Complicity and Criminal Liability in Rwanda: A Situationist Critique
- First Online:
- Cite this article as:
- Ciurria, M. Res Publica (2011) 17: 411. doi:10.1007/s11158-011-9157-z
- 140 Views
In Complicity and the Rwandan Genocide (2010b), Larry May argues that complicity can be the basis for criminal liability if two conditions are met: First, the person’s actions or inactions must contribute to the harm in question, and secondly, the person must know that his actions or inactions risk contributing to this harm. May also states that the threshold for guilt for criminal liability is higher than for moral responsibility. I agree with this latter claim, but I think that it casts doubt on May’s account of criminal liability, particular in so-called performance cases in which low-level participants merely fail to help. This is because it is far from clear that passive non-helpers are morally responsible for their participation in widespread harms. Situationism purports to show that passive bystanders typically are not morally responsible for their role in such harms, because they were behaving reasonably subject to the constraints they faced. In this paper, I assess this claim, and defend it on the basis of O. W. Holmes’ standard of the reasonable person as a guide to judging criminal complicity. Finally, I provide a situationist account of the Rwandan genocide, which focuses on the systemic causes and primary perpetrators of the genocide, rather than low-level participants.
First, the person’s actions or inactions must be causally efficacious at least in the sense that had the person not committed these actions or inactions the harm would have been made significantly less likely to occur. Second, the person must know that his or her actions or inactions risk contributing to a harmful enterprise, and must intend that these actions or inactions risk making this contribution (May 2010a, p. 158)
This analysis permits criminal prosecution in cases of bystander complicity, including ‘performance cases’ in which spectators provoke an offense by failing to intervene. May gives the example of a crowd of spectators who gather around two men apparently on the verge of fighting, and in doing so, ‘not only encourage, but play such an important causal role that they “make the fight” and can perhaps be held liable in ways that even the boxers themselves cannot’ (May 2010a, p. 161).
May wishes to distinguish this type of case—which merges legal and moral complicity—from similar cases which he says are ‘normally only counted’ as morally complicit (May 2010a, p. 165): for example, the famous Kitty Genovese case, in which Ms. Genovese was stabbed, the perpetrator fled, and then, when the neighbours failed to react, he returned to kill her. Although May attempts to provide an ‘in-principle basis’ for separating these cases (May 2010a, p. 165), I do not think that he succeeds. His argument is that legal complicity, unlike mere moral complicity, requires two factors: (1) in terms of actus reus, the person’s actions or inactions must be causally efficacious in the sense that had she not committed these actions or inactions, the harm would have been less likely to occur, and (2) in terms of mens rea, the person must know that her actions or inactions risk causing serious harm. May is clear that the actus reus is not the deciding factor: He says, ‘there must be some serious mens rea element, in addition to the actus reus component, in order to avoid the infantilization of the defendant’ (May 2010a, p. 169). However, the mens rea element is extremely loose, as it does not require that the perpetrator intends ‘the outcome of the principal’s enterprise,’ but only that she ‘knows that [she] is in a sense participating in an enterprise, by [her] commission or omission, which risks harm’ (May 2010a, p. 170). This is a highly qualified condition, and May acknowledges that it represents a ‘lower standard of mens rea than is often adopted’ (May 2010a, p. 169). It is adopted, apparently, to accommodate cases of recklessness, which are serious criminal offenses. However, in lowering the bar, May risks including cases such as Kitty Genovese’s, in which the non-helpers easily fit the description of (1) being counterfactually causally efficacious in a harmful enterprise, and (2) knowing that their actions risk contributing to such an enterprise (even if it is not the principal’s intended harm). Indeed, even if Ms. Genovese’s non-helping neighbours did not know that the principal was going to murder his victim, surely they must have known that his action ‘risked causing serious harm,’ and also that their inaction risked contributing to this harmful enterprise ‘in some sense.’ And this is all that is required to satisfy (2).
If this is right, then there is no in-principle basis for drawing a distinction between the boxing case and the Genovese case. Even if one type of non-helping is more criminally liable than another, this does not mean that it is exclusively criminally liable. On the other hand, since the two types are not categorically different, it is possible that neither is criminally liable. At the far end of the spectrum, we can argue that neither type is even morally complicit—that is, both types are legally and morally neutral. This is the interpretation that I favour. This is because it is far from clear that passive non-helping is even morally culpable, let along criminally liable. And, as May points out, ‘criminal liability generally requires crossing a higher bar than other forms of legal liability and certainly than is needed for moral responsibility’ (May 2010a, p. 167).
In moral philosophy, situationism has been leveled against the assumption that passive participants are morally responsible for their behavior. This criticism does not affect the bulk of May’s argument: it leaves intact his account of legal liability in cases of active legal complicity and passive duty-based responsibility (where one fails to do one’s duty). However, it calls into question his account of passive complicity in ‘performance cases’ such as the boxing match, the Genovese case and non-helping in Rwanda.
Gilbert Harman and the Milgram Experiments
‘the step-wise character of the shift from relatively unobjectionable behaviour to complicity in a pointless, cruel, and dangerous ordeal’, making it difficult to find a rationale to stop at one point rather than another… ‘the difficulty in moving from the intention to discontinue to the actual termination of their participation’, given the experimenter’s refusal to accept a simple announcement that the subject is quitting,’… [and the fact that] as the experiment went on, ‘the events that unfolded did not “make sense” or “add up”…. The subjects’ task was that of administering severe electric shocks to a learner who was no longer attempting to learn anything…. [T]here was simply no way for [subjects] to arrive at a stable “definition of the situation”.’ (Harman 1999, pp. 323–24, citing Ross and Nisbett 1991).
If Harman is right, then low-level participants are not morally responsible for their complicity in harmful enterprises where authority figures control the situation. Rather, the relevant details of the situation are to blame.
This seems to suggest that we can also excuse non-helpers in widespread harms such as genocide. However, it might be objected that Harman’s standard of moral evaluation, which is essentially folk-psychological, is not the correct one to apply in criminal proceedings. In particular, it might be argued that the law should employ a perfectionistic standard of evaluation (based on the ideal of the model citizen), rather than one that makes allowances for the typical frailties of normal human beings. This stricter view favours May’s account, as it entails holding passive bystanders criminally liable for their role in atrocities, and justifies criminal proceedings for the (at least) tens of thousands of non-helpers in the Rwandan genocide. However, it is not the approach taken in Western criminal law, where, since Vaughan v. Menlove (1837),1 the courts have applied the standard of the reasonable person. This view received an early defense in Oliver Wendell Holmes’ The Common Law (2009), which contends that in order to maintain social cohesion, the law must consider ‘what would be blameworthy in the average man, the man of ordinary intelligence and prudence, and determines liability by that’ (Holmes 1881, p. 73). This account has both inculpating and exculpating implications. On the one hand, if a person is ‘born hasty and awkward’ and is ‘always having accidents,’ his neighbours and the courts can nonetheless require him ‘to come up to their standard’ (Holmes 1881, p. 73). But on the other hand, ‘when a man has a distinct defect of such a nature that all can recognize it as making certain precautions impossible, he will not be held answerable for not taking them’ (Holmes 1881, p. 74).
If it strains credulity to imagine what the “ordinary man” would do in the position of a battered spouse, it is probably because men do not typically find themselves in that situation. Some women do, however. The definition of what is reasonable must be adapted to circumstances which are, by and large, foreign to the world inhabited by the hypothetical “reasonable man” (1990, p. 277).
Thus, the standard was re-defined to accommodate the situation of the battered woman.
I believe that a similar sort of adjustment needs to be made in cases of widespread harms such as genocide. That is, in trying to determine what the ‘average’ person ‘of ordinary intelligence and prudence’ would do in the context of an atrocity, we cannot begin by considering what we think the average Canadian or American or European would do, as these are several quite different epistemic positions. Instead, we must place special deliberative weight ex post on what the majority of people in the atrocity actually did under the inordinate pressures of their situation.
This type of approach is defended by Peter Stone in his insightful paper, ‘Rationality, intelligibility and interpretation’ (2009), where he suggests that although the natural interpretive approach is to view others as ‘behaving by and large’ as we think that we would behave in their situation (Stone 2009, p. 51), there are cases where this approach is precisely backward: namely, if a whole group of individuals appear to be acting irrationally under a similar set of circumstances. Then we must consider the possibility that they face constraints that we ‘did not recognize before,’ and that, under these constraints, we would likely behave the same way. Hence, we must consider the possibility that these individuals are acting reasonably subject to their constraints (Stone 2009, p. 50). This offers a compelling modification to Holmes’ standard of the reasonable person, i.e., in cases of collective harms, the standard of the reasonable person must be appropriately contextualized in light of how the majority of people responded to the constraints of their situation. Based on this standard, we would be justified in exonerating the large number of passive non-helpers in Rwanda.
This accords with Harman’s extenuating judgment of the Milgram participants, and militates against May’s account of passive complicity, which applies a higher standard than that typically assumed in Western criminal law. For the sake of fairness and consistency, participants in the Rwandan genocide must be judged according to an appropriately contextualized standard of evaluation—one that is likely to excuse merely passive non-helpers.
Philip Zimbardo, the Stanford Prison Experiment (SPE) and Abu Ghraib
Philip Zimbardo, the author of the famous Stanford Prison Experiment (SPE) (1971), in which 24 students were assigned to role-play prisoners and guards in a simulated prison environment, criticizes ‘individualistic’ explanations, which blame the individual wrongdoer, rather than the ‘system forces’ that ‘create and maintain situations’ (Zimbardo 2007, p. viii). In The Lucifer Effect (2007), Zimbardo laments his authorial role in the SPE, in which the ‘guards’ quickly resorted to subjecting the ‘prisoners’ to humiliating and dehumanizing treatment, including sleep deprivation, solitary confinement, starvation, force-feeding, and sexual humiliation. Witnessing this, Zimbardo terminated the experiment after 6 days (rather than the scheduled 14), but only after his then-girlfriend accused him, ‘What you are doing to those boys is a terrible thing!’ (Zimbardo 2007, p. 168). As an outsider, she brought a more objective perspective to the situation.
Zimbardo compares this climate of abuse to the Abu Ghraib Prison, where in 2004 US Military Police were charged with ‘punching, kicking and slapping detainees… forcing detainees to remove their clothing and keeping them naked for several days at a time… arranging naked male detainees in a pile and then jumping on them… using military working dogs (without muzzles) to intimidate and frighten detainess…’ (Zimbardo 2007, p. 357), among other offenses. In October 2004, Zimbardo testified as an expert witness for Staff Sergeant Ivan ‘Chip’ Frederick, who was accused of maltreatment, assault and indecent acts. Zimbardo’s defense was, in brief, that Sergeant Frederick was not a ‘bad apple’; instead, Abu Ghraib was a ‘rotten barrel.’ Frederick was, Zimbardo points out, a previously psychologically normal, ‘All-American poster soldier’ (Zimbardo 2007, p. 344), who, after arriving at Abu Ghraib, succumbed to the pressures of ‘job stress and burnout potential’ (Zimbardo 2007, p. 347), nonstop night work, irregular meals, unsanitary conditions, and a pre-existing culture of abuse and sexual humiliation. Consequently, Zimbardo says, Frederick was not culpable, but rather, the system, and particularly then-President George W. Bush and his advisers, were to blame ‘for their role in defining torture as an acceptable practice’ (Zimbardo 2007, p. 378).
I should clarify that Zimbardo does not think that no one is ever morally culpable. Rather, he believes that in cases of widespread harm, the systemic causes, or primary instigators, typically deserve more blame than low-level participants—hence his accusation of the Bush administration for facilitating the abuses at Abu Ghraib. He also suggests that personal responsibility may depend on the agent’s motivations or subjective construal of the situation. Hence, he describes one of the more abusive ‘guards’ in the SPE, Hellmann, as being ‘able to go beyond the demands of the situation’ (Zimbardo 2007, p. 218) to fulfill his own supra-sadistic desires. He also takes responsibility for his administrative role, which he calls ‘distressing,’ ‘improper’ and ‘bizarre’ (Zimbardo 2007, p. 218). Others have pointed out that the participants in the Milgram experiment manifested varying degrees of distress, based on their emotional orientations; yet a vast majority ‘expressed considerable distress’ (Smith and Mackie 2000, p. 396). Connecting this with the Holmes-Stone hypothesis that we cannot blame someone for doing what it was reasonable for someone in his position to do, we may say that Hellman, Zimbardo and the Bush administration acted unreasonably, since they went beyond what other participants did or intended in the same type of situation. Specifically, Hellmann enjoyed his abusive role, and Zimbardo embraced his administrative role in the SPE (albeit temporarily), while the other participants expressed disdain and recoiled at their actions. And the Bush administration willfully authorized torture, whereas the Abu Ghraib guards typically performed their roles either against their inclination, or under psychological duress.
The implication is that, whereas primary instigators may be held accountable for their role in atrocities, most low-level participants are not morally responsible for their participation. Admittedly, some participants may demonstrate extraordinary callousness; but it is unreasonable, based on the evidence, to assume that most participants are prone to callousness and voluntary maltreatment. It is more likely, as situationists have proposed, that they are extremely vulnerable to situational pressures, and thus, that situations and principals are to blame.
This calls into question May’s claim that ‘we should allow for many gacaca, “trials in the grass,” of low level participants in genocide’ (May 2010a, p. 135). For if, as May attests, the ‘threshold for guilt’ for criminal liability is higher than for moral responsibility, and, as I have tried to show, most low-level participants in Rwanda are not morally culpable, then many gacaca trials are not warranted. [May says that ‘something like the gacaca process is at least potentially justifiable, even though it involves tens of thousands persecuted for the genocide’ (May 2010a, p. 151), emphasis mine]. However, if I am right, and the number of culpable actors is small, it should be possible for national courts to handle all genocide cases, eliminating the need for gacaca.
A Situationist Account of the Rwandan Genocide
the colonial ideology of racial division; the economic and political crises of the 1980s and early 1990s; the previously very highly organised nature of Rwandan society; and the fragile regional and class base of a political faction determined to hold on to state power at any cost. Other important contributing factors were the use of sophisticated propaganda techniques, and the escalatory cycle of violence within Rwanda and Burundi (Hintjens 1999, p. 287)
This type of explanation differs from the ‘individualistic’ accounts that Zimbardo criticizes, as well as ‘dispositional analyses,’ which focus on the moral dispositions of particular agents. Nonetheless, this type of explanation is theoretically compatible with morally blaming, and criminally prosecuting, high-level instigators who are responsible for triggering the causes cited, e.g., propaganda and escalating cycles of violence. These include the Rwandan Patriotic Front, who orchestrated the genocide, and perhaps also key members of the World Bank/IMF who implemented ‘a package of structural adjustment measures’ that immediately devalued Rwanda’s national currency by two thirds, and, in conjunction with other pressures, ‘dramatically worsened the lives of most Rwandans’ (Hintjens 1999, p. 257). These instigators are morally blameworthy and legally liable because, like Guard Hellmann from the SPE, they went beyond the demands of the situation, far surpassing the contributions of the majority of participants even in high-ranking positions, such as the Bahutu elite. Thus, they at least prima facie surpassed what a reasonable person would do in their position. Indeed, they may even have performed acts that they themselves realized were wrong at the time, manifesting a particularly culpable motivational structure.
May seems to prefer an individualistic approach. As we saw, he favours ‘wide-ranging trials’ (May 2010a, p. 137), and claims that prosecuting ‘tens of thousands’ of defendants is ‘potentially justified’ (May 2010a, p. 151). He also says that complicity should remain ‘focused on the individual agent rather than the group’ (May 2010a, b, p. 149), which differs from the situationists’ focus on groups and systems. May defends this position because he believes that ‘having so many trials… makes the members of the community, those who were legally complicit and those who were only morally complicit, confront their connection to the genocide,’ which he hopes will ‘lead to changes in their behavior in the future’ (May 2010a, p. 151). However, it is far from clear that gacaca proceedings advance this aim. Sarkin (2001) writes that ‘so many years have elapsed since the genocide of 1994 that the authorities cannot, and should not, seek to prosecute all those accused of participating in the slaughter. Perhaps only the architects and main organizers ought to be prosecuted’ (Sarkin 2001, p. 144). He worries that the courts cannot ‘dispense justice in a fair and impartial manner’ (Sarkin 2001, p. 162). On a purely philosophical level, there is reason to worry that gacaca trials will inadvertently scapegoat low-level participants, diverting attention away from the broader range of causal factors, both human and inhuman. These are the principals cited by Hintjens, including Eurpoean colonizers, Rwandan propagandists, and members of the World Bank/IMF.
May recognizes this general concern. In a later chapter in his book, he notes that ‘demonization of the defendant retards reconciliation,’ and thus ‘demonization should be shunned’ (May 2010a, b, p. 252). He also repudiates scapegoating, but he thinks that scapegoating happens when criminal proceedings target only principal perpetrators, and omit complicit bystanders (May 2010a, b, p. 259). As a situationist, I agree that this focus embodies a type of scapegoating; but I also believe that prosecuting numerous complicit bystanders engenders a type of scapegoating, as it diverts attention away from the main causes of the genocide, including foreign financial powers that precipitated a devaluation of Rwanda’s national currency and an inflation of Rwanda’s external debt. With respect to demonization, I worry that blaming complicit bystanders may inadvertently demonize them, as blame, in general, suggests that the target has acted on beliefs that a basically rational and moral person would eschew, and thus deserves special unhappiness or misfortune. (Michael Philips, for instance, defines blame as adopting an attitude that someone is ‘less worthy of happiness or good fortune’ and ‘perhaps deserves to suffer’ (Philips 1987, p. 143)). Thus, blaming complicit bystanders in Rwanda suggests that they lacked rationality and moral understanding—i.e., that they were, to an extent, demonic. Yet the social psychological evidence suggests that most complicit wrongdoers are reasonable and moral, but—like every human being—vulnerable to overwhelming situational pressures. They are not demons, but rather ordinarily sensitive and vulnerable human beings. Whether this sensitivity is pro-social or anti-social ultimately depends on the circumstances.
I have disputed May’s claim that wide-ranging gacaca trials, potentially involving tens of thousands of low-level participants, should be implemented to ease post-genocidal tensions in Rwanda. However, gacaca trials do advance some legitimate aims. In particular, they help educate Rwandans about the causes of the genocide, thereby militating against future recurrences. May points out that trials in general play a ‘public educative role’ (May 2010a, p. 151). However, gacaca trials are also supposed to play a judicial role, and it is this role which is misguided. For if, as situationism suggests, passive bystanders very well may not be morally responsible for their role in widespread harmful enterprises, and criminal liability requires crossing a higher bar than moral culpability, then it is dubious whether wide-ranging trials are justified. Furthermore, since trials typically target particular individuals rather than groups or systems, they risk diverting attention away from the principal causes of genocide. Thus, it may be more fair as well as more productive as a means of preventing future atrocities to develop educational initiatives rather than resorting to criminal prosecution.
In this case, the defendant was found criminally negligent after building a haystack on the border of the plaintiff’s property, and ignoring repeated warnings that the haystack was in danger of catching fire, saying that he ‘would chance it.’ Predictably, the haystack caught fire and destroyed the plaintiff’s buildings. (1837) 3 Bing. N.C. 468.