Why Serious International Crimes Might Not Seem ‘Manifestly Unlawful’ to Low-level Perpetrators

Article 33 of the International Criminal Court Statute allows low-ranking perpetrators to ç in exceptional cases ç rely on the defence of superior orders. By doing so, Article 33 might be seen as an acknowledgement that within a specific context orders to commit international crimes might not always be manifest unlawful. Article 33(2), however, restricts the possibility to rely on this defence to perpetrators of war crimes and denies perpetrators of crimes against humanity and genocide a similar defence, since according to Article 33(2), such orders are considered always to be manifestly unlawful. This contribution questions whether such a distinction should be made. Many low-ranking perpetrators involved in such crimes by following superior orders seem to genuinely believe that they were doing the ‘right thing’. This article seeks to explain how these perpetrators might have come to such a belief, and the challenge this might represent to the core principles which underpin the concept of individual criminal responsibility.


Introduction
Starting with the very first post-World War II trials the question of whether superior orders could be a (complete) defence has been at the forefront of discussion. Many defendants claimed, expressly or by implication, that they did nothing more than follow orders, that they could not have done otherwise, that they were just small cogs in a big machine, and thus that they should not be punished. The Nuremberg Tribunal rejected this argument with the famous statement that '[c]rimes against international law are committed by men, not abstract entities, and only by punishing individuals who commit such crimes can the provisions of international law be enforced.' 1 When in 1998 the International Criminal Court (ICC) Statute was agreed upon, Article 33(1) of the ICC Statute provided: The fact that a crime within the jurisdiction of the court is committed pursuant to an order of a Government or of a superior, whether military or civilian, shall not relieve that person of criminal responsibility, unless: (a) the person was under a legal obligation to obey orders of the government or the superior in question; (b) the person did not know that the order was unlawful; and (c) the order was not manifestly unlawful.
This provision was an innovation compared to past formulas, in that it acknowledged that there were indeed cases where potentially an order to commit international crimes might not be manifestly unlawful. The practical consequence being that in those cases, and if all other criteria would have been met, low-ranking perpetrators could rely on the defence of superior orders. This is furthermore an important acknowledgement of the core principle of criminal law that only those who can be blamed, i.e. who have a 'guilty mind' can be prosecuted and punished since it allows low-ranking perpetrators who were not aware of the unlawfulness of the order to rely on this defence. The subsequent paragraph however clarifies that the defence of superior orders may only be raised in cases of war crimes and not for orders to commit crimes against humanity or genocide because these crimes are, according to the Article 33(2), deemed to be manifestly unlawful by definition. Article 33(2) literally reads: 'For the purpose of this article, orders to commit genocide or crimes against humanity are manifestly unlawful.' 2 From theoretical, legal and moral perspectives, nothing is wrong with restricting the extent to which perpetrators who commit horrendous crimes can rely on the defence of superior orders as Article 33(1) does by applying strict criteria. The problem, however, is that Article 33(2) ç if read in combination with Article 33(1) ç seems to suggest that orders to commit crimes against humanity and genocide are always and by definition manifest unlawful and recognized as such by all recipients of such orders. Yet this is not the case. This contribution will show that extensive research on perpetrators of international crimes has shown that many low-ranking perpetrators do not always necessarily recognize such orders as manifestly illegal. Quite the contrary: many of them believe that ç within the specific context in which they were given the orders and because of the fact that the orders were given by someone whom they perceived as a legitimate source of authority ç they are doing the right thing and thus their behaviour is legitimate and lawful.
Modern history is, sadly, replete with terrible atrocities. Confronted with such atrocities, we instinctively tend to distance ourselves from the crimes, as well as the perpetrators thereof whom we tend to regard as evil or insane. 3 International criminal law may reflect a similar assumption, since the idea of 'international criminals' is often associated with the historical legal notion of hostes humanis generis: enemies of all mankind. 4 This qualification stands in stark contrast to the findings of scholars from social sciences and genocide studies, who have found that such perpetrators are frequently ordinary people, not so very different from us, who believe they did the right thing even when carrying out the conduct for which they were subsequently tried or punished. 5 These findings challenge the very notion that orders which eventually amount to crimes against humanity or genocide are always recognized by the recipients of these orders as manifest unlawful.
This article does not deny that orders to commit crimes against humanity and genocide should always be deemed unlawful. It challenges the assumption in Article 33(2) of the ICC Statute that such orders are always, and should always be, recognized by perpetrators at the time the crimes are committed, as being manifestly illegal.
Although international crimes can of course be committed by state actors and non-state actors, this article will for the sake of clarity focus on state officials who commit international crimes on orders of the state or through the state apparatus, such as the police or military only. In order to understand the perpetrators of such state-induced international crimes we first need to look at the political and institutional context in which they commit their crimes. Next the focus will be on 'crimes of obedience' as a concept and it will be shown how perpetrators generally change and transform after their first 'kill'.
We will then consider how perpetrators look back on their actions, and lastly draw some conclusions, discuss their implications and consider lessons learned.

The Political Context
The use of force by the state and its agents ç whether against another state or an armed group or even its individual citizens ç is not by definition illegal. States have a monopoly on the use of force, and may within certain limits employ force lawfully, e.g. to fight crime, maintain order and defend the country against external threats. Thus, although states are obviously bound by both national and international law (for the state as such the United Nations Charter and its ban on the use of force are paramount), including international human rights law and international humanitarian law, the legality of particular applications of force may depend on broader legal and factual factors, of which the direct perpetrator may not always be aware. Accordingly, with the exception of some obvious prohibitions such as extreme torture or certain forms of sexual violence (which violate non-derogable rights), a state official may well become accustomed to acting on the basis of orders to use force without having an independent basis to assess their legality. State officials may thus come to trust their superiors to have 'taken care of the law', and so suspend the natural reticence in employing force when simply ordered to do so.
Of course, we know that trust in the state is not always well-founded. Superiors cannot always safely be relied upon. In particular, senior persons in state hierarchies may well abuse the state's monopoly of the use of force, and subvert it to their own ends. On occasion, this may be accompanied by an obvious departure from 'business as usual' ç for example, a sudden coup in a hitherto stable democracy ç but equally it may not. The descent from legitimate uses of force (or, at least, routine uses of force, even if not in compliance with international human rights law) into international crimes may not be so clear. But indeed, this can also be true even when both functionaries and leaders act in 'good faith', addressing a threat that they perceive as real, even if it leads to a disproportionate and objectively improper response.
The response of the United States to the 9/11 attacks is a good example. The Bush administration launched the 'War on Terror' as a response to these attacks. They felt that their country and way of life was under attack and felt the strong urge to defend themselves. In doing so, their apparent resort in some cases to unlawful means and methods seems to have been motivated by a belief that lawful means were insufficient to achieve their legitimate goal. There were nonetheless clear signs of an effort to confront the boundaries of national and international law, and to push them sufficiently far back as to cloak apparent violations in spurious legality. 7 This effort, with inevitable implications for the subjective perceptions of state functionaries, was further influenced by a worldview which seemed to condone violating those rules, in pursuit of a 'higher purpose'. In essence, the conflict was reduced to a storybook portrayal of good versus evil, in which the 'end' seemed to justify the 'means'. Such 'rationalizing' behaviour enabled state functionaries to commit international crimes without realizing that the orders they followed were manifestly unlawful. More so because, other countries were sympathetic to the US cause: protecting the western way of life against religious extremism and terrorism. Left unchecked, this can contribute to a widespread 're-framing' of conduct, from being deviant to being acceptable. 8 It is striking that at least parts of the American public may continue to accept torture as a legitimate and necessary means to fight some kinds of threats. 9 For all these reasons, to understand the perpetrator's view of many international crimes, we have to understand that many perpetrators ç no matter what side they are fighting on ç not only believe they are fighting in a battle of good versus evil, but also that they are on the 'good' side. Indeed, people may have a general tendency to divide the world into 'good' and 'bad', and believe in what Baumeister dubbed the 'myth of pure evil'. 10 Such a view may easily lead to the misconception that all means and measures in pursuit of the 'good' are acceptable and legitimate. 11 And thus many evil acts are 'performed by people who believe they are doing something supremely good'. 12 Evidence of this kind of thinking among perpetrators of international crimes is shown by the following illustrative quotes. Of the My Lai killings, in Vietnam: Of the 'dirty war', in Argentina: I might have made some small mistakes, but in the big things I don't repent anything. :::We were all convinced that it was the best thing that could be done for the country ::: I believed that both my commanding officers and myself were the saviors of the fatherland. 14 As a human being, face to face with the enemy, when you kill you have to question it. I told you that I came back from the first flight feeling bad. Bad. I did not feel good, but I did not doubt that militarily I had carried out an order I was completely convinced about. 15 It was necessary to persecute the terrorists and to interrogate them using whatever means. The rationale was to torture some terrorists to protect the majority of the population. 16 Involvement in the dirty war was service to re-establish a democratic style of life in Argentina. 17 Of the genocide, in Rwanda: It was just my job. At that time, I did not realize that it was genocide, I was just doing my job: kill the enemy. Only later did I realize that it was genocide. No one pushed me, I did it willingly. I was told that the enemy had to be killed and I believed it, and I knew that they had to be killed. ::: at the time I thought I was doing a good job. 18 All these perpetrators believe that they were doing the right thing. On what basis can we say that these perpetrators should have known better? Are there universal, consistent, legal and moral norms which each and every individual knows and recognizes to be relevant to their conduct? International law has developed over the last 70 years and that reflects a relatively sudden and marked departure from centuries of previous practice in tolerating, or indeed perpetuating, conduct including public execution, torture 19 and slavery. Today, all of these behaviours are considered international crimes. But they did not always. So what is deemed lawful or unlawful changes over time and furthermore decided upon by states which have the executive, legislative and judicial power to do so. In this context, it is not unsurprising that state functionaries, quite rationally, form the view that the state remains the main arbiter of what is right and wrong in a given situation. And, as the following paragraphs further explain, state agents are further conditioned by governmental institutions to accept their judgements and just do as they are told. Thus, one functionary in Argentina during the Dirty War, described his compulsion to follow orders, because if he did not: '::: he would be betraying the confidence the nation places in him when it entrusts him with its most precious things: the care of its land and traditions and the blood of its children'. 20 It is thus not so remarkable that state agents may place a disproportionate degree of weight in the legal and factual evaluations of their superiors (who will typically be higher state functionaries), and may not themselves question even the most objectively dubious orders, and ask themselves whether these orders are right or wrong.

The Institutional Context
Most perpetrators of state-induced international crimes are members of militarized units such as the army, police or special/secret forces. Militarized organizations are characterized by a very clear and strict hierarchy in which people tend to be recognized by their ranks or functions, rather than by their name or personal identity. Recruits are, above all, required to be loyal and obedient to their superiors and conform to the rules, norms and values within the militarized organization. 21 Disobeying orders can lead not only to disciplinary measures but also sometimes criminal prosecution and punishment for some of the most serious offences such as high treason, which especially in wartime often carries the most severe sentence. For all these reasons, many soldiers consequently may feel that they have no other option than to obey: it is their task and duty, whether they like the order or not. The Greek torturer Petrou who was convicted by the Greek courts for torture despite stating that he was ordered and trained to do so explained: 'If the officers ordered you to beat, you beat. If they ordered you to stop, you stopped. You never thought you could do otherwise.' 22 Military training has several purposes. Of course, recruits need to be fit enough to do the job, and they have to acquire certain skills including, perhaps most importantly, how to use weapons. But training is perhaps even more significant for its emotional and moral impact, rather than just its technical aspects. Recruits have to be desensitized to violence and learn to overcome natural inclinations not to use violence; ultimately, they have to be prepared to kill a fellow human being ç and perhaps risk their own life ç for no reason other than that they are told to do so. 23 Research by Marshall has shown that, during World War II, only 15^20% of the American soldiers returned enemy fire. 24 Yet this apparent resistance to the use of force could be altered by altering the training. A focus on drills in which recruits were conditioned to fire their weapons in an automated response seemed to show success ç by the Korean war, 50% are reported to have returned fire, while in the Vietnam War, 90% did so. 25 This example graphically illustrates how soldiers can be trained to overcome their resistance to certain conduct, including the use of force. Likewise, training is important in ensuring that recruits obey all orders, unconditionally. This is important as soldiers need to fight and win wars, and thus need to be determined and courageous to overcome, perhaps, the most fundamental human fear: personal death. In order to achieve this, soldiers are to a large extent deindividualized and depersonalized during training: they get an identical uniform and haircut, minimal privacy, their position is determined by number or rank not name and they eat, sleep, and train as a group. Non-conformity is likely to be punished. 26 Recruits are measured and rewarded only according to the needs and values of the military: whether or not they are good soldiers. Thus, a typical introduction to a new recruit: From this time on I will be your mother, your father, your sister, and your brother. I will be your best friend and your worst enemy. I will be there to wake you up in the morning, and I will be there to tuck you in at night. You will jump when I say 'frog' and when I tell you to s__ your only question will be 'what colour'. IS THAT CLEAR? 27 Lastly, and importantly, these social pressures do not stop when training stops. As recruits quickly learn, military organizations place a value on establishing and maintaining a certain corp d'esprit which is close to holy. Military bases are a different world from civilian life, with their own rules and traditions. And not only are soldiers thus marked out from civilians, but soldiers may also mark themselves out from one another, by separating themselves into more elite, more dedicated groups (e.g. special forces). Such transitions may be marked by initiation rites, whether formal or informal. 28 Fundamentally, however, all soldiers have to support each other, and be prepared to kill and die for each other, throughout their careers. This is one of the defining principles of military life. By contrast, in civilian society, the rule against killing is one of the most prominent and universally accepted norms and values. Ideally, recruits should also be told that the use of force even within a period of war is limited and that the rules of international 24 As quoted by Grossman, supra note 23, at 4. humanitarian law should be followed. Within peace-loving states, which are not at war, soldiers are indeed told this but the many crimes committed throughout the last century ç also since the time the Geneva Conventions have been in place ç show that most recruits, especially those banded together in armed groups during periods of emerging conflict, have not been trained to do so or that they have easily forgotten these rules or disregarded them because of the overriding need to follow orders as given, as well as the need to protect their nation.

A Crime of Obedience
State-induced international crimes committed by state agents can be qualified as so-called 'crimes of obedience'. This term was defined by Kelman and Hamilton as: 'an act performed in response to orders from authority that is considered illegal or immoral by the international community'. 29 In other words, for most international crimes, the perpetrators may come to believe that certain behaviour (i.e. killing, raping, torturing or maiming) is in fact considered acceptable by the authorities, and therefore condoned or even encouraged. They believe that the behaviour is ç under the very specific circumstances, which may distinguish the situation from the 'normal' rules ç legitimate. Moreover, although orders to perpetrate such conduct may sometimes be direct and clear, they are far more often indirect or implicit, and phrased in a manner which may seem to conceal their criminal nature. For instance, rather than issuing overt instructions to subordinates to kill or torture, superiors may simply order 'you know what to do with them' or 'soften them up'. Yet, no matter whether the orders are direct or indirect, the law still seems to assume that subordinates can recognize orders which are manifestly illegal. But in this regard, there may be at least three well-established aspects of human behaviour which show that recognizing the manifest illegality is not so self-evident as it might seem: the human tendency towards obedience to authority, social conformity and the 'neutralizing' use of language. We shall briefly consider each in turn. First, the experiments conducted by Stanley Milgram suggest that people tend to obey orders from an authority that they perceive as legitimate. 30 Thus, in these experiments, the subjects were made to believe that they were participating in an experiment testing the effect of punishment on the ability to learn and remember. But in reality, however, the experiment was designed to test whether ordinary people would follow an order which can be seen as improper and immoral (to give a fellow human being electric shocks of up to 450 volts). It turned out that 65% of the subjects did as requested, showing how obedient people tend to be when given orders by someone whom they believed was a legitimate source of authority. 31 The experiment also showed that people are not by nature cruel or enthusiastic about hurting others. 32 When interviewed subsequently, the subjects confirmed that they felt they were caught in a dilemma between two very different norms: the rule to obey an authority and the rule not to hurt others. Yet, the former proved to be stronger than the latter. This in turn may reflect the pressure felt by military subordinates in the field when they receive an order. A perpetrator explains: 'The order had been given and I had to carry it out.' Milgram's experiment is of course, a laboratory situation and thus very different from real life. However, it remains true that most people recognize the state ç with its executive, legislative and judicial power ç as the ultimate authority, and thus it is likely that, since humans do have a 'tendency towards obedience', as Milgram's experiment shows, this will apply par excellence to orders from state authorities. As Hoess, camp commander at Auschwitz, further explained: 'Whether the reason behind the extermination of the Jews was necessary or not was not something on which I could allow myself to form an opinion::: .' 33 This comment is notable because it illustrates not only that people may tend to obey, or even feel compelled to obey, the instructions of a perceived authority, but also that in so doing they may tend to absolve themselves of any responsibility in relation to the ordered conduct and assume that the order is necessary and legitimate, taken the circumstances. Recent research in neuroscience and biology shows that 'coercive instructions caused participants to experience less agency over the harmful outcomes of their actions'. 34 Thus, when perpetrators state that they 'only followed orders', they may not just be trying to excuse their conduct but revealing a genuine change in their perception of their own responsibility. 35 This may also be consistent with some anecdotal evidence from torturers during the colonels regime in Greece, for example: 'I feel the need to say to the respected Tribunal and to the Greek people that I am a person like you, like your neighbour, like a friend. And when I tortured, the officer Spanos tortured with my hand. 36  'They changed us into instruments. People without a will of their own. Who obey ::: . You were trained not to think.' 37 And, finally, 'You have to realize that you yourself have not killed anyone and that you are just an instrument of the state.' 38 They generally seem to feel that they merely have a functional/ institutional responsibility, i.e. to follow orders while their superiors have the duty to ensure that these orders are legitimate. Subordinates believe that their superiors make such judgements and trust them to do so.
Secondly, there is also evidence that people have a general tendency to conform to the behaviour of their peers. Humans are social beings. 39 We have a very strong natural inclination to look at others in order to find clues on what is right and wrong, and how we should behave. Asch, for example, tested how social influences shape a person's judgements and beliefs by showing a group of people a set of lines, and asking them to indicate which lines were equal in length. When alone, very few subjects made mistakes: only 1%. However, when working within the group ç whose other members on instruction of the experimenter all deliberately gave a wrong answer ç 38% of the subjects yielded to the majority opinion and likewise gave the wrong answer. Asch proposed two possible explanations: either the subjects sincerely doubted their own perception because it was apparently inconsistent with that of their peers, or that the subjects knew perfectly well that their perception was right, but did not want or dare to stand out from the group. 40 Another good example of the tendency to conform may be the 'bystander effect': the reported phenomenon that, the greater the number of people witness an emergency, the less likely it becomes that any individual intervenes. Again, this might be explained by each individual's reference to the reactions of other people as a guide to their own behaviour. Thus, although a person might intervene if they were the only witness to someone drowning, they might not intervene if others are present yet do not react ç in such circumstances, people may tend to reason that there is no genuine emergency at all, since no one else has intervened. People in other words strongly rely on others when assessing a specific situation. 41 This is equally true when in a period of war, and orders are given. If others follow up on these orders, recruits will believe the orders to be legitimate.
Ordinarily, intentional and unlawful violence is not only considered criminal but in peace time morally reprehensible.Yet, in the context of an armed conflict or a widespread or systematic attack against a civilian population, this may change and people might come to see killing a fellow human being as legitimate or even a moral or social duty. 42 This is well-illustrated by Christopher Browning in his book Ordinary Men, recounting that many perpetrators joined in mass executions because 'they did not want to be considered a coward'. 43 In extreme circumstances, socially and morally acceptable behaviour may become completely reversed, with violence not only normalized but seen as constructive or desirable. This is clearly illustrated by a speech given by Heinrich Himmler to the Einsatzgruppen in October 1943 in which he tried to turn crimes into heroic acts: Most of you know what it means when 100 corpses lie there, or 500 lie there, or 1000 lie there. To have gone through this and ç apart from the exceptions caused by human weakness ç to have remained decent, that has hardened us. That is a page of glory in our history never written and never to be written. 44 The influence of this appeal is further illustrated by the separate observation of a Schutzstaffel officer that: We were Germans best and hardest. Every single one of us dedicated himself to others.What held us together was an alliance of comradeship. Not even the bond of marriage can be stronger. Comradeship is everything. It gave us the mental and physical strength to do what others were too weak to do. 45 Crimes ordered by the state are often framed as conduct which is necessary or even heroic ç which brings us to our third point: the use of language. Language can be used to hide the true meaning and moral implications of human conduct, and thus to help perpetrators avoid confronting the reality of their actions. For example, in many cases, criminal conduct is frequently described or re-characterized by the use of euphemisms. When policymakers in Nazi Germany decided to exterminate the Jews, they did not call it genocide or murder but the 'Final Solution to the Jewish Problem'. Using this term not only denied what was going on (mass killing) but even seemed to provide some legitimacy: the notion of a solution to a problem which must be solved. Likewise, in South Africa, the policy of racial discrimination was called ' Apartheid' ç a term which seemed merely to indicate that differences among people should be taken into account. In various places and times, torture has been described euphemistically as 'gathering intelligence', 'softening them up' or as 'enhanced interrogation techniques'. In the same way, even specific techniques might be given neutral, every day or light-hearted names: 'parrot peach', 'refrigerator', 'tea party' and so on. 46  referred to by their names but instead dehumanized. By progressively identifying, stigmatizing, isolating and vilifying a victim group, they can be effectively placed outside the moral universe of obligation. In Nazi Germany, the Jews were called 'lebens unwu« rtiges Leben' ç literally, lifes unworthy of living. In Rwanda, the Tutsis were called 'cockroaches'. By using this type of language the crimes are presented as legitimate behaviour. As a consequence perpetrators may well stick to their own moral norms while committing crimes since they exclude certain groups from their right to being treated as human beings: 'It's not easy to beat a person and it isn't easy to beat a woman and a child ::: beating a person is unpleasant. But ::: . [t]his is a different population ::: [t]his is a different kind of citizen.' 47 For all these reasons, therefore, subordinates who receive a 'criminal' order may not only feel a strong pressure to obey but may also often subjectively consider the order to be legitimate. In such circumstances, very few individuals are likely to disobey. Moreover, interestingly, it may be that those few courageous individuals who do wish to disobey will resort to evasive behaviour rather than open defiance. This can take many forms including deliberately missing when told to shoot someone, turning up late for work or even self-inflicted injury. One of the soldiers, in My Lai, preferred to shoot himself in the foot rather than to challenge his company commander. 48 Unfortunately, for the majority of people who feel compelled to carry out the criminal order and actually do so, things will get even worse as they fall victim to a number of further psychological processes resulting from their first crime which will make them believe that they were doing the right thing.

After the First 'Kill' ç Transformative Psychological Processes
Having committed their first crime, almost all perpetrators are shocked by what they have done. 49 Some will suddenly realize that the victim is human after all ç a husband, mother or someone's friend ç and thus not so very different after all. Perpetrators may feel horror, shock, distress, disgust, anxiety, and fear or animal pity. They may also experience physical reactions including dizziness, nausea, vomiting, crying, panic and, in a later stage, depression, nightmares and hallucinations. 50 Academic literature is divided about the nature of this so-called 'perpetrator disgust': some scholars call it a sign of human morality while others call it 'aesthetic'disgust which has nothing to do with morality. 51 Whatever the origin of these feelings and emotions, they give rise to a strong feeling of cognitive dissonance. The perpetrators believe themselves to be doing the right thing ç yet suddenly they feel terrible about what they have just done. This state is well described in psychological literature, and usually is experienced as a nagging feeling which people wish to resolve as quickly as possible. 52 This can occur either by changing behaviour ç never repeating the crime again ç or their perception of that behaviour. Sadly, although some perpetrators may manage (and have the opportunity) to modify their behaviour, 53 most will not. Instead, they will seek out ways to soothe their consciences, to justify or legitimize what they have done and thus to come to terms with the horror of the act itself. This in turn not only increases the risk of re-offending, but also further shifts their moral compass. As one perpetrator recalled: This need to 'get used' to the horror is partly a conscious process ç but also partially an unconscious one, a reaction to protect our own psychological welfare and well-being. People want to feel good about themselves ç indeed, want to be good people, according to their own norms and values ç and, hence, when they act inconsistently with that desire, they tend to experience a strong need to absolve themselves of blame or guilt. Aronson notes that 'people are motivated to justify their own actions, beliefs, and feelings. When they do something they will try, if at all possible, to convince themselves (and others) that it was a logical, reasonable thing to do'. 55 Thus, in order to stay psychologically healthy, perpetrators may tend to force themselves to get used to the violence. One means of doing so, as Bandura explained, is through moral disengagement. 56 Seeking moral justification for their actions, perpetrators may cognitively reconstruct the meaning of their behaviour. Thus, a killing is no longer murder, but a necessary means to a noble end: to protect the country or to create a better world. By this means, a perpetrator may continue to see himself as a moral agent. Bandura concluded that: 'this is especially evident when social pressures impel people to act counter to standards of conduct that would ordinarily make them feel ashamed and self-contemptuous. By engaging in self-deceptive machinations, they can behave inhumanely without self-condemnation'. 57 An American soldier involved in My Lai remembers that after killing the first few people, killing became a routine: I went to turn her and there was a little baby with her that I had also killed. The baby's face was half gone. My mind just went. The training came to me and I just started killing. Old men, women, children, water buffaloes, everything. We were told to leave nothing standing. We did what we were told, regardless of whether they were civilians. They were the enemy. Period. Kill. If you don't follow a direct order you can be shot yourself. Now, what am I supposed to do? You're damned if you do and you're damned if you don't. You didn't have to look for people to kill, they were just there. I cut their throats, cut off their hands, cut out their tongue, their hair, scalped them. I did it. A lot of people were doing it and I just followed. I just lost all sense of direction. I just started killing any kinda way I can kill. It just came. I didn't know I had it in me. After I killed the child my whole mind just went. It just went. And after you start it's very easy to keep going on. The hardest is to kill the first time but once you kill, then it becomes easier to kill the next person and the next one and the next one. Because I had no feelings, no emotions. Nothing. 58 Perpetrators get used to the violence they commit and are gradually desensitized. People also learn by doing, just as much in this context as in everyday life, as Staub explained. They progress through a continuum of destructiveness in which greater harm doing is not only possible but probable. 59 The violence becomes a matter of routine. An illustrative quote from a South American torturer: I can only say that when you first start doing the job, it is hard :::. You hide yourself and cry, nobody can see you. Later on, you do not cry, you only feel sad. You feel a knot in your throat but you can hold back the tears. And after ::: not wanting to ::: but wanting to, you start getting used to it. Yes, definitely there comes a moment when you feel nothing about what you are doing. 60 Bureaucrats involved in genocide use similar mechanisms to absolve themselves of blame. They substitute moral responsibility for a technical and organizational responsibility, in which they focus on 'doing their duty' and the task at hand within the organization, rather than on the ultimate outcome of the bureaucratic operation. Indeed, bureaucratic systems may positively help actors to divorce themselves from the consequences of their conduct. Bureaucracy is a modern, rational and impersonal form of organization characterized by a clear functional division of labour: the work is compartmentalized and fragmented. 61 Accordingly, bureaucrats tend to feel that they are small cogs in a big machine, without responsibility for what the machine does or any power to stop it. Such a bureaucracy can turn genocide into a technique, 62 or an administrative process, as reflected by Hoess's comment about Auschwitz: I must admit that this gassing set my mind at rest, for the mass extermination of the Jews was to start soon ::: Now we had the gas, and we had established a procedure. I always shuddered at the prospect of carrying out exterminations by shooting, when I thought of the vast numbers concerned, and of the women and children ::: I was therefore relieved to think that we were to be spared all these blood-baths, and that the victims too would be spared suffering until their last moment came. 63 It is evident from some of these accounts that people can go a long way in their own moral justifications, even to the extent of creating a separate, distorted reality. 64 One extreme example: I made the effort, and it was possible for me, to shoot only children. It so happened that the mothers led the children by the hand. My neighbour then shot the mother and I shot the child that belonged to her, because I reasoned with myself that after all without its mother the child could not live any longer. 65 The more apparently abhorrent the conduct, the greater the need to find a way to justify or rationalize it. Thus, the murder of children might be viewed as an act of mercy. Perpetrators may be so desperate to find excuses to soothe their consciences that Goleman termed such self-deceptions as 'vital lies'. 66 Ultimately, perpetrators may become so accustomed and desensitized to violence that they cease to experience any emotions at all in this context. Thus, a medical doctor involved in selections at Auschwitz: When you see a selection for the first time ç I'm not talking only about myself. I'm talking about the most hardened SS people ::: you ::: how children and women are selected. Then you are so shocked ::: that it just cannot be described. And after a few weeks one can be accustomed to it. And that cannot be explained to anybody. 67 And one of the soldiers involved in My Lai: All these examples show that many of the perpetrators who act on orders feel compelled to follow these orders and while they might ç after their first 'kill' realize the abhorrent nature of their crimes, the instinctive reaction of most will be to rationalize and justify their own behaviour. They in fact learn to come and see this behaviour as nothing out of the ordinary, as something which is necessary and legitimate and they gradually get used to it. It is not only their environment which leads them to this belief but also their own inner psychological forces. They lose their capacity to see their behaviour in the same way outsiders would see it.

How Do Perpetrators Look Back?
The way perpetrators look back is diverse. Some will simply always deny their involvement, while others will point towards their superiors and the orders they received or try to convince anyone who cares to listen, that they did the right thing. Some might feel that they are heroes, or feel proud of their role. 69 By contrast, some perpetrators may start to realize what they have done. Such 'remorseful' perpetrators are likely to suffer from psychological problems including nervous breakdowns, nightmares, depression and post-traumatic stress disorder. Their testimonies show that they did not judge their behaviour at the time of commission the same way as they would now in hindsight: The things you did, you think back and say, 'I can't believe I did that.'At the time, it seemed right. But now you know what you did was wrong. The killing gets to you. The nightmares get to you. You just can't escape it. You can't escape the past. 70 There are many examples like the one above which show how some perpetrators struggle. Perpetrator trauma has been much neglected in studies on perpetrators; this is a pity since we can learn a lot from it. 71 Such research shows that perpetrators at the time when given the order believed that the right thing to do was to follow this order. It shows that many ç at the time ç believed that the orders were legitimate. And it shows that taken out of this particular context some perpetrators realize the true nature of these orders: realize they were actually getting orders to commit horrendous crimes. We prefer not to see and acknowledge this. As Mohamed concluded, '[t]o admit and acknowledge that perpetrators of atrocious crimes experience those crimes as trauma is to admit and acknowledge their status as people rather than as monsters.' 72 We prefer to be able to see perpetrators as 'bad people' very different from us and we prefer not to believe their (in our eyes) petty excuses that they believed they were doing the right thing. We prefer to believe that we are ordinary people who, in their place, would instantly recognize the manifest illegality of what they did, and would refrain from doing it. The academic literature, however, suggests differently.

Conclusion and Implications: Lessons to Be Learned
One of the most important lessons we can learn from the research described in this article is that, within the very particular circumstances in which international crimes are usually committed, many low-level perpetrators do not necessarily recognize the manifest illegality of the orders they receive. Quite the contrary, they may perceive orders to commit genocide and crimes against humanity as necessary or legitimate. Lauterpacht has defined manifest illegality as 'obvious to a person of ordinary understanding'. 73 The point here is that the perpetrators' sense of the legitimacy of such orders (and their own conduct) has become distorted. This is a result of the deliberate creation of a social context which facilitates and promotes crimes, a social context in which they, as good people, fight bad and evil people who threaten their country. They believe they have to use force and violence in order to defend themselves. Furthermore, they are trained to trust their superiors and the government to assess what is and what is not legitimate. They are often shocked by their first crimes but then quickly learn to rationalize and justify their own behaviour. They do this in order to sooth their conscience. They do so in order to psychologically survive. Hence, they come to sincerely believe their superiors in qualifying their behaviour (read: their crimes) as necessary and legitimate within the given circumstances. In short, the literature suggests that perpetrators often no longer recognize the manifest unlawfulness of such orders.
The law acknowledges that orders to commit war crimes are not always manifestly illegal in Article 33(1) but does not do so in relation to crimes against humanity and genocide. The consequence is that the law might punish someone who is not fully aware of the manifest unlawfulness of the order that is given to him. One might wonder if in those cases one of the core principles of (international) criminal law is still met. This principle is that only those with a 'criminal mind', i.e. only those who can be blamed for their crimes, who know that what they are doing is wrong will be prosecuted and punished.
The law is meant to set standards to certain behaviour and to punish those who violate these standards and by doing so ideally prevents or at least limits future violations. I, however, doubt whether Article 33(2) is effective in changing the behaviour of foot soldiers since they often do not recognize the manifest illegality of the orders they get. It would make much more sense to punish not only those who order and instigate their subordinates to commit crimes, but also those who abuse their positions of authority and mislead low-ranking soldiers into believing their unlawful orders are in fact necessary and lawful.