly with central ideas in that research, without necessarily writing a substantial article in response. The blog provides that platform.

Rapid intervention in practice – I wanted to create a stage for presenting research quickly, in ways that could inform time-sensitive real-world processes. For example, I used EJIL Talk! to post a summary of several years of research into aiding and abetting in international criminal law to influence a judicial debate on these topics. Blogging can replace amicus briefs.

Predominantly theory – I read many very good blogs to get information about latest developments in the field of international criminal justice, but I also wanted to create a blog that would act as a clearinghouse for the best scholarly debates in these fields. I hope this site will be conceptual, eclectic, and scholarly, thereby complementing the other great sites that already exit.

Generating plural intellectual community – I’m not really intending to use this site to broadcast my own views as much as I am hopeful that it will become a place where a wide variety of scholars, from different backgrounds and sensibilities, engage in spirited debate. In a way, the site is an attempt at creating an inclusive intellectual community, instead of thinking of scholarly life in individualistic terms.

Celebrating excellent scholarship – From time to time, we all come across scholarship that’s so good it moves us. For me, Mirjan Damaška’s The Shadow Side of Command Responsibility was one of the first examples of this, and there’ve been several in the many years since. I wanted to create an outlet to celebrate other people’s work that registers at this level, to promote excellence, creativity and insight in fields that interest me.

So, with this short introduction, welcome again to the blog! I can’t guarantee that it will be the busiest site you visit, but I do hope it will be consistently intellectually stimulating.