1 Osiel, M., ‘The Banality of Good: Aligning Incentives against Mass Atrocities’, (2005) 105 Columbia Law Review 1751Google Scholar, at 1764 (quoting from Enrique Gimbernat Ordeig, ‘Autor y complice en derecho penal’ [‘Perpetrator and Accomplice in Criminal Law’] 187 (1966)).

2 It is legally misleading because these doctrines only attribute unlawfulness rather than ‘liability’. The better term is ‘modes of attribution’, since whether a defendant is ‘liable’ once a particular unlawful act is attributed to her requires a further assessment of justifications and excuses. Admittedly, this nomenclature is premised on a preference for the normative theory of guilt. G. P. Fletcher, The Grammar of Criminal Law: American, Comparative, and International, Vol. 1 (2007), 319, 329. In terms of origin, it is also unclear where international criminal justice acquired the term ‘modes of liability’ and why it gained such ascendency in the discipline. Early international judgments used the more appropriate phrase ‘modes of participation’; Prosecutor v. Tadić, Judgement, Case No. IT-94–1-A, Appeals Chamber, 15 July 1999, para. 227 (referring to joint criminal enterprise as a ‘mode of participation’); Prosecutor v. Delić, Judgement, Case No. IT-04–83-T, Trial Chamber I, 15 September 2008, para. 56 (referring to superior responsibility as a ‘mode of participation’); Prosecutor v. Blaškić, Judgement, Case No. IT-95–14-T, Trial Chamber, 3 March 2000, para. 266 (discussing ordering, planning, instigating, or otherwise aiding and abetting as ‘modes of participation’). This accords with the descriptor adopted in most domestic criminal systems. In German criminal law, the overarching concept is Beteiligung, which experts translate as ‘participation’; see M. Bohlander, Principles of German Criminal Law (2008), 154. French criminal theory also refers to criminal participation; see C. Lazerges, ‘La participation criminelle’, in C. Lazerges, Réflexions sur le Nouveau Code Pénal (1995), 11; for historical antecedents, see also B. Getz, De la soi-disant participation au crime (1876); in many Anglo-American jurisdictions, the tendency is to describe modes of liability as those rules that determine parties to crime; see W. R. LaFave, Criminal Law (2010), 701 (employing the term ‘parties to crime’); A. Simester and G. R. Sullivan, Criminal Law: Theory and Doctrine (2007), 195–246 (discussing modes of participation).

3 A large number of international criminal courts expressly profess a commitment to only prosecuting those ‘who bear the greatest responsibility’ for crimes within their jurisdiction. See Agreement between the United Nations and the Government of Sierra Leone on Establishing a Special Court for Sierra Leone (with Statute), Art. 1.1, Sierra Leone–UN, 16 January 2002, 2178 UNTS 137 (‘The Special Court shall, except as provided in subparagraph (2), have the power to prosecute persons who bear the greatest responsibility for serious violations of international humanitarian law and Sierra Leonean law committed in the territory of Sierra Leone since 30 November 1996, including those leaders who, in committing such crimes, have threatened the establishment of and implementation of the peace process in Sierra Leone.’); ICC Office of the Prosecutor, ‘Paper on Some Policy Issues before the Office of the Prosecutor’, September 2003, at 7 (‘The global character of the ICC, its statutory provisions and logistical constraints support a preliminary recommendation that, as a general rule, the Office of the Prosecutor should focus its investigative and prosecutorial efforts and resources on those who bear the greatest responsibility, such as the leaders of the State or organisation allegedly responsible for those crimes’); Extraordinary Chambers in the Courts of Cambodia Internal Rules (Rev. 4) as revised on 11 September 2009, Preamble (‘WHEREAS the Cambodian authorities have requested assistance from the United Nations in bringing to trial senior leaders of Democratic Kampuchea and those who were most responsible for the crimes and serious violations of Cambodian penal law, international humanitarian law and custom, and international conventions recognized by Cambodia, that were committed during the period from 17 April 1975 to 6 January 1979’); for scholarly opinion endorsing this view, see A. Danner, ‘Enhancing the Legitimacy and Accountability of Prosecutorial Discretion at the International Criminal Court’, (2003) 97 AJIL 510; for a more critical assessment, see Alvarez, J. E., ‘Crimes of States/Crimes of Hate: Lessons from Rwanda’, (1999) 24 Yale JIL 365Google Scholar.

4 M. Osiel, Making Sense of Mass Atrocity (2009), 85. The only caveat is that Osiel's comment assumes an objective theory of perpetration, whereby the perpetrator is the person who actually releases the gas into the concentration camps. As we will later see, the objective theory is discredited, but this does not undermine Osiel's point that rank-and-file perpetrators are generally viewed as less culpable than their superiors in international criminal justice.

5 Damaška, M., ‘The Shadow Side of Command Responsibility’, (2001) 49 American Journal of Comparative Law 455, at 456CrossRefGoogle Scholar. Aside from Damaška's excellent article, the critical literature is extensive. For some of the best exemplars, see, in particular, Olásolo, H., ‘Reflections on the International Criminal Court's Jurisdictional Reach’, (2005) 16 Criminal Law Forum 279CrossRefGoogle Scholar; Danner, A. M. and Martinez, J. S., ‘Guilty Associations: Joint Criminal Enterprise, Command Responsibility, and the Development of International Criminal Law’, (2005) 93 CLR 75Google Scholar; Robinson, D., ‘The Identity Crisis of International Criminal Law’, (2008) 21 LJIL 925CrossRefGoogle Scholar; Ohlin, J. D., ‘Three Conceptual Problems with the Doctrine of Joint Criminal Enterprise’, (2007) 5 JICJ 69Google Scholar.

6 Robinson, supra note 5, at 931.

7 There is a considerable and important literature dealing with the doctrine and policy of complicity in international criminal justice but, to the best of my knowledge, none of it explores the objectionable peripheries of the doctrine; see, e.g., Clapham, A. and Jerbi, S., ‘Categories of Corporate Complicity in Human Rights Abuses’, (2000) 24 HICLR 339Google Scholar (explaining three categories of policy implication derived from the application of complicity); L. Lunde, M. Taylor, and A. Huser, Commerce or Crime? Regulating Economies of Conflict (2003) (providing a helpful synthesis of the law of complicity in 16 different jurisdictions); Keitner, C. I., ‘Conceptualizing Complicity in Alien Tort Cases’, (2008) 60 Hastings Law Journal 61Google Scholar (discussing ATCA cases that employ complicity); for a notable exception, see Dubber, M. D., ‘Criminalizing Complicity: A Comparative Analysis’, (2007) 5 JICJ 977Google Scholar (discussing the need for international criminal justice to craft a law of complicity specific for its purposes).

8 The French Criminal Law is a good example of this paradox. Art. 121(6) of the French Criminal Code stipulates that ‘[s]era puni comme auteur le complice de l'infraction’. Simultaneously, leading experts define complicity as ‘un mode d'imputation dirigé contre une personne qui a aidé à la realization d'une situation infractionnelle sans pour autant accomplir elle-même aucun des actes visé par le texte d'incrimination’. J.-H. Robert, Droit pénal général (2005), 343.

9 Gardner, J., ‘“Aid, Abet, Counsel, Procure”: An English View of Complicity’, in Eser, A., Huber, B., and Cornils, K. (eds.), Einzelverantwortung und Mitverantwortung im Strafrecht (1998), 228Google Scholar. L. Steyn, then of the British House of Lords, also put the point succinctly in the Pinochet litigation when he cited ‘an elementary principle of law, shared by all civilised legal systems, that there is no distinction between the man who strikes, and a man who orders another to strike’, L. Steyn in R v. Bow Street Metropolitan Stipendiary Magistrate, ex parte Pinochet Ugarte, 3 WLR 1456 (HL 1998), at 54.

10 Damaška, supra note 5, at 464.

11 Danner and Martinez, supra note 5, at 78 (‘International human rights law, domestic criminal law, and transitional justice. Each one, to varying degrees, informs the purposes and principles of international prosecution, and their interaction creates conflicts within international criminal law itself’); Robinson, supra note 5, at 961 (‘Interpretive, substantive, structural, and ideological assumptions of human rights and humanitarian law have been absorbed into ICL discourse, distorting methods of reasoning and undermining compliance with fundamental principles’); to be fair, Darryl Robinson's excellent piece also points out that this may only be part of the problem and that domestic systems depart from basic principles, too. Robinson, supra note 5, at 927–30; Greenawalt, A. K. A., ‘The Pluralism of International Criminal Law’, (2011) 86 Indiana Law Journal 1111Google Scholar (observing doubts about tribunals’ commitments to core principles of justice ‘that many domestic legal systems take for granted’, and arguing that ‘[w]hile greater reliance on domestic law might not offer a complete solution, it may offer at least one positive step in ICL's rediscovery of a criminal law that better aspires to ICL's liberal aims’).

12 Dressler, J., ‘Reforming Complicity Law: Trivial Assistance as a Lesser Offense?’, (2008) 5 Ohio State Journal of Criminal Law 427, at 427Google Scholar.

13 Tallgren, I., ‘The Sensibility and Sense of International Criminal Law’, (2002) 13 EJIL 561CrossRefGoogle Scholar, at 566–7.

14 Moore, M. S., ‘The Superfluity of Accomplice Liability’, in Moore, M. S. (ed.), Causation and Responsibility: An Essay in Law, Morals, and Metaphysics (2009), 280CrossRefGoogle Scholar; Sullivan, B., ‘Principals and Accomplices: A Necessary and Useful Division?’, in Duff, A. and Wong, C. (eds.), Foundational Issues in the Philosophy of Criminal Law (2007), 151Google Scholar; Kadish, S. H., ‘Reckless Complicity’, (1996) 87 Journal of Criminal Law and Criminology 369CrossRefGoogle Scholar (although Kadish does not advocate for the abolition of complicity, his position is closest to that I advance here); L. Alexander and K. Kessler Ferzan, Crime and Culpability: A Theory of Criminal Law (2009), 7–10 (arguing that insufficient concern is the baseline for all forms of criminal responsibility); D. Kienapfel, Der Einheitstäter im Strafrecht (1971); T. Rotsch, ‘Einheitstäterschaft’ statt Tatherrschaft: Zur Abkehr von einem differenzierenden Beteiligungsformensystem in einer normativ-funktionalen Straftatlehre (2009).

15 Moore, M. S., ‘Causing, Aiding, and the Superfluity of Accomplice Liability’, (2007) 156 University of Pennsylvania Law Review 395Google Scholar. The description of complicity as ‘superfluous’ is overly forgiving of the violations of theoretical principles that complicity presently entails, but, in fairness to Michael Moore, his analysis does not consider the mental element of aiding and abetting where the most conspicuous violations of culpability occur. Moreover, his assessment of the physical element does not deal with standards adopted in international criminal justice, which deviate from basic principles elsewhere. Both these points are explored further below.

16 The countries are Austria, Brazil, Denmark, Italy, and Poland. For further information, see J. Pradel, Droit pénal comparé (2002), 121, at 133; Ambos, K., ‘Development of a Common Substantive Criminal Law for Europe Possible? Some Preliminary Reflections’, (2005) 12 Maastricht Journal of European and Comparative Law 173CrossRefGoogle Scholar, at 182–5 (setting out examples from various unitary jurisdictions).

17 Although the Nuremberg and Tokyo Charters adopted differentiated doctrines of complicity, the majority of cases merely considered whether an accused was ‘concerned in’, ‘connected with’, ‘inculpated in’, or ‘implicated in’ international crimes. For an overview of these cases, see United Nations War Crimes Commission, Digest of the Laws and Cases, Law Reports of the Trials of War Criminals, Vol. 15 (1949), 49–58. Like H. Olásolo, I conclude that this amounts to a unitary theory of perpetration in so far as it fails to distinguish modes of participation. See H. Olásolo, The Criminal Responsibility of Senior Political and Military Leaders as Principals to International Crimes (2010), 21.

18 Keiler, J., ‘Towards a European Concept of Participation in Crime’, in Klip, A. (ed.), Substantive Criminal Law of the European Union (2011)Google Scholar; B. Weißer, Täterschaft in Europa: Ein Diskussionsvorschlag für ein europäisches Tätermodell auf der Basis einer rechtsvergleichenden Untersuchung der Beteiligungssysteme Deutschlands, Englands, Frankreichs, Italiens und Österreichs (2011); W. Schöberl, Die Einheitstäterschaft als europäisches Modell: Die strafrechtliche Beteiligungsregelung in Österreich und den nordischen Ländern (2006); Ambos, supra note 16, at 182–5; for rare exceptions to the rule that scholars do not consider the unitary theory of perpetration for international crimes, see Olásolo, supra note 17, at 14–20; E. van Sliedregt, The Criminal Responsibility of Individuals for Violations of International Humanitarian Law (2003), 61–5.

19 Gardner, J., ‘Complicity and Causality’, (2007) 1 Criminal Law and Philosophy 127CrossRefGoogle Scholar, at 128.

20 To find individual criminal responsibility pursuant to a joint criminal enterprise, the elements that must be established are: (i) a plurality of persons; (ii) the existence of a common plan, design, or purpose that amounts to or involves the commission of a crime provided for in the Statute; and (iii) the participation of the accused in the common plan involving the perpetration of one of the crimes provided for in the Statute. For a particularly recent affirmation, see Prosecutor v. Krajišnik, Judgement, Case No. IT-00–39-A, Appeals Chamber, 17 March 2009, paras. 156–157.

21 Danner and Martinez, supra note 5, at 132.

22 Prosecutor v. Tadić, supra note 2, at 196. Note that this language is not always consistent: see Prosecutor v. Vasiljević, Judgement, Case No. IT-98–32-A, Appeals Chamber, 25 February 2004, para. 97 (finding that ‘[t]he first category is a “basic” form of joint criminal enterprise. It is represented by cases where all co-perpetrators, acting pursuant to a common purpose, possess the same criminal intention’).

23 In JCE II, the common plan in JCE I is merely replaced by ‘an organized criminal system’, such as an extermination or concentration camp. There is, therefore, general consensus that this ‘systematic’ category in JCE II is only a subset of the ‘basic’ form in JCE I; see, e.g., Prosecutor v. Tadić, supra note 2, at 203 (‘this category of cases . . . is really a variant of the first category’); Prosecutor v. Kvočka et al., Judgement, Case No. IT-98–30/1-A, Appeals Chamber, 28 February 2005, para. 82 (describing JCE II as ‘a variant of the first form’); K. Ambos, ‘Amicus Curiae Brief in the Matter of the Co-Prosecutors’ Appeal of the Closing Order against Kaing Guek Eav “Duch” Dated 8 August 2008’, (2009) 20 Criminal Law Forum 353, at 374 (concluding that JCE II can be viewed as an element of JCE I if interpreted narrowly).

24 Prosecutor v. Kvočka et al., supra note 23, at 83.

25 In fact, there is good authority for the idea that the standard is actually objective foreseeability, lowering the mental element required for JCE III even further; see Cassese, A., ‘The Proper Limits of Individual Responsibility under the Doctrine of Joint Criminal Enterprise’, (2007) 5 JICJ 109Google Scholar, at 121 (arguing that most courts actually apply an objective standard of foreseeability for JCE III).

26 Prosecutor v. Tadić, supra note 2, para. 204. For discussion, see Haan, V., ‘The Development of the Concept of Joint Criminal Enterprise at the International Criminal Tribunal for the Former Yugoslavia’, (2005) 5 International Criminal Law Review 167, at 177CrossRefGoogle Scholar. For other arguments regarding the international use of Pinkerton liability, see Fletcher, G. P., ‘New Court, Old Dogmatik’, (2011) 9 JICJ 179Google Scholar, at 179–90.

27 Prosecutor v. Tadić, supra note 2, at 193. In support of this proposition, the Tribunal cited law from France, Italy, England and Wales, Canada, the United States, Australia, and Zambia that also criminalizes a version of JCE III; ibid., at 224.

28 Ambos, K., ‘Joint Criminal Enterprise and Command Responsibility’, (2007) 5 JICJ 159Google Scholar, at 174 (concluding that, relative to other aspects of JCE, ‘the conflict of JCE III with the principle of culpability is more fundamental’); Fletcher, G. P. and Ohlin, J. D., ‘Reclaiming Fundamental Principles of Criminal Law in the Darfur Case’, (2005) 3 JICJ 539Google Scholar, at 548 (arguing that ‘the doctrine [JCE] itself is substantively overbroad and transgresses basic principles of legality that limit punishment to personal culpability’); Ohlin, supra note 5, at 85 (discussing the violation of culpability occasioned by punishing different degrees of contribution equally); Danner and Martinez, supra note 5, at 134 (arguing that JCE poses significant challenges to the culpability principle).

29 In a sense, desert is synonymous with meritocracy. If an individual performs well in an exam, she deserves an excellent mark. If she kills her mother, she deserves punishment. For more on the positive and negative notions of desert, see J. Feinberg, Doing and Deserving: Essays in the Theory of Responsibility (1974), 55. See also P. H. Robinson, Distributive Principles of Criminal Law: Who Should Be Punished, How Much? (2008), 135 (discussing vengeful, deontological, and empirical concepts of desert).

30 Damaška, M., ‘What Is the Point of International Criminal Justice?’, (2008) 83 Chicago–Kent Law Review 329Google Scholar; I. Tallgren, supra note 13, at 561.

31 According to Hart, ‘[w]hat is needed is the realization that different principles (each of which may in a sense be called a “justification” [for punishment]) are relevant at different points in any morally acceptable account of punishment’. Furthermore, ‘it is perfectly consistent to assert both that the General Justifying Aim of the practice of punishment is its beneficial consequences and that the pursuit of this General Aim should be qualified or restricted out of deference to principles of Distribution which require that punishment should be only of an offender for an offense’; H. L. A. Hart, Punishment and Responsibility: Essays in the Philosophy of Law (1984), 3, 9 (emphasis in original).

32 Ibid., at 5–6.

33 G. P. Fletcher, Rethinking Criminal Law (1978), 415.

34 For a helpful overview of the different intensities of retribution and their intersection with utilitarian justifications for criminal law, see Alexander and Kessler Ferzan, supra note 14 (discussing weak, moderate, and strong conceptions of retributivism); for a summary of similar thinking in German criminal theory, see V. Krey, Deutsches Strafrecht: Allgemeiner Teil 2 (2002), 118 (discussing dominant theories of punishment in German criminal law, none of which advocates extending liability beyond an individual's desert).

35 Since restorative fault emphasizes a defendant's responsibility for rectifying harm that he has caused, John Braithwaite has argued that assessments of fault should be moved from their current point of assessment at the time the crime is perpetrated ‘to fault based on how restoratively the offender acts after the crime’, Braithwaite, J., ‘Intention versus Reactive Fault’, in Naffine, N., Owens, R. J., and Williams, J. M. (eds.), Intention in Law and Philosophy (2001), 345Google Scholar. Few courts have adopted restorative theories of punishment in cases involving international crimes, such that Braithwaite's vision of culpability is less germane for present purposes. This leaves open the question of whether, in preferencing some version of retributive punishment, international criminal lawyers may have ‘hitched themselves to a dead horse’; G. Simpson, Law, War and Crime: War Crimes Trials and the Reinvention of International Law (2007), 137.

36 J. Braithwaite and P. Pettit, Not Just Deserts: A Republican Theory of Criminal Justice (1993), 168 (‘We agree with the negative retributivists, for republican reasons, that indeed only the guilty ought to be punished’).

37 Damaška, supra note 5, at 470.

38 Prosecutor v. Tadić, supra note 2, at 186; Prosecutor v. Brima et al., Judgement, Case No. SCSL 04–16-A, Appeals Chamber, 22 February 2008, para. 15; Prosecutor v. Vasiljević, supra note 22, at 29. Strikingly, the better formulation was at Nuremberg: the Tribunal claimed that its reasoning was ‘in accordance with well-settled legal principles, one of the most important of which is that criminal guilt is personal, and that mass punishments should be avoided’, Trial of the Major War Criminals before the International Military Tribunal, Vol. 22 (1947), 499.

39 M. S. Moore, Placing Blame: A Theory of the Criminal Law (2010), 403 (acknowledging the dual meanings of culpability, but emphasizing that responsibility entails a voluntary and unjustified act that proximately causes harm, coupled with the obligation that ‘one must have done so culpably’); Fletcher, supra note 33, at 461 (stipulating that the components of desert are wrongdoing and culpability). Note that culpability bears several meanings here. On the one hand, it is frequently used in a normative sense – that is, a person is culpable only if she is justifiably to blame for her conduct – as compared with the use of the term ‘culpability’ in the US Model Penal Code to designate mental elements. For further discussion, see ibid., at 398.

40 It is difficult to reconcile the extent of strict liability in many Anglo-American national systems with the view frequently expressed in international criminal scholarship that national departures from culpability are highly exceptional. In a survey of 165 new offences created within England and Wales in 2005, Andrew Ashworth shows that strict liability was sufficient in 40%, plus an additional 26% were strict liability but watered down slightly by a proviso that the offence must be carried out ‘without reasonable excuse’; A. Ashworth, ‘Criminal Law, Human Rights and Preventive Justice’, in B. McSherry, A. Norrie, and S. Bronitt (eds.), Regulating Deviance: The Redirection of Criminalisation and the Futures of Criminal Law (2008). Strict liability is really just the tip of the iceberg. For a wide range of violations of culpability in the United States, see Robinson, P. H., ‘Imputed Criminal Liability’, (1983) 93 Yale Law Journal 609CrossRefGoogle Scholar, at 617–18 (discussing Pinkerton liability, the felony-murder rule, vicarious liability of officials of organizations, RICO, and others).

41 1948 Convention on the Prevention and Punishment of the Crime of Genocide, 78 UNTS 277, approved 9 December 1948, S. Treaty Doc. No. 1, 81st Cong., 2d Sess. (registered 12 January 1951) in Art. II.

42 R. Lemkin, Axis Rule in Occupied Europe (2008), 90.

43 Prosecutor v. Brðanin, Decision on Interlocutory Appeal, Case No. IT-99–36-A, Appeals Chamber, 19 March 2004, para. 6 (holding that, even when the crime charged is genocide, ‘the Prosecution will be required to establish that it was reasonably foreseeable to the accused that an act specified in Article 4(2) would be committed and that it would be committed with genocidal intent’). For similar conclusions relating to other special-intent crimes, see Prosecutor v. Milutinović et al., Judgement, Case No. IT-05–87-T, 26 February 2009, para. 471 (convicting Šainović of persecution for the murder of Kosovo Albanians ‘even though falling outside of the object of the JCE, [the murders carried act to persecute were] reasonably forseeable to Šainović’).

44 Nersessian, D. L., ‘Whoops, I Committed Genocide: The Anomaly of Constructive Liability for Serious International Crimes’, (2006) 30 Fletcher Forum of World Affairs 81Google Scholar, at 82.

45 Ibid., at 82. What this thoughtful criticism does not reveal is how aiding and abetting is also constructive, and that this point was instrumental in leading international courts to define JCE III similarly; Prosecutor v. Brðanin, supra note 43, at 5, 8 (‘As a mode of liability, the third category of joint criminal enterprise is no different from other forms of criminal liability which do not require proof of intent to commit a crime on the part of an accused before criminal liability can attach. Aiding and abetting, which requires knowledge on the part of the accused and substantial contribution with that knowledge, is but one example’).

46 Nersessian, supra note 44, at 96–8.

47 Ashworth, A., ‘The Elasticity of Mens Rea’, in Tapper, C. (ed.), Crime, Proof and Punishment: Essays in Memory of Sir Rupert Cross (1981), 45Google Scholar, at 53–6 (referring to ‘representing labelling’ as ‘the belief that the label applied to an offence ought fairly to represent the offender's wrongdoing’; Ashworth's prototypical illustration was the impropriety of merging the hitherto separate crimes of theft and obtaining by deception, which were thought to convey separate moral wrongs). For further discussion, see also Williams, G., ‘Convictions and Fair Labelling’, (1983) 42 Cambridge Law Journal 85CrossRefGoogle Scholar; Chalmers, J. and Leverick, F., ‘Fair Labelling in Criminal Law’, (2008) 71 Modern Law Review 217CrossRefGoogle Scholar. To my mind, this principle does not enjoy a separate existence from culpability. This is because the label of a criminal conviction is a key component of a defendant's punishment and therefore must be reconciled with desert.

48 Ashworth, supra note 47, at 53–6 (referring to ‘representing labelling’ as ‘the belief that the label applied to an offence ought fairly to represent the offender's wrongdoing’).

49 Chalmers and Leverick make this point well in describing the preferences of rape victims to have their assailants prosecuted for rape, even if this leads to lesser jail terms; see Chalmers and Leverick, supra note 47, at 217.

50 Robinson, supra note 5, at 941.

51 Ibid., at 941.

52 A. P. Simester offers arguably the most famous defence of joint criminal liability at the national level; see Simester, A. P., ‘The Mental Element in Complicity’, (2006) 122 Law Quarterly Review 578Google Scholar, at 599 (‘[b]y forming a joint enterprise, S signs up to its goal. In so doing, she accepts responsibility for the wrongs perpetrated in realising that goal, even though they be done by someone else. Her joining with P in a common purpose means that she is no longer fully in command of how the purpose is achieved. Given that P is an autonomous agent, S cannot control the precise manner in which P acts. Yet her commitment to the common purpose implies an acceptance of the choices and actions that are taken by P in the course of realizing that purpose. Her responsibility for incidental offences is not unlimited: S cannot be said to accept the risk of wrongs by P that she does not foresee, or which depart radically from their shared enterprise, and joint enterprise liability rightly does not extend to such cases. Within these limitations, however, the execution of the common purpose – including its foreseen attendant risks – is a package deal. Just as risks attend the pursuit of the common purpose, an assumption of those risks flows from S's subscription to that purpose’); see also G. P. Fletcher's helpful outline of the common justification for felony-murder, G. P. Fletcher, Basic Concepts of Criminal Law (1998), 193 (‘The state justifiably threatens robbers who cause death with an additional punishment in order to make them, as it were, “careful” robbers – they should do everything possible to minimize the risk of death. Imposing this additional burden on them is not considered unjust, for they, as robbers, have embarked on a forbidden course of endangering human life’).

53 Danner and Martinez, supra note 5, at 79.

54 Ibid., at 151.

55 Ibid., at 146.

56 Cassese, supra note 25, at 121.

57 Ibid., at 121.

58 Ibid., at 121.

59 Ibid., at 121; although Cassese treats complicity as an exception to these rules, this final aspect of his reasoning conflicts with his dual rationale for limiting JCE III at all – an accomplice is convicted of the same crime as a principal, too. In adopting the same conclusion, E. van Sliedregt explains the difference by pointing to a distinction between perpetration and participation. The former forbids escalation; the later tolerates this. van Sliedregt, E., Joint Criminal Enterprise as a Pathway to Convicting Individuals for Genocide, (2007) 5 JICJ 184–207, at 201Google Scholar. I am grateful to Thomas Weigend for confirming that this represents accepted reasoning within German criminal theory, too. I return to this argument later; see note 221, infra, and accompanying text. For now, it is sufficient to observe that the philosophical basis for this difference in treatment remains unexplained in the literature.

60 H. Grotius et al., Hugonis Grotii De Jure Belli ac Pacis Libri Tres: In Quibus Jus Naturae et Gentium, Item Juris Publici Prćcipua Explicantur (1925), 523.

61 In Re: Yamashita, 327 US 1 (1946), at 15; for a detailed history of the case, see R. L. Lael, The Yamashita Precedent: War Crimes and Command Responsibility (1982).

62 Prosecutor v. Halilović, Judgement, Case No. IT-01–48-T, Trial Chamber I, Section A, 16 November 2005, para. 53; A. J. Sepinwall, ‘Failures to Punish: Command Responsibility in Domestic and International Law’, (2008) 30 Mich. JIL 251, at 267 (completing a survey of previous practice by concluding that ‘there is overwhelming support for the mode of liability view’).

63 K. J. Heller, The Nuremberg Military Tribunals and the Origins of International Criminal Law (2011), 262–3.

64 E.g., Art. 7(2) of the ICTY Statute states: ‘The fact that any of the acts referred to in articles 2 to 5 of the present Statute was committed by a subordinate does not relieve his superior of criminal responsibility if he knew or had reason to know that the subordinate was about to commit such acts or had done so and the superior failed to take the necessary and reasonable measures to prevent such acts or to punish the perpetrators thereof’ (Updated Statute of the International Criminal Tribunal for the Former Yugoslavia, Art. 7(2), September 2008).

65 Damaška, supra note 5, at 468.

66 Ibid., at 468.

67 Ibid., at 469.

68 V. Tadros, Criminal Responsibility (2007), 49 (emphasis in original).

69 Fletcher, supra note 33, at 420.

70 See ibid., at 61–2 (describing a basic cleavage in the criminal law, between crimes of harmful consequences and crimes of harmful actions). Fletcher's taxonomy elsewhere refers to patterns of manifest criminality, harmful consequences, and subjective criminality; Fletcher, ibid., at 388–90. The last of these labels describes inchoate offences, and therefore is not directly relevant here; this tripartite taxonomy emulates German criminal theory. Krey, supra note 34, at 151–3 (discussing Erfolgsdelikte (result-oriented crimes); schlichte Tätigkeitsdelikte (non-result-oriented crimes); and Verletzungsdelikte/Gefährdungsdelikte (crimes constituted by violation of legal interests/mere endangerment of legal interests). The same distinction is true in both French and Spanish criminal law; see A. Eser, Individual Criminal Responsibility (2002), 105 (describing a distinction between delito de mera actividad and delito de resultado in the former, and infraction formelle and infraction materielle in the latter).

71 H. L. A. Hart and T. Honoré, Causation in the Law (1985), 79 (emphasis in original).

72 The criticism of the traditionalist division between conduct and harm-type offences is best made in M. S. Moore, Causation and Responsibility: An Essay in Law, Morals, and Metaphysics (2009), 101 (‘The thesis is that all complex descriptions of actions share with “killing” a built-in, second causal element: the bodily movement (that is caused by a volition) must itself cause some further, independent event to occur, like a death in the case of “killing”’); but see Gardner, J., Moore on Complicity and Causality, (2008) 156 University of Pennsylvania Law Review 432Google Scholar, at 432–53 (disagreeing that rape requires causation, because the offence demands ‘no result . . . other than the action in question having been performed’). For helpful discussion, including arguments that would disagree with both these positions, see Duff, R. A., ‘Acting, Trying, and Criminal Liability’, in Shute, S., Gardner, J., and Horder, J. (eds.), Action and Value in Criminal Law (1995), 75Google Scholar, at 83–5.

73 Moore, supra note 72, at 101.

74 For the classic discussion of this in English-speaking literature, see T. Nagel, ‘Moral Luck’, Proceedings of the Aristotelian Society, Supplementary Vol. 50 (1976); for more recent discussion, see A. Ashworth, ‘Taking the Consequences’, in Shute, Gardner, and Horder, supra note 72, at 107–24; Alexander and Kessler Ferzan, supra note 14, at 171–5.

75 For an excellent overview of these arguments, see Alexander and Kessler Ferzan, supra note 14, at 171–96 (arguing that only culpability, not resulting harm, affects desert). For a response to these claims, which asserts the orthodox position that harm matters, see Moore, supra note 72, at 30 (arguing that we feel very differently about a drunk driver's responsibility for swerving and only missing a child crossing the street by an inch than we do if the drunk driver actually hits and kills the child).

76 Prosecutor v. Ferdinand Nahimana, Jean–Bosco Barayagwiza, Hassan Ngeze, Judgement, Case No. ICTR-99–52-A, Appeals Chamber, 28 November 2007, para. 723.

77 Prosecutor v. Ferdinand Nahimana, Jean–Bosco Barayagwiza, Hassan Ngeze, Judgement, Case No. ICTR-99–52-T, Trial Chamber I, 3 December 2003, para. 1015; this aspect of the Trial Chamber's reasoning was affirmed on appeal. See Prosecutor v. Nahimana et al., supra note 76, at 678; for a concise articulation of the difference between instigation as mode of liability and incitement as inchoate offence, see Prosecutor v. Callixte Kalimanzira, Judgement, Case No. ICTR-05–88-T, Trial Chamber III, 22 June 2009, para. 512 (‘Instigation under Article 6 (1) is a mode of liability; an accused will incur criminal responsibility only if the instigation in fact substantially contributed to the commission of one of the crimes under Articles 2 to 4 of the Statute. By contrast, direct and public incitement is itself a crime, requiring no demonstration that it in fact contributed in any way to the commission of acts of genocide’).

78 The ICC Elements of Crimes define the two crimes as including the following elements: ‘(1) The perpetrator killed one or more persons, including by inflicting conditions of life calculated to bring about the destruction of part of a population’ and ‘The perpetrator deported or transferred one or more persons to another State or to another location’, Elements of Crimes, Official Records of the Assembly of States Parties to the Rome Statute of the International Criminal Court, First Session, New York, 3–10 September 2002 (hereafter, ‘ICC Elements of Crimes’), at 6 and 17, respectively.

79 Prosecutor v. Milutinović et al., supra note 43, at 92 (‘liability for aiding and abetting under the Statute cannot be inchoate: the accused cannot be held responsible under Article 7(1) for aiding and abetting if a crime or underlying offence is never actually carried out with his assistance, encouragement, or moral support’); Prosecutor v. Laurent Semanza, Judgement, Case No. ICTR-97–20-T, Trial Chamber III, 15 May 2003, para. 378 (‘Article 6(1) does not criminalize inchoate offences’).

80 Moore, supra note 72.

81 Ibid.; H.-H. Jescheck and T. Weigend, Lehrbuch des Strafrechts: Allgemeiner Teil (1996), 618 (also doubting that omissions are causal in so far as they lack ‘a real source of energy’).

82 Fletcher, supra note 52, at 64.

83 Tadros, supra note 68, at 171–2.

84 Prosecutor v. Krnojelac, Judgement, Case No. IT-97–25-A, Appeals Chamber, 17 September 2003, para. 171.

85 I view this language as largely cosmetic because it conceals the long history of holding the superior responsible ‘for the crimes of his subordinates’ and, more significantly, belies the ongoing practice of using superior responsibility to convict military and civilian commanders of ‘rape’, ‘pillage’, and ‘genocide’ carried out by underlings; e.g., ‘The Accused Ljubomir Borovèanin is found GUILTY pursuant to Article 7(3) of the Statute, of the following counts: Count 4: Murder, as a crime against humanity’, Prosecutor v. Popović et al., Judgement, Case No. IT-05–88-T, Trial Chamber II, 10 June 2010, para. 835. Thus, it seems clear that the superior is still convicted of the crime his subordinates perpetrated. For a more detailed confirmation of this reasoning, see Robinson, supra note 5, at 951–2.

86 Prosecutor v. Hadžihasanović and Kubura, Judgement, Case No. IT-01–47-T, Trial Chamber, March 2006, 620–8, para. 15 (finding Amir Kubura, for instance, ‘GUILTY of failing to take the necessary and reasonable measures to punish the murder of Mladen Havranek at the Slavonija Furniture Salon in Bugojno on 5 August 1993’; the judgement's entire disposition followed this approach).

87 Ibid., at 625, 627 (sentencing Enver Hadzihasanovic to five years’ imprisonment and Amir Kubura to two and a half years for failing to prevent or punish war crimes).

88 Prosecutor v. Hadžihasanović and Kubura, Judgement, Case No. IT-01–47-A, Appeals Chamber, 22 April 2008, para. 38, citing Prosecutor v. Blaškić, Judgement, Case No. IT-95–14-A, Appeals Chamber, 29 July 2004, para. 77; Prosecutor v. Kordić and Čerkez, Judgement, Case No. IT-95–14/2-A, Appeals Chamber, 17 December 2004, para. 832.

89 T. Weigend, Bemerkungen zur Vorgesetztenverantwortlichkeit im Völkerstrafrecht (2004), 116, 999, 1021; Jia, B. B., ‘The Doctrine of Command Responsibility Revisited’, (2004) 3 Chinese Journal of International Law 1CrossRefGoogle Scholar (‘[b]ut it makes no sense to see failure to punish in the same light, which should be treated as an offence independent of subordinate crimes that raise the issue of command responsibility in the first place’); Robinson, supra note 5, at 951 (‘[e]ven if we agree that failure to punish crimes is worthy of criminalization, it is simply inaccurate to label such a failure as “genocide”’); C. Meloni, ‘Command Responsibility: Mode of Liability for the Crimes of Subordinates or Separate Offence of the Superior?’, (2007) 5 JICJ 619, at 636 (‘with regard to the failure to punish, where no real causal link subsists between the subsequent failure to act of the superior and the crime previously committed, the conviction of the superior for the same crime committed by the subordinates is difficult to justify’); van Sliedregt, E., ‘Article 28 of the ICC Statute: Mode of Liability and/or Separate Offense’, (2009) 12 New Criminal Law Review 420, at 431–2Google Scholar (arguing that superior responsibility has different structures in different jurisdictions, but recommending the ‘splitting solution’ involving treating failures to punish as a separate crime).

90 Weigend, supra note 89, at 1021.

91 Ibid., at 1021.

92 Ibid.; see opinion to similar effect at note 83, supra.

93 Weigend, supra note 89, at 1021.

94 Triffterer, O., ‘Causality, a Separate Element of the Doctrine of Superior Responsibility as Expressed in Article 28 Rome Statute?’, (2002) 15 LJIL 203CrossRefGoogle Scholar (arguing that causation is embedded in the structure of superior responsibility); G. Mettraux, The Law of Command Responsibility (2009), 42–3 (accepting that superior responsibility is a mode of liability, and rejecting the view that the doctrine of superior responsibility contains no requirement of causality).

95 Triffterer, supra note 94, at 203. Use of the word ‘absolve’ in the text is my own. I use this in anticipation of a criticism that Triffterer's ‘second chance’ is causally unnecessary if the failure to control is already adequate. With this modification, I believe his account is coherent causally, even if I harbour grave doubts about whether the causal element could ever match the requisite subjective element of the subordinate's crime at the time of perpetration.

96 Weigend, supra note 89, at 1021; Meloni, supra note 89, at 630.

97 Sepinwall, supra note 62, at 289.

98 Ibid., at 292.

99 Ibid., at 295.

100 Her argument is, for instance, a wonderful explanation of why failures to punish must be criminalized in the face of fears of over-criminalization more broadly; see D. N. Husak, Overcriminalization (2008). Nonetheless, my own view is that a separate conduct-based crime remains the appropriate form of liability, since the superior who fails to subsequently punish the perpetrators under her command makes no difference to a completed atrocity.

101 Kadish, S. H., Complicity, Cause and Blame: A Study in the Interpretation of Doctrine, (1085) 73 CLR 323CrossRefGoogle Scholar, at 337 (‘the notion of derivativeness can be expressed as well in terms of the requirement of a result: just as causation doctrine requires that the prohibited result occur before there can be an issue of the actor having caused it, so in complicity doctrine there must be a violation of law by the principal before there can be an issue of the secondary party's liability for it’).

102 K. J. M. Smith, A Modern Treatise on the Law of Criminal Complicity (1991), 4.

103 Sullivan, supra note 14, at 154; Fletcher, supra note 33, at 649–50 (‘Aiding another person to commit a crime renders one an accomplice, and being an accomplice is simply one way of “being guilty of an offence”’).

104 Strafgesetzbuch, Bundesgesetzblatt I (2009), 3214, paras. 26, 27, 49; for a modern English translation, see M. Bohlander, The German Criminal Code: A Modern English Translation (2008), 43, at 50.

105 Is it not possible that some rogue aider is five-sixths as culpable as the perpetrator? As M. Moore has argued, ‘[o]ne could say that, on average, accomplices are less-substantial causers than are the principals they aid, and this is true enough. Yet this is only a rule of thumb, something that is true in the general run of causes’, Moore, supra note 15, at 423; and, in fact, the intuition that the indirect nature of the accomplice's acts renders her less culpable is ‘surprisingly difficult to justify’; C. Kutz, Complicity: Ethics and Law for a Collective Age (2000), 147 (discussing whether complicit actors are less culpable than direct actors). For further discussion, see Fletcher, supra note 33, at 654–7 (addressing the rationale for categorically mitigating the accessory's punishment).

106 The Accessories and Abettors Act 1861 (24 & 25 Vict. c.98), Section 8 (emphasis added).

107 Fletcher, supra note 33, at 651.

108 Ibid., at 582 (‘the actor is punished for a violation of the same prohibitory norm that covers standard cases of perpetration. There is no crime of . . . “being an accessory”’).

109 In England and Wales, Sections 44, 45, and 46 of the Serious Crimes Act 2007 create three new inchoate crimes of intentionally encouraging or assisting an offence, encouraging or assisting an offence believing it will be committed, and encouraging or assisting offences believing one or more will be committed. For commentary, see A. Ashworth, Principles of Criminal Law (2009), 458–61; in the United States, similar offences are labelled criminal facilitation; for discussion, see Weisberg, R., ‘Reappraising Complicity’, (2000) 4 Buffalo Criminal Law Review 217CrossRefGoogle Scholar, at 261–70.

110 C. Kutz, ‘Causeless Complicity’, (2007) 1 Criminal Law and Philosophy 239, at 301 (arguing that causality should play no role in complicity, in part in order to limit the role of moral luck in criminal law). For similar arguments, see R. Buxton, ‘Complicity and the Law Commission’, 1973 Criminal Law Review 223 (‘the way out of these and other difficulties would be to create a general offence of aiding or encouraging crime, committed by one who does acts which are known to be likely to be of assistance or encouragement to another in committing crime, whether or not that principal crime is in fact committed’); Yeager, D., ‘Helping, Doing, and the Grammar of Complicity’, (1996) 15 Criminal Justice Ethics 25CrossRefGoogle Scholar (‘harm principles should have little or nothing to do with the law of complicity’); Michael Moore summarizes these arguments succinctly by claiming that ‘(on this view) accomplice liability is just inchoate liability in the special cases when the evil sought to be prevented by the law has occurred (even though the accomplice did not cause it to occur)’; Moore, supra note 15, at 401. Note, however, that several of these authors would eliminate this special requirement that the harm occurred, making complicity resemble attempt even more closely than Moore suggests.

111 Fletcher, supra note 33, at 679.

112 Yeager, supra note 110, at 31.

113 Fletcher, supra note 33, at 637 (defining accessorial liability as ‘all those who are held derivately liable for another's committing the offense’).

114 The typical international trial alleging complicity concludes abruptly by declaring: ‘The Accused RADOSLAV BRĐANIN is found . . . GUILTY pursuant to Article 7(1) of the following counts: Count 3 – Persecutions’, Prosecutor v. Brðanin, Judgement, Case No. IT-99–36-T, Trial Chamber I, 1 September 2004, para. 1152; a comprehensive review of all convictions for aiding and abetting, instigation, planning, and ordering reveals that international courts and tribunals follow this format in 59 other situations – that is, making no mention of the mode of liability within the disposition of the judgement. In only three scenarios, international courts state something like: ‘The Chamber finds the Accused Haradin Bala GUILTY, pursuant to Article 7(1) of the Statute, of the following counts: Count 4 – Torture, a violation of the laws or customs of war, under Article 3 of the Statue, for having aided the torture of L12’, Prosecutor v. Limaj et al., Judgement, Case No. IT-03–66-T, Trial Chamber II, 30 November 2005, para. 41. Thus, in over 95% of cases, international courts' dispositions make no mention of complicity, even though it was the basis for conviction.

115 Ibid.

116 See United Nations War Crimes Commission, ‘Trial of Bruno Tesch and Two Others “The Zyklon B Case”’, 1 LRTWC 93.

117 Assume for the sake of this argument that Bruno Tesch's company Tesch & Stabenow was the only available supplier of the means of exterminating such a large number of civilians, such that their contribution was an indispensable cause of the crime. In fact, as I detail below, this is not factually accurate, but my minor factual modification makes the normative point indisputable.

118 Actually, so convicting those who intentionally provide chemicals for a genocide that subsequently takes place is even more objectionable than convicting an actual murderer of attempted murder, because the label ‘attempted murder’ at least communicates the gravity of the offence involved. Criminal facilitation communicates nothing of the sort and, to the extent that the label of the crime is a key element in the punishment inflicted on an accused, this significantly underrepresents desert.

119 Fletcher, supra note 33, at 637 (defining accessorial liability as ‘all those who are held derivately liable for another's committing the offense’).

120 Dubber, supra note 7, at 982 (showing how a putative accomplice would ‘[e]scape trial and punishment if the principal was never found, was never prosecuted, was acquitted, was convicted but had his conviction overturned or was pardoned’).

121 Sayre, F. Bowes, ‘Criminal Responsibility for the Acts of Another’, (1929) 43 Harvard Law Review 689CrossRefGoogle Scholar, at 695; Smith, supra note 102, at 20–3.

122 Bowes Sayre, supra note 121, at 695.

123 H.-L. Schreiber, ‘Problems of Justification and Excuse in the Setting of Accessorial Conduct’, 1986 Brigham Young University Law Review 611, at 620 (highlighting how the notion of limitierte Akzessorietät was not developed in German criminal law until 1943).

124 For even more complicated variants of my example, see Fletcher, supra note 33, at 667–9 (discussing inconsistent American case law on the issue of whether a confederate of a criminal who is justifiably shot by police while fleeing the crime scene can be an accomplice in the death of his confederate); Schreiber, supra note 123, at 629–30 (discussing a scenario in which the ‘accomplice’ deliberately initiates a situation in which you kill Mr W in self-defence).

125 Ibid., at 528.

126 Prosecutor v. Akayesu, Judgement, Case No. ICTR-96–4-T, Chamber I, 2 September 1998, para. 528.

127 Robert, supra note 8, at 351; see also Salvage, P., ‘Le lien de causalite en matière de complicité’, (1981) 25 Revue de science criminelle et de droit pénal comparé 41Google Scholar.

128 Kadish, supra note 101, at 337.

129 Prosecutor v. Vasiljević, supra note 22, at 102; Prosecutor v. Moinina Fofana and Allieu Kondewa, Judgement, Case No. SCSL 04–14-T, Trial Chamber I, 2 August 2007, para. 145; Prosecutor v. Ntakirutimana, Judgement, Case Nos. ICTR-96–10-A and ICTR-96–17-A, Appeals Chamber, 13 December 2004, para. 501.

130 Rome Statute of the International Criminal Court, 2187 UNTS 90 (entered into force on 1 July 2002), Art. 25(3)(c).

131 Westerfield, L., ‘The Mens Rea Requirement of Accomplice Liability in American Criminal Law: Knowledge or Intent’, (1980) 51 Mississippi Law Journal 155, at 178Google Scholar.

132 Ibid.

133 Ashworth, supra note 109, at 415; this also reflects the state of the law in German criminal law; see Bohlander, supra note 2, at 168 (discussing doppelter Anstifter- und Gehilfenvorsatz – twofold intent of the aider or abettor).

134 Smith, supra note 102, at 141–97 (reviewing English and American jurisprudence requiring that the accessory must intend his acts of assistance or encouragement and be aware of their ability to assist or encourage the principal offender, then exploring the complexities of these two variations); G. L. Williams, Criminal Law: The General Part (1953), 394–6 (reviewing the multiple mental elements required for the accomplice); Eser, supra note 70, at 923–4 (discussing the ‘double intent’ in complicity derived from German criminal law).

135 For comprehensive analyses, see Keitner, supra note 7, at 86–96; Cassel, D., ‘Corporate Aiding and Abetting of Human Rights Violations: Confusion in the Courts’, (2007) 6 Northwestern University Journal of International Human Rights 304Google Scholar, at 308–15.

136 I say perhaps unfaithfully because the Model Penal Code also has a strange provision requiring 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 Philadelphia: American Law Institute, Model Penal Code Commentaries (1985), at 296, para. 2.06 (emphasis added). The significance of this provision is opaque; see Mueller, G. G. E., ‘The Mens Rea of Accomplice Liability’, (1987) 61 Southern California Law Review 2169Google Scholar, at 2178 (pointing out that there is some ambiguity arising from this provision about how to address the accomplice's knowledge of circumstances, but arguing that the accomplice should be required to show the same mental element as that required for perpetration); in any event, even US federal standards of complicity have varied wildly, involving knowledge, purpose, recklessness, and a unitary theory; see Weiss, B., ‘What Were They Thinking: The Mental States of the Aider and Abettor and the Causer under Federal Law’, (2001) 70 Fordham Law Review 1341Google Scholar, at 1486 (reviewing the relevant case law and advocating for the unitary theory, which he describes as the ‘derivative approach’).

137 Prosecutor v. Furundžija, Judgement, Case No. IT-95–17/1-T, Trial Chamber, 10 December 1998, para. 246 (emphasis added); Prosecutor v. Blaškić, supra note 88, at 50; Prosecutor v. Kvočka et al., Judgement, Case No. IT-98–30/1-T, Trial Chamber, 2 November 2001, para. 255; Prosecutor v. Naletilić and Martinović, Judgement, Case No. IT-98–34-T, Trial Chamber, 31 March 2003, para. 63; Prosecutor v. Strugar, Judgement, Case No. IT-01–42-T, Trial Chamber II, 31 January 2005, para. 350; other cases refer to accessorial liability for ‘foreseeable consequences’ of one's actions; see Prosecutor v. Kvočka et al., Judgement, Case No. IT-98–30/1-T, Trial Chamber, 2 November 2001, at 262 (‘The aider or abettor of persecution will . . . be held responsible for discriminatory acts committed by others that were a reasonably foreseeable consequence of their assistance or encouragement’); Prosecutor v. Tadić, Judgement, Case No. IT-94–1-T, Trial Chamber, 7 May 1997, para. 692 (stating that the aider and abettor ‘will . . . be responsible for all that naturally results from the commission of the act in question’).

138 G. Williams, ‘Complicity, Purpose and the Draft Code – 2’, (1990) Criminal Law Review 98, at 99.

139 For a small selection of further examples, see Prosecutor v. Kunarac et al., Judgement, Case Nos. IT-96–23-T and IT-96–23/1-T, Trial Chamber, 22 February 2001, para. 759 (convicting the accused Kovać for handing over and/or selling two women to other soldiers whom he knew would ‘most likely continue to rape and abuse them’); Prosecutor v. Stakić, Judgement, Case No. IT-97–24-T, Trial Chamber II, 31 July 2003, para. 602 (convicting the accused, Stakić, for deliberately placing civilians in harm's way ‘with the knowledge that, in all likelihood, the victims would come to grave harm and even death’); Prosecutor v. Brima, Case No. SCSL 04–16-T, Trial Chamber II, 20 June 2007, para. 1786 (convicting the accused, Brima, for killings because he ‘was aware of the substantial likelihood that his presence would assist the commission of the crime by the perpetrators’).

140 Prosecutor v. Brðanin, supra note 114, at 272.

141 Ibid., at 473.

142 See note 11, supra.

143 LaFave, supra note 2, at 725–7 (discussing the ‘natural and probable consequence’ rule in various American jurisdictions, which is very similar to that adopted in international criminal justice); Robert, supra note 8, at 350 (setting out how an accomplice's acts are unlawful if the crime actually committed injures the same legal interest as that the accomplice considered); Bohlander, supra note 2, at 167–73 (indicating that in German law, dolus eventualis will suffice for the accomplice's intent); Ashworth, supra note 109, at 415–20 (discussing English jurisprudence that makes it adequate that the accomplice knows of the ‘type’ of crime the perpetrator will commit).

144 Ashworth, supra note 109, at 419 (cogently pointing out that ‘the accomplice knows that one or more of a group of offences is virtually certain to be committed, which means that in relation to the one(s) actually committed, there was knowledge only of a risk that it would be committed – and that amounts to recklessness’).

145 Prosecutor v. Ntakirutimana and Ntakirutimana, Judgement, Case Nos. ICTR-96–10-A and ICTR-96–17-A, Appeals Chamber, 13 December 2004, para. 497 (‘this standard (knowledge) does not extinguish the specific intent requirement of genocide. To convict an accused of aiding and abetting genocide based on the “knowledge” standard, the Prosecution must prove that those who physically carried out crimes acted with the specific intent to commit genocide’).

146 The most famous use of complicity to escalate responsibility occurred in the Krstić case, in which the Appeals Chamber of the ICTY overturned the Trial Chamber's conviction of General Krstić as a principal perpetrator in genocide, substituting a conviction for the same crime through complicity. This was necessary, according to the Appeals Chamber, because ‘[t]here was a demonstrable failure by the Trial Chamber to supply adequate proof that Radislav Krstić possessed the genocidal intent’; Prosecutor v. Krstić, Judgement, Case No. IT-98–33-A, Appeals Chamber, 19 April 2004, para. 134. There are, however, many more cases that adopt the same position. For a small subset of such cases across various international criminal tribunals, see Prosecutor v. Krnojelac, supra note 84, at 52 (finding for the crime against humanity of persecution that ‘the aider and abettor in persecution, an offense with a specific intent, must be aware . . . of the discriminatory intent of the perpetrators of that crime’, but ‘need not share the intent’); Prosecutor v. Fofana, supra note 129, at 145 (‘In the case of specific intent offences, the aider and abettor must have knowledge that the principal offender possessed the specific intent required’); Prosecutor v. Bagosora, Judgement, Case No. ICTR-98–41-T, Trial Chamber I, 18 December 2008, para. 2009 (‘[i]n cases of specific intent crimes such as persecution or genocide, the aider and abetter must know of the principal perpetrator's specific intent’).

147 See the discussion of French, German, Swiss, English, Canadian, and Australian law within the Krstić Appeal Judgement, which ultimately carried the day in what would rapidly become the accepted position across all international criminal tribunals; Prosecutor v. Krstić, supra note 146, at 141.

148 B. Fisse and C. Howard, Howard's Criminal Law (1990), 332.

149 Ibid.

150 Ibid. Stephen Shute sees an ability to predict natural phenomena like the sun rising in the east as undermining the argument that ‘[i]n the strictest sense of the word one cannot “know” that something will be the case in the future’, S. Shute, ‘Knowledge and Belief in the Criminal Law’, in S. Shute and A. P. Simester (eds.), Criminal Law Theory: Doctrines of the General Part (2002), 171, at 186–7; I am doubtful whether these examples do enough to account for Fisse's point about knowledge. Natural phenomena like the sun rising tomorrow are merely examples of future events we can predict with the highest degree of certainty, but these illustrations do not mean that awareness of a probability is automatically equivalent to knowledge. To my mind, G. R. Sullivan offers a more accurate explanation by accepting that we can know what the laws of physics will produce in the future, but that many cases involving decisions about what defendants knew of other people's future acts ‘afford graphic demonstrations of how statutory language is sometimes completely overridden’, G. R. Sullivan, ‘Knowledge, Belief, and Culpability’, in Shute and Simester, supra, at 215.

151 Fisse and Howard, supra note 148, at 332.

152 Smith, supra note 102, at 13.

153 Williams, supra note 138, at 101.

154 Kadish, supra note 101, at 353 (‘A pall would be cast on ordinary activity if we had to fear criminal liability for what others might do simply because our actions made their acts more probable’). As I set out in the next paragraph, Kadish does not himself agree with this argument.

155 Williams, supra note 138, at 101. Apart from the responses to this line of argument I set out below, I also find Glanville Williams's argument that recklessness ‘requir[es] me to enquire’ misleading. One either assists someone, believing that there is a substantial probability that they will use your assistance to perpetrate a crime, or one declines to offer that assistance. In either scenario, the duty to enquire does arise (since you are judged based on your actual awareness), such that the metaphor of unpaid auxiliary police is unfounded. The position also prematurely assimilates complicity with omission liability perfectly, when there are important differences between these two types of derivative liability; see Fletcher, supra note 33, at 676–7 (concluding a comparison between omission and complicity by highlighting the differences between the two, many of which undermine Williams's arguments).

156 Kadish, supra note 14, at 369, 387 (‘It is not evident to me that subjecting actors in these circumstances to liability for a crime of recklessness need greatly imperil the security of otherwise lawful activities, certainly not any more than holding actors liable for recklessly “causing” harms, which the law regularly does. People aren't all that unpredictable’); Alexander, L., ‘Insufficient Concern: A Unified Conception of Criminal Culpability’, (2000) 88 CLR 931CrossRefGoogle Scholar, at 944–7 (defending reckless complicity against arguments of overreach). For the traditional response to these arguments, see Simester, supra note 52, at 588–91.

157 See note 142, supra.

158 The offence need not mention knowledge at all. For a war crime like declaring no quarter be given, the basic requirement is only that the perpetrator declared or ordered that there shall be no survivors, and that the declaration or order ‘was given in order to threaten an adversary or to conduct hostilities on the basis that there shall be no survivors’. These elements make no mention of knowledge, leaving a great deal of ambiguity about what the accomplice needs to know in order to be convicted of the offence. This ambiguity also leads to terrible complexity. For example, see Law Commission, LAW COM No. 305, Participating in Crime (2007) (UK) (detailing the tremendous complexity of the knowledge-based system within England and Wales).

159 Weisberg, supra note 109, at 233 (exploring different interpretations of these three elements); Mueller, supra note 136, at 2174 (arguing that, because of these multiple points of inquiry, confusion has existed concerning the mens rea element of accomplice liability for years).

160 LaFave, supra note 2, at 324.

161 See the summary of inconsistent approaches and the advent of the double intent in Prosecutor v. Orić, Judgement, Case No. IT-03–68-T, Trial Chamber II, 30 June 2006, paras. 286–288.

162 See note 128, supra.

163 ‘ICC Elements of Crimes’, supra note 78, at 26 (pillage requires that the perpetrator intended to deprive the owner of the property and to appropriate it for private or personal use); I am compelled to add that I strongly disagree that this notion of private or personal use is workable or reflects customary international law; see J. G. Stewart, Corporate War Crimes: Prosecuting Pillage of Natural Resources (2010), 19–23; ibid., at 14 (the war crime of torture requires the perpetrator to be aware of the factual circumstances that established the protected status of the victim under the Geneva Conventions of 1949).

164 Evidently, the point was the subject of heated discussion; see American Law Institute, supra note 136, at 318–19, para. 2.06 (noting that, while the Chief Reporter for the Model Penal Code favoured a standard broader than purpose, the Institute rejected the position after tense debate); see also Simester, supra note 52, at 583 (arguing that the protection of potential victims and the preservation of liberties for potential defendants ‘demand more stringent mens rea standards for secondary liability than is needed to establish culpability’).

165 Dennis, I. H., ‘The Mental Element for Accessories’, in Smith, P. (ed.), Criminal Law: Essays in Honour of J. C. Smith (1987), 60Google Scholar (criticizing the purpose standard as insensitive to retributive notions of desert); Sullivan, supra note 14, at 154–5 (arguing that, in order for the accomplice to be convicted of the same offence as the perpetrator, retributive theory would require an equivalence of culpability); for a competing perspective, see Simester, supra note 52, at 600 (arguing that ‘culpability is not enough . . . the better approach is to distinguish culpability from responsibility, and to focus on the latter’).

166 If, to paraphrase I. Dennis, more reckless facilitators are deterred, perhaps fewer atrocities will transpire; Dennis, supra note 165, at 60.

167 Duff, R. A., ‘“Can I Help You?”: Accessorial Liability and the Intention to Assist’, (1990) 10 Legal Studies 165CrossRefGoogle Scholar; Williams, supra note 134, at 366–80; Bronitt, S., ‘Defending Giorgianni – Part Two: New Solutions for Old Problems in Complicity’, (1993) 17 Criminal Law Journal 305Google Scholar; for myself, I doubt whether business deserves the privileged status it often receives in the theoretical discussions of this topic, given that it merely represents one facet of social interaction where influence is rampant. What, for instance, about families, literature, music, and teachers?

168 P. Robinson and J. M. Darley, Justice, Liability and Blame: Community Views and the Criminal Law (1996), 103.

169 Ibid., at 103 (concluding that ‘[f]rom the point of view of our respondents, the culpability requirement as to result should not be elevated to purposeful . . . instead, the offense should be graded according to the degree of culpability that the accomplice shows’).

170 A. Duff argues that there is a conceptual problem with punishing at the supranational level, to the extent that ‘[c]alling someone to answer, holding someone responsible, is a communicative endeavor which presupposes normative community; normative community requires at least a modicum of mutuality’; A. Duff, ‘Can We Punish the Perpetrators of Atrocities?’, in T. Brudholm and T. Cushman (eds.), The Religious in Responses to Mass Atrocity (2009), 93. Although these criticisms are important, they do tend to overlook the growing practice of national courts’ prosecuting their own nationals for international crimes within domestic courts, and the supranational principle of complementarity, which seeks to institutionalize that shift towards trials in national communities. Moreover, the enforcement of international criminal norms in regional international courts, which represent a more homogeneous community, may improve the case for international criminal adjudication; see, e.g., Burke-White, W. W., ‘Regionalization of International Criminal Law Enforcement: A Preliminary Exploration’, (2003) 38 Texas ILJ 729Google Scholar.

171 Prosecutor v. Blaškić, supra note 88, at para. 48 (emphasis added); for a different rendering of the same idea, see Prosecutor v. Lukić and Lukić, Judgement, Case No. IT-98–32/1-T, Trial Chamber III, 20 July 2009, para. 901 (‘There is no requirement of a causal relationship between the conduct of the aider or abettor and the commission of the crime’). The same standard has spread to other international criminal tribunals; Prosecutor v. Fofana, supra note 129, at 143 (same); Prosecutor v. Bagilishema, Judgement, Case No. ICTR-95–1A-T, Trial Chamber I, 7 June 2001, para. (‘the assistance given by the accomplice need not constitute an indispensable element, i.e. a conditio sine qua non, of the acts of the perpetrator’).

172 Prosecutor v. Blaškić, supra note 88, at para. 48 (emphasis added); Prosecutor v. Blagojević and Jokić, Judgement, Case No. IT-02–60-A, Appeals Chamber, 9 May 2007, para. 187 (‘the Appeals Chamber reiterated that one of the requirements for the actus reus of aiding and abetting is that the support of the aider and abettor have a substantial effect upon the perpetration of the crime’); Ndindabahizi v. The Prosecutor, Judgement, Case No. ICTR-01–71-A, Appeals Chamber, 16 January 2007, para. 117 (‘a conviction for aiding and abetting presupposes that the support of the aider and abetter has a substantial effect upon the perpetrated crime’); Prosecutor v. Brima, supra note 38, at 775 (‘The actus reus of “aiding and abetting” requires that the accused gave practical assistance, encouragement, or moral support which had a substantial effect on the perpetration of a crime’).

173 The idea is reminiscent of a cartoon in the New Yorker magazine that depicts a meeting between three businesspeople, where one comments to another ‘we want to include you in this decision without letting you affect it’; see The Cartoon Bank (2011), available online at www.cartoonbank.com/2011/we-want-to-include-you-in-this-decision-without-letting-you-affect-it/invt/137184.

174 Gardner, supra note 72, at 443.

175 The International Law Commission's Draft Code of Crimes against Peace and Security of Mankind adopted a definition of aiding and abetting that required the accomplice to assist ‘directly and substantially’, but the only justification for this was that this was ‘intended to limit the application of the Code to those individuals who had had a significant role in the commission of a crime’; International Law Commission, Summary of the 2437th Meeting, Consideration of the Draft Articles on Second Reading, 6 June 1996, para. 26. The two international judgments that initially endorsed the substantial-effect standard relied on the ILC recommendation, together with a selection of Second World War case law that made no direct mention of substantial effect; see Prosecutor v. Tadić, supra note 137, at 688–92; Prosecutor v. Furundžija, supra note 137, at 219–31; accordingly, I conclude that the true criminological motivations for the substantial-effect doctrine are mainly unarticulated.

176 Hart and Honoré, supra note 71, at 41 (‘A deliberate human act is therefore most often a barrier and a goal in tracing back causes in such inquiries: it is something through which we do not trace the cause of a later event and something to which we do trace the cause through intervening causes of other kinds’); Hart and Honoré, supra note 71, at 129 (‘the free, deliberate and informed act or omission of a human being, intended to produce the consequence which is in fact produced, negatives causal connection’).

177 Kadish, supra note 101, at 333 (arguing that, otherwise, we do violence to notions of agency and the conception of a human action as freely chosen upon which we depend to convict the perpetrator).

178 Moore, supra note 72, at Part IV, The Legal Presupposition of There Being Intervening Causes (criticizing Hart and Honore's views that voluntary actions are intervening causes); Feinberg, supra note 29 (arguing, contrary to Hart and Honoré, that ‘there is no conceptual barrier, at least none imposed by common sense, to our speaking of the causes of voluntary actions’); in particular, Feinberg's conceptual distinction between ‘causing a person to act’ and ‘making him act’ offers a strong critique of Hart and Honoré's thesis; Feinberg, supra note 29, at 161, 165 (arguing that, although a mother clearly played some (albeit extremely remote) causal role in her 30-year-old son's crime by merely having given birth to the perpetrator, it would be ‘misleading in the extreme’ to suggest that his mother, 30 years earlier, ‘made’ him perpetrate the crime; for further criticism of Hart and Honoré's thesis, see Smith, supra note 102, at 68–70 (‘it is possible to construct counter-examples where actions, while voluntary within the meaning accorded by Hart and Honoré, are in “common speech” reasonably describable as “caused” by another’). For a similar position in German criminal theory, see G. Freund, in W. Joecks, K. Miebach, and B. von Heintschel-Heinegg (eds.), Münchener Kommentar zum Strafgesetzbuch: Band 1 (2003), paras. 1–51; C. Roxin, Strafrecht. Allgemeiner Teil Band I: Grundlagen, Der Aufbau der Verbrechenslehre (2006), 363.

179 Moore, supra note 15, at 422–3; Kutz, supra note 110, at 294 (in fairness to Hart and Honoré, they viewed instigation as an exception to their general rule that voluntary action breaks causal chains but, as J. Feinberg retorts, ‘they put forward no more general principle to explain why the exceptions are exceptions’); Feinberg, supra note 29, at 153.

180 International criminal courts and tribunals confirm as much. For instance, in the media case in which representatives of the Radio télévision libre des mille collines (RTLM) were convicting of inciting genocide, the Rwanda Tribunal held that ‘[t]he nature of media is such that causation of killing and other acts of genocide will necessarily be effected by an immediately proximate cause in addition to the communication itself. In the Chamber's view, this does not diminish the causation to be attributed to the media, or the criminal accountability of those responsible for the communication’, Prosecutor v. Nahimana et al., supra note 77, at 952.

181 There is some support for this thesis. Both of the first cases to address complicity in modern international criminal justice refer to the problem of over-determination within the context of discussions of the substantial-effect doctrine; see Prosecutor v. Tadić, supra note 137, at 688 (acknowledging that ‘in virtually every situation, the criminal act most probably would have occurred in the same way had not someone acted in the role that the accused in fact assumed’); Prosecutor v. Furundžija, supra note 137, at 224 (discussing a Second World War case the defendant claimed that his conduct in no way contributed to the crimes because others would have taken his place). Nonetheless, this explanation is not entirely convincing, since the same judgement also acknowledged that ‘the culpability of an aider and abettor is not negated by the fact that his assistance could easily have been obtained from another’; ibid.

182 United Nations War Crimes Commission, supra note 116, at 102; indeed, even if the SS were bent on using Zyklon B for the purposes, there were many other sources. Representatives of the firm I.G. Farben were also prosecuted for supplying large quantities of Zyklon B that ‘was actually used in the mass extermination of inmates of concentration camps, including Auschwitz’; United Nations War Crimes Commission, ‘Trial of Carl Krauch and Twenty-Two Others “The I.G. Farben Trial”’, (1947) 10 LRTWC 1, at 23–4.

183 For an elegant philosophical discussion of the problem, which draws on examples of complicity, see J. Glover, ‘It Makes No Difference Whether or Not I Do It’, (1975) 49 Proceedings of the Aristotelian Society (discussing over-determination with reference to a scientist producing chemical and biological weapons); Kutz, supra note 105, at 115–24 (exploring the responsibility of pilots in the Dresden firebombing on the basis of over-determined causes); J. Gardner, Offences and Defences: Selected Essays in the Philosophy of Criminal Law (2008), 57 (discussing over-determined causality in the context of complicity, although he does not use the term ‘over-determination’).

184 Hart and Honoré, supra note 71, at 117–19 (discussing what they describe as additional causes, and the need for assessing sine qua non based on events that occurred ‘in this particular way’); Smith, supra note 102, at 84 (‘the sine qua non condition is concerned with an event's exact occurrence, including time, place, extent and type of harm, and so on’); Hörnle, T., ‘Commentary to “Complicity and Causality”’, (2006) 1 Criminal Law and Philosophy 143CrossRefGoogle Scholar, at 144 (using the example of a firing squad to show how the ‘subtraction method’ of calculating causation leads to injustice, which might be overcome by focusing on events ‘as they happened’); for criticism that this approach misuses the term ‘causation’, see Yeager, supra note 110, at 29 (arguing that this approach ‘simultaneously uses a word [‘cause’ or ‘causation’] in a special or technical sense that need not confirm to our ordinary use of the word, while still trading on what we normally mean by it’). Likewise, for further criticisms, see Moore, supra note 15, at 406–7.

185 In addition to the English-language authorities cited immediately above, see also F. Dencker, Kausalität und Gesamttat (1996), 86; H. Koriath, Kausalität und objektive Zurechnung (2007), 145–7; Roxin, supra note 178, at 359.

186 RGSt 58, 113, at 114–15 (Entscheidungssammlung des Reichsgerichts in Strafsachen, Vol. 58, 113, at 114–115); see also C. Roxin, Strafrecht: Allgemeiner Teil Band II: Besondere Erscheinungsformen der Straftat (2003), at 194, para. 26, marginal number 186.

187 See W. Joecks, in Joecks, Miebach, and von Heintschel-Heinegg, supra note 178, paras. 1–51 StGB and 27, marginal numbers 23–37; S. A. Osnabrügge, Die Beihilfe und ihr Erfolg: Zur objektiven Beziehung zwischen Hilfeleistung und Haupttat (2002), 159–60, 261, para. 27 StGB; Roxin, supra note 178.

188 Joecks, Miebach, and von Heintschel-Heinegg, supra note 178, at marginal number 27; Roxin, supra note 178.

189 To recall, the accepted position in international criminal justice is that ‘proof of a cause–effect relationship between the conduct of the aider and abettor and the commission of the crime, or proof that such conduct served as a condition precedent to the commission of the crime, is not required’, see note 166, supra, and accompanying text.

190 Fletcher, supra note 33, at 590 (‘Because the causal link is limitless, some new concept must be devised to eliminate far-flung effects from the range of liability. Common lawyers speak about proximate cause’); in German criminal theory, normative attribution (objektive Zurechnung) is considered an additional element of any actus reus, in order to restrict the broad effect of causality; C. Roxin, Strafrecht: Allgemeiner Teil: Grundlagen, der Aufbau der Verbrechenslehre (2006), 372; Koriath, supra note 185, 15 (discussing the implications of normative attribution); M. Maiwald, Kausalität und Strafrecht: Studien zum Verhältnis von Naturwissenschaft und Jurisprudenz (1980), 4–5, 9; for a helpful English-language summary, see Krey, supra note 34, at 59–101.

191 I borrow the example from Joshua Dressler, although he uses it in a different context; Dressler, J., ‘Reassessing the Theoretical Underpinnings of Accomplice Liability: New Solutions to an Old Problem’, (1985) 37 Hastings Law Journal 91Google Scholar, at 133.

192 Gardner, supra note 9, at 231. Gardner views the paradox as more apparent than real because causal relations come in stronger and weaker versions; ibid.

193 Moore, supra note 39, at 423.

194 Art. 12, Austrian Criminal Code, translated from the French version in Pradel, supra note 16, at 133. It is interesting to note that, according to para. 15, section 2, of the Austrian Penal Code, the attempt to facilitate an offence is not punishable. This reveals that the Austrian system still requires a distinction between instigators and aiders and cannot therefore be considered a pure unitary system. I am grateful to Thomas Weigend for the point.

195 In my view, any time a court refers to a mode of attribution as sui generis, the Latin acts as a mask for the departure from basic principles. The phrase is thus a telltale sign that the mode of liability cannot be philosophically justified; Prosecutor v. Halilović, supra note 62, at 78 (‘The Trial Chamber further notes that the nature of command responsibility itself, as a sui generis form of liability, which is distinct from the modes of individual responsibility set out in Article 7(1), does not require a causal link’); Prosecutor v. Orić, supra note 161, at 293 (‘the superior's responsibility under 7(3) of the Statute can indeed be called a responsibility sui generis’).

196 The standard was initially adopted in Prosecutor v. Thomas Lubanga Dyilo, Decision on the Confirmation of Charges, Case No. ICC-01/04–01/06, Pre-Trial Chamber I, 29 January 2007, para. 341; see also Prosecutor v. Germain Katanga and Mathieu Ngudjolo Chui, Decision on the Confirmation of Charges, Case No. ICC-01/04–01/07, Pre-Trial Chamber I, 30 September 2008, para. 486 (finding that the criminal responsibility of a person ‘must be determined under the control over the crime approach to distinguishing between principals and accessories’). The concept was also employed at the ICTY by one German judge, but the use of the doctrine was rejected on appeal; see Prosecutor v. Stakić, supra note 139, at 440; Prosecutor v. Stakić, Judgement, Case No. IT-97–24-A, Appeals Chamber, 22 March 2006, para. 62 (finding that ‘[t]his mode of liability, as defined and applied by the Trial Chamber, does not have support in customary international law or in the settled jurisprudence of this Tribunal, which is binding on the Trial Chambers’).

197 Fletcher, supra note 33, at 655 (pointing out that the subjective test was unworkable in practice because a trier of fact could not easily determine the attitude of the suspect at the time of the deed); Schreiber, supra note 123, at 626 (detailing the criticism that the person swinging the machete could consider herself an accomplice, and thereby benefit from lower penalties afforded accessories). Ultimately, some consider that this could also lead to a situation in which differences of opinion among assailants mean that there are no perpetrators of a crime at all. This would arise where all participants in a criminal offence believed that they left the decision of whether to commit the crime to others; for a full discussion, see Olásolo, supra note 17, at 30–3.

198 See the discussion of Organisationsherrschaft (control over an organization), in Prosecutor v. Germain Katanga and Mathieu Ngudjolo Chui, supra note 196, at 498–518.

199 Weigend, T., ‘Perpetration through an Organization: The Unexpected Career of a German Legal Concept’, (2011) 9 JICJ 91Google Scholar, at 105 (‘Since criminal liability for ordering or instigation is a sufficient basis for imposing severe sentences on responsible figures in the background of the actual crimes, adopting the notion of “perpetration through an organization” may create more problems than it solves’).

200 Prosecutor v. Bemba, Decision Pursuant to Article 61(7)(a) and (b) of the Rome Statute on the Charges of the Prosecutor against Jean-Pierre Bemba Gombo, Case No. ICC-01/05–01/08, Pre-Trial Chamber II, 15 June 2009, para. 326 (emphasis added). I add emphasis to the words ‘the distinguishing criterion’ because of the highly contestable assumptions that a distinguishing criterion is conceptually necessary and that there is only one.

201 I refer to ‘important criminal energy’ because it is the classic justification for a distinction between perpetrators and accomplices founding German criminal theory. The argument is that extensive participation shows important criminal energy, and that qualitatively significant contributions are more culpable. So, in contrast to the principal and instigator, whose contributions drive the wrongdoing, the aider's contribution is of minor relative significance; C. Roxin, supra note 186, at 231; see also W. Joecks, K. Miebach, and G. M. Sander, Münchener Kommentar zum Strafgesetzbuch: Gesamtwerk (2003), at 3.

202 Gardner, supra note 183, at 63.

203 Gardner later acknowledges in response to critics that ‘the distinction between principals and accomplices might perhaps be excised from the law (e.g. for rule of law reasons), but, as [he] tries to show, it cannot be excised from life’; Gardner, supra note 19, at 135. I am tempted to read this as a concession that a unitary theory of perpetration is legally justifiable provided moral distinctions are respected within the sentencing phase, but I no doubt read more into the comment than he might accept.

204 I doubt this because the point seems entirely contingent on the construction of the particular offence in question. If, for example, an offence is defined as ‘causing rape’, then the criminal offence itself collapses the distinction between perpetrators and accomplices. Thus, if there is a distinction between perpetrators and accomplices, it is a by-product of the drafting of criminal codes, not an innate property of principal perpetration or accessorial liability themselves. Curiously, the point is not entirely academic for international criminal justice – the war crime of ‘wilfully causing great suffering or serious injury and cruel treatment’ presumably furnishes a practical illustration. Conceptually, there is no difference between perpetrators and accomplices of this war crime, since treaty-makers have expunged any difference by employing causation in the crime's definition.

205 For background to the derivative nature of complicity, see section 2.1.2, infra.

206 Fletcher, supra note 33, at 636.

207 See section 2.1.2, infra.

208 Kadish, supra note 101, at 373; Smith, supra note 102, at 107–10; Moore, supra note 15, at 418–20; Gardner, supra note 19, at 136.

209 Prosecutor v. Nyiramasuhuko et al., Judgement, Case No. ICTR-98-42-T, Trial Chamber II, 24 June 2011, paras. 6087–6088.

210 For proof of this in international criminal justice, see note 109, supra (showing that the dispositions of international courts and tribunals make no mention of complicity in over 95% of complicity cases that lead to conviction. Instead, they merely declare the name of the crime with which the accomplice is convicted).

211 Fletcher, supra note 33, at 654–6 (highlighting reasons for the departure from the objective theory of perpetration); Schreiber, supra note 123, at 614 (placing the objective theory of perpetration in historical context, and demonstrating the passage to a subjective theory, then the turn to ‘control over the act’); Dubber, supra note 7, at 983 (highlighting how the position in the US Model Penal Code resembles the objective theory, perhaps explaining why the non-proxyable problem remains so vital in English-language theory).

212 As mentioned earlier, the war crime of ‘wilfully causing great suffering’ already takes this form; see note 202, supra.

213 Moore, supra note 15, at 417.

214 Kutz, supra note 110, at 303.

215 C. Roxin, in B. Jähnke, H. W. Laufhütte, and W. Odersky (eds.), Strafgesetzbuch: Leipziger Kommentar: Erster Band (2003), para. 25, marginal number 6; see also Jescheck and Weigend, supra note 81, at 646.

216 Roxin et al., supra note 215, marginal number 8.

217 Rotsch, supra note 14 (showing how considerations of culpability and sentencing prompt practitioners and theorists to choose somewhat arbitrarily between modes of attribution to apply); evidently, these practices have a long history. Schreiber explains that, during the Weimar Republic reform, efforts were undertaken to relax strict insistence on the derivative nature of complicity so as ‘to constrain the scope of indirect perpetration and to relegate many of the cases that were thus being dealt with as a species of perpetration back to the category of complicity’; Schreiber, supra note 123, at 620. This is also true in France, where the fact that complicity does not attach to the lowest form of crimes (called contraventions) leads the French Cour de cassation to declare accomplices co-perpetrators to avoid their acquittal. For details, see B. Bouloc, G. Stefani, and G. Levasseur, Droit pénal général (2005), 287–8. For further modern examples from the Netherlands and elsewhere, see Keiler, supra note 18, at 186–90.

218 HR, 23 October 1990, NJ 1991, 328, cited in Keiler, supra note 18, at 187.

219 Keiler, supra note 18, at 190.

220 For discussion of the fair-labelling principle, see note 46, supra, and accompanying text.

221 Van Sliedregt, supra note 59 (arguing that whether members of a JCE must comply with the full mens rea of genocide turns on whether they are perpetrators or participants); Cassese, supra note 25, at 121 (arguing that it is a logical impossibility for someone who does not have the necessary mens rea for genocide to ‘commit’ the crime, but accepting that he or she may aid and abet the crime nonetheless). I am grateful to Thomas Weigend for confirming that the fact that JCE and superior responsibility purport to act as forms of ‘committing’ a crime whereas complicity is a mere means of participation is the normative basis upon which the two concepts might diverge within a differentiated model.

222 For one of the most thoughtful discussions, see H. van der Wilt, ‘Joint Criminal Enterprise and Functional Perpetration’, in A. Nollkaemper and H. van der Wilt (eds.), System Criminality in International Law (2009), available online at http://dx.doi.org/10.1017/CBO9780511596650.

223 See note 104, supra.

224 Ibid.

225 See note 208, supra.

226 Rotsch, supra note 14, at 468.

227 See van der Wilt, supra note 222; H. van der Wilt also argues that we should not underestimate the symbolic expressive value of JCE. While I agree with everything in this article, I disagree with the framework best capable of achieving these desires.

228 Bothe, M., ‘War Crimes’, in Cassese, A., Gaeta, P., and Jones, J. R. W. D. (eds.), The Rome Statute of the International Criminal Court (2002), 392Google Scholar.

229 Draper, G. I. A. D., ‘The Modern Pattern of War Criminality’, in Dinstein, Y. and Tabory, M. (eds.), War Crimes in International Law (1996), 160Google Scholar.

230 See, e.g., ‘1949 Geneva Convention Relative to the Protection of Civilian Persons in Time of War’, 75 UNTS 287 (1958), A.T.S. No. 21, Art. 146(3) (‘The High Contracting Parties undertake to enact any legislation necessary to provide effective penal sanctions for persons committing, or ordering to be committed, any of the grave breaches of the present Convention defined in the following Article’).

231 This is evident, for instance, from the statement of one negotiator at the time, who observed that ‘[t]he Conference is not making international penal law’, Final Record of the Diplomatic Conference of Geneva of 1949, Vol. II (1949), Section B, at 116.

232 Art. 10 of the ICC Statute stipulates that ‘[n]othing in this Part shall be interpreted as limiting or prejudicing in any way existing or developing rules of international law for purposes other than this Statute’.

233 M. Koskenniemi, ‘The Pull of the Mainstream’, (1990) 88 Michigan Law Review 1946, at 1952. This, as Koskenniemi memorably argues, ‘because the interpretation of “state behavior” or “state will” is not an automatic operation but involves the choice and use of conceptual matrices that are controversial and that usually allow one to argue either way’; Koskenniemi's principal point is that it is really our moral certainty that something should be prohibited that is driving the analysis of custom, not some objectively ascertainable standard that might be obtained in some dispassionate positivist fashion.

234 B. Simma and P. Alston, ‘Sources of Human Rights Law: Custom, Jus Cogens, and General Principles’, (1988) 12 Australian Year Book of International Law 82, at 88; the attempts to justify concepts in human rights (and international criminal law) as pre-existing aspects of customary international law even though they are not easily reconciled with normal standards for identifying custom is, I suspect, an example of what David Kennedy calls ‘a combination of overly formal reliance on textual articulations that are anything but clear or binding and sloppy humanitarian argument’; see Kennedy, D., ‘International Human Rights Movement: Part of the Problem?’, (2002) 15 HHRJ 101, at 120Google Scholar.

235 Prosecutor v Ieng et al., Case No: 002-19-2007-ECCC/OCIJ (PTC38), para. 83 (‘For the foregoing reasons, the Pre-Trial Chamber does not find that the authorities relied upon in Tadic . . . constitute a sufficiently firm basis to conclude that JCEIII formed part of customary international law at the time relevant to Case 002’).

236 For an excellent overview of this history together with modern manifestations of the problem, see Van Schaack, B., ‘Crimen Sine Lege: Judicial Lawmaking at the Intersection of Law and Morals’, (2008) 97 GLJ 119Google Scholar.

237 Damaška, supra note 5, at 469.

238 Van Schaack, supra note 236, at 125.

239 Sloane, R. D., ‘The Expressive Capacity of International Punishment: The Limits of the National Law Analogy and the Potential of International Criminal Law’, (2007) 43 Stanford JIL 39Google Scholar, at 42–4 (arguing that expressive theories of punishment are likely to capture the nature of international sentencing better than other retributive or utilitarian conceptions); see also M. Drumbl, Atrocity, Punishment, and International Law (2007), 173–9 (discussing the role of expressivism in international criminal punishment).

240 Tallgren, supra note 13, at 583.

241 For an insightful discussion, see H. van der Wilt, ‘Equal Standards? On the Dialectics between National Jurisdictions and the International Criminal Court’, (2008) 8 International Criminal Law Review 229, at 244–5. This concern is apparent in the United States, too, where courts asked to deliberate on Alien Tort Act cases must decipher whether they apply domestic or international standards of complicity, before they attempt to ascertain which of the competing standards represents customary international law. Keitner, supra note 7, at 73–9.

242 UNSC, The Rule of Law and Transitional Justice in Conflict and Post Conflict Societies, Report of the Secretary-General, UN Doc. s/2004/616 (2004); M. Drumbl, ‘Collective Violence and Individual Punishment: The Criminality of Mass Atrocity’, (2005) 99 Northwestern University Law Review 539 (calculating average cost of trials as $25 million per case).

243 Zacklin, R., ‘The Failings of Ad Hoc Tribunals’, (2004) 2 JICJ 541Google Scholar, at 543 (‘The delays in bringing detainees to trial – and the trials themselves – have generally been so lengthy that questions have been raised as to the violation by the tribunals of the basic human rights guarantees set out in the [ICCPR]’); one commentator described the ICTY's proceedings as ‘as annoying and interminable as the Tour de France’; P. Hazan and J. T. Snyder, Justice in a Time of War: The True Story behind the International Criminal Tribunal for the Former Yugoslavia (2004), 187 (citing J. Finci).

244 Greenawalt, supra note 11; van der Wilt, supra note 241, at 244–5.

245 Bohlander, supra note 2, at 153 (discussing the use of unitary theory of perpetration (Einheitstaterbegriff) for administrative offences (Ordnungswidrigkeiten); for further information about administrative offences, see Krey, supra note 34, at 21.

246 Sullivan, supra note 14, at 156.

247 A. Norrie, Crime, Reason, and History: A Critical Introduction to Criminal Law (1993), 8–9.

248 Danner and Martinez, supra note 5, at 132.

249 Bronitt, supra note 167, at 55.

250 Fletcher, supra note 2, at 20.