Gavin Dingwall - Journal of International and Comparative Law

[(2016) 3:2 JICL 375–409]

CIRCUMSTANCE, CHOICE AND THE DENIAL OF A SUPERIOR ORDERS DEFENCE IN INTERNATIONAL AND

COMPARATIVE CRIMINAL JUSTICE

Gavin Dingwall*

Abstract: Claims that an actor who followed orders issued by a superior should be granted a defence to a criminal offence are largely rejected in international and domestic criminal law. Various justifi cations have been offered: such a defence would often excuse participants in the gravest forms of criminal activity; a degree of choice remains and so some culpability is present; and deterrence would be compromised if the defence could be invoked too readily. This article will assess these claims with particular regard to the notions of “circumstantial luck” and choice. It will be argued that the rejection of a superior order defence is justifi able, even though some of the orthodox rationales for rejection appear weak. Instead, the relevance (if any) of a superior order claim should be considered at the sentencing stage as part of an overall assessment of individual culpability.

Keywords: superior orders; comparative approaches; international criminal justice; defences; mitigation

I. Introduction

Criminal law scholarship has long debated the availability of defences to a criminal charge.1 Recently there has been a resurgence of interest fuelled in part by the modifi cation of existing defences2 and the introduction of new defences at a domestic

* Professor of Criminal Justice Policy at De Montfort University, Leicester, UK and Visiting Professor of Law, University of Modena and Reggio Emilia, Italy. I benefi ted from discussing this article with my colleagues Tim Hillier and Omar Madhloom and wish to record my thanks for their input.

1 Classic works would include Joshua Dressler, “Refl ections on Excusing Wrongdoers: Moral Theory, New Excuses and the Model Penal Code” (1987) 19 Rutgers Law Journal 671; Kent Greenawalt, “The Perplexing Borders of Justifi cation and Excuse” (1984) 84(8) Columbia Law Review 1897; Paul H Robinson, Jane Grall and Myron Moskovitz, Criminal Law Defenses (St Paul, MN: West, 1984). More recent contributions include Anthony Duff, Answering for Crime: Responsibility and Liability in the Criminal Law (London: Bloomsbury Publishing, 2007); Claire Finkelstein, “Excuses and Dispositions in Criminal Law” (2002) 6(1) Buffalo Criminal Law Review 317; John Gardner, “The Gist of Excuses” (1998) 1(2) Buffalo Criminal Law Review 575; Alan Reed and Michael Bohlander with Nicola Wake and Emma Smith (eds), General Defences in Criminal Law: Domestic and Comparative Perspectives (Farnham: Ashgate, 2014); Peter Westen, “An Attitudinal Theory of Excuse” (2006) 25(3) Law and Philosophy 289; William Wilson, “The Structure of Criminal Defences” [2005] Crim LR 108.

2 See Oliver Quick and Celia Well, “Partial Reform of Partial Defences: Developments in England and Wales” 45(3) Australian & New Zealand Journal of Criminology 337 for an overview. Diminished responsibility serves as an example of a modifi ed existing defence, see s.52 of the Coroners and Justice

JICL-3(2).indb 375JICL-3(2).indb 375 01/11/16 14:5701/11/16 14:57

376 Journal of International and Comparative Law

level.3 It is right that defences are subjected to this level of scrutiny; someone who ostensibly should incur criminal liability, and the associated penalty, should not readily be excused as that would appear to defeat the purpose of criminalisation and, in a practical sense, would lead to inconsistency of application, which raises questions of legitimacy. The impact of a defence can be enormous: criminal liability can be removed in many cases and reduced in others. Often there is no doubt that a defence should exist. Debate centres on the exact circumstances where the defence should apply. An example is provided by self-defence. The need for individuals to have the legal right to protect themselves through the use of force appears obvious,4 but the situations where force should be legitimate and the degree of force that can be justifi ed are areas of dispute.

Superior orders, the subject matter of this article, needs consideration at a more basic level because there is no general acceptance that such a defence exists or is desirable; instead, there is considerable inconsistency at a comparative level. The debate then is more about whether a defence in any form should be available and less about the ambit and scope of such a defence. This makes the argument presented in this article more complicated because reliance on the existing law will only get us so far. Reasons may be given as to why a superior orders defence is necessary or why it should be refused, though the courts frequently arrive at a conclusion without articulating their reasoning, but a ledger of where a defence exists and where it does not fails to provide an answer as to whether and why a defence should or should not be allowed. Any analysis needs to be more philosophical and must address some fundamental, and often complex and diffi cult, issues concerning individual culpability. If it is decided that a defence should be available, secondary issues would then need to be considered. An example of such a secondary concern would be whether a defence should be available to all offences or should be limited. However, even if it was decided that following the commands of a superior does affect culpability, this does not mean that a defence is appropriate. Rather than creating (or in some jurisdictions confi rming) a superior orders defence, the issue could be resolved at the sentencing stage by recognising that compliance with an order from a superior should be viewed as personal mitigation. Sentencers would have the discretion to determine what culpability (and in extreme cases it may be minimal) attaches to the inferior’s compliance with the command. The primary distinction between the approaches is that a conviction would still result if superior

Act 2009 and comment by Ronnie D Mackay, “The Coroners and Justice Act 2009 – Partial Defences to Murder: (2) The New Diminished Responsibility Plea” [2010] Crim LR 290.

3 Loss of control provides an English example of what is probably best regarded as a new defence due to the fundamental differences from the existing defence of provocation. See Alan Norrie, “The Coroners and Justice Act 2009: Partial Defences to Murder (1) Loss of Control” [2010] Crim LR.

4 See the chapters in Christian Coons and Michael Weber (eds), The Ethics of Self-Defense (New York: Oxford University Press, 2016); Tyler Doggett, “Recent Work on the Ethics of Self-Defense” (2011) 6(4) Philosophy Compass 220; Helen Frowe, “A Practical Account of Self-Defence” (2010) 29(3) Law and Philosophy 245.


The Denial of a Superior Orders Defence 377

orders were to be treated as mitigation whereas a defence would result in acquittal. Questions then of fair labelling and reputation are also pertinent.5

The paucity of existing jurisprudence is not the only reason why it is necessary to subject the rationales for having a superior orders defence to a sustained theoretical critique. Superior orders as a potential defence differs from most forms of defence. One distinction relates to context: although there are examples regarding orders from a civilian,6 most cases arise in a military context and often during confl ict. Conversely, although many “domestic defences” are recognised in international criminal justice,7 their use would be highly unusual. This must be recognised because it is arguably in the fi eld of international criminal law that the need for careful evaluation of defences becomes most acute. Extraordinary crime has no direct parallel in domestic criminal justice due to the barbarity and scale of the offending. In itself, any system of criminal law would struggle to respond appropriately to harm of this magnitude.8 However, the context makes the attribution of criminal liability diffi cult despite this evident and extreme harm.9 Prosecution will be arbitrary and therefore tokenistic, for example, given the systemic nature of the offending, leading to legitimate concerns that a very limited number of individuals are being scapegoated for the sins of many. The danger of scapegoating is compounded by the command structure inherent in the armed forces. Frequently, individual acts of violence are committed by low-ranking combatants who are following orders from those more senior in the hierarchy. Those who issue such commands can be held criminally liable where appropriate, as can those whose wilful blindness allows subordinates to act in this manner.10 What, though, of the

5 See James Chalmers and Fiona Leverick, “Fair Labelling in Criminal Law” (2008) 71(2) Modern Law Review 217; John J Child, “Prior Fault: Blocking Defences or Constructing Crimes” in Alan Reed and Michael Bohlander with Nicola Wake and Emma Smith (eds), General Defences in Criminal Law (n.1).

6 See R v Trainer (1864) 4 F & F 105, 176 ER 488. 7 See Antonio Cassese, “The Statute of the International Criminal Court: Some Preliminary Refl ections”

(1999) 10(1) European Journal of International Law 144. 8 See Gavin Dingwall and Tim Hillier, “The Banality of Punishment: Context Specifi city and Justifying

Punishment of Extraordinary Crimes” (2010) 6(1) International Journal of Punishment and Sentencing 6; Mark A Drumbl, Atrocity, Punishment, and International Law (Cambridge, NY: Cambridge University Press, 2007): Alexander KA Greenawalt, “International Criminal Law for Retributivists” (2014) 35(4) University of Pennsylvania Journal of International Law 969; David Wippman, “Atrocities, Deterrence, and the Limits of International Justice” (1999–2000) 23 Fordham International Law Journal 473; Anthony Keane Woods, “Moral Judgments and International Crimes: The Disutility of Desert” (2011) 52(3) Virginia Journal of International Law 633.

9 See Gavril Paraschiv, “On the Gravity of the Offence in International Criminal Law” (2013) 5(1) Geopolitics, History and International Relations 136.

10 This article will not consider the potential liability of those who issue illegal commands. See Grady K, “International Crimes in the Courts in England and Wales” [2014] Crim LR 693 and, more generally, Mirjan Damaska, “The Shadow Side of Command Responsibility” (2001) 49(3) The American Journal of Comparative Law 455; Alison Marston Danner and Jenny S Martinez, “Guilty Associations: Joint Criminal Enterprise, Command Responsibility, and the Development of International Criminal Law” (2005) 93(1) California Law Review 75. With regards to genocide, crimes against humanity and war



individual perpetrator who maintains that the only reason he acted in this way was because he was commanded to do so?

Claims that a defence should be granted to those who followed orders have a signifi cant historical pedigree.11 With the development of international criminal justice in the aftermath of the Second World War, a more reasoned response to the age-old “soldier’s dilemma” was required. The fact that authorities are comparatively sparse refl ects the limits of enforcement; despite the involvement of thousands in most wars, few participants have faced justice in international tribunals. Of those who have, only junior combatants could potentially claim that they were acting on a senior offi cer’s command. Neither is such a defence likely to be run with any frequency in domestic courts. In addition the fact that such a scenario is likely to be rare, and one where there is supporting evidence even rarer, the lack of domestic case law can often be explained by a settled position in a criminal code, for example in France, or by the availability of another established and accepted defence. One must, however, guard against measuring the importance or the contestability of a point of law with reference only to the frequency with which it emerges in the courts. Many of the most vexed questions of substantive law, for example the situations where intention can properly be inferred, seldom arise in practice. It remains the case that such questions should not be ignored and deserve a considered response.

In a detailed analysis of the American and English authorities, Brownlee outlines three possible approaches that could be taken with regards to those who seek to excuse their criminal conduct on the basis that they were following superior orders:12

crimes, specifi c statutory provisions apply in England and Wales. The International Criminal Court Act 2001, s.65 provides:

“(2) A military commander, or a person effectively acting as a military commander, is responsible for offences committed by forces under his effective command and control, or (as the case may be) his effective authority and control, as a result of his failure to exercise control properly over such forces where –

(a) he either knew, or owing to the circumstances at the time, should have known that the forces were committing or about to commit such offences, and

(b) he failed to take all necessary and reasonable measures within his power to prevent or repress their commission or to submit the matter to the competent authorities for investigation and prosecution.

(3) With respect to superior and subordinate relationships not described in subsection (2), a superior is responsible for offences committed by subordinates under his effective authority and control, as a result of his failure to exercise control properly over such subordinates where –

(a) he either knew, or consciously disregarded information which clearly indicated, that the subordinates were committing or about to commit such offences,

(b) the offences concerned activities that were within his effective responsibility and control, and

(c) he failed to take all necessary and reasonable measures within his power to prevent or repress their commission or to submit the matter to the competent authorities for investigation and prosecution.

(4) A person responsible under this section for an offence is regarded as aiding, abetting, counselling or procuring the commission of the offence”.

11 See R v Cooke (1660) 5 St Tr 1077; R v Axtell (1661) Kel 1, 13, 84 ER 1055, 1060.12 Ian D Brownlee, “Superior Orders – Time for a New Realism?” [1989] Crim LR 396, 397–398.



(1) Position A. Absolute and automatic justifi cation of the subordinate concerned. The superior who was the author of the orders becomes the principal to any offence on the maxim, respondeat superior.

(2) Position B. Personal liability of the subordinate. This “no defence” position is effected by affi xing the individual soldier with both an unavoidable obligation to evaluate each order that he or she is given and a personal responsibility for ensuring that only lawful orders are obeyed. It is sometimes referred to as the position of “absolute” liability.

(3) Position C. The middle position, which has been called, “the manifest illegality principle”. According to this approach a soldier is entitled to a defence on the basis of superior orders where the order on which the defence is based was not of such a nature that it should have been palpably obvious to the reasonable recipient that it called for an illegal act to be done.

This framework will be adopted throughout this article both as a means of classifying existing approaches in comparative and international law and when considering whether a defence (of whatever type) should be available in appropriate cases. It is necessary to go through several stages to assess the desirability of a defence of superior orders. An important starting point is to scope existing legal practice to see if a consensus has emerged and, regardless of whether that is the case, the factors that have infl uenced the courts. The next three sections consider, in turn, superior orders in English law, superior orders in comparative law and superior orders in international criminal justice. Later sections adopt a more theoretical approach. It will be argued that the nature of a superior orders claim has distinct similarities with the widely-accepted defence of duress and so the fi rst of these sections considers whether there is a difference in terms of culpability where an individual offends after being commanded to do so by a superior as opposed to where they have been threatened with severe violence. Reviewing the arguments for having a defence of duress assists us in deciding whether a superior orders defence should be provided, but other concerns are relevant. The section following provides this broader appraisal of the case for a distinct defence. An alternative option — namely that superior orders should be viewed as a form of mitigation — is considered in the penultimate section.

It will be submitted in the conclusion that, whilst cognisance should be taken of the soldier’s plight and allowance must be made for the consequent reduction in individual culpability, a superior orders defence is not the preferable means to achieve justice. An absolute defence (Brownlee’s Position A) absolves individuals of all responsibility for their actions whereas the “manifest illegality” option (Brownlee’s Position C), whilst initially attractive because it links culpability to the gravity of the harm, also has the potential to be permissive and excuse all but a limited number who commit the most serious offences. Where illegal superior orders have been followed, sentencers should recognise that this reduces, but does not remove, individual liability: some degree of mitigation is appropriate, but the weight that should be attached to it will vary with each case. This conclusion is, in



the author’s opinion, consistent with the existing law in England and Wales and, therefore, in this jurisdiction, amounts to a defence of the status quo.

II. Superior Orders in English Law

A useful starting point in considering the current position in England and Wales is this statement found in the current Manual of Service Law:

“A person who is bound to obey a superior is under a legal duty to refuse to carry out an order received from that superior, to do some act or make some omission, if the order is manifestly illegal. If the illegal order is carried out, an offence may be committed. Where the order is not manifestly illegal, an accused will not be excused if they carried out the order and in doing so commits an offence. However, the accused may have a defence on other grounds because, for example, the order may negate a particular intent on the accused’s part (which may provide a complete defence) or it may reduce the offence to one of a less serious nature or it may excuse what otherwise appears to be negligence. Evidence of superior orders which fall short of providing the accused with a defence to the offence may still be a strong mitigating factor.”13

The Manual is not a source of law14 and this unequivocal position masks the fact that the available law is “sparse and in confl ict”.15 An exposition of the law for military personnel must state the law with clarity and a lengthy exposition of contentious areas of legal debate would detract from that clarity. The earliest authorities, Cooke16 and Axtell,17 would support the position that superior orders could not form the basis of a defence in English law. Both individuals had participated in the execution of Charles I Cooke, in his capacity as Chief Justice of Ireland, had presented the indictment whilst Axtell had ordered a subordinate to carry out the execution. The courts refused to accept that their compliance with lawful orders should give rise to a defence and both were convicted of treason. Similarly, in Thomas,18 a defence was denied to a naval sentinel who, in contravention of an order, shot and killed someone on board a boat even though it was accepted that

13 Ministry of Defence, The Manual of Service Law (London: Ministry of Defence, Vol 1, 2013) section 2, para.1-12-7.

14 Adam Roberts and Richard Guelff, Documents on the Laws of War (Oxford: Oxford University Press, 3rd ed., 2000) p.13.

15 Dennis J Baker, Glanville Williams Textbook of Criminal Law (London: Sweet and Maxwell, 4th ed., 2015) p.842.

16 Cooke (n.11).17 Axtell (n.11).18 (1851) 105 ER 89.



Thomas had fi red with the mistaken belief that it was his duty to do so. According to Brownlee’s classifi cation, these rulings adopt Position B: namely, a position of absolute liability.19

Two nineteenth-century judgments, though, demonstrated some willingness to allow a defence, although neither formulation of a potential defence would have succeeded on the facts of the cases just discussed. R v James20 is arguably the more interesting authority despite the defence ultimately being denied. The grounds for rejection suggest that in other circumstances, a superior orders defence could be advanced successfully. What was determinative in this case was that Lord Abinger CB refused the defence on the basis that murder was malum in se, hence a servant would know that an order which would lead to death should not be followed. The employer in this case had allegedly ordered employees to block the airway of a competitor’s mine; regardless of the veracity of this account, James could not avail himself of a defence because he must have known that the order should have been disobeyed. What we would appear to have is the emergence of a qualifi ed defence that is either offence-specifi c, murder, for example, falling outside the ambit of the defence, or alternatively a defence dependent on an objective appreciation of whether the order should be obeyed based on the seriousness of the harm. A couple of points are worth developing. First, the approach that is adopted for determining the scope of the defence matters. Whilst certain offences are evidently malum in se, this is not always the case, with the result that classifi cation can become arbitrary and subjective on the margins. Second, as James was refused the defence, one can merely speculate when the defence would be available. However, as the case envisages a defence for some offences based on a conception of offence severity, the approach appears to equate to Brownlee’s Position C, which allows for a defence except where the act is manifestly illegal. Admittedly, conduct which is malum in se and behaviour which is manifestly illegal are not wholly synonymous. The latter imparts a gravity which is not necessarily obvious in the former. Yet, by stating that, because murder is malum in se, a servant would have known not to follow the order, Lord Abinger is in effect relying on the manifest illegality of the crime to infer that a servant would know not to comply with the order.

19 “Absolute liability”, of course relates solely to the operation of a superior orders defence. It does not remove or restrict the possibility of the defendant being able to rely upon another defence, perhaps most commonly duress (see Section V). This is recognised in the Ministry of Defence, Manual of Service Law (n.13):

“However, the accused may have a defence on other grounds because, for example, the order may negate a particular intent on the accused’s part (which may provide a complete defence) or it may reduce the offence to one of a less serious nature or it may excuse what otherwise appears to be negligence.”

Strictly speaking, this statement is not without problems because it equates absence of mens rea with an affi rmative defence. A more accurate claim would be that the defendant would be entitled to an acquittal or conviction on a lesser charge as the instant offence is not made out. Yet, even if the offence is made out, a defence may be available.

20 (1837) 8 Car & P 131, 173 ER 429.



Trainer21 categorically supported this approach. A train driver and a fi reman were charged with manslaughter after a failure to stop at a signal resulted in the loss of life. The fi reman raised a superior orders defence on the basis that company rules stipulated that he had to follow a driver’s directions and the driver had chosen to ignore the signal. Willes J stated that an inferior should have a defence if he was obeying orders from a superior unless they were obviously against the law. “Obvious” and “manifest” illegality are not identical — manifest may suggest a higher threshold — but here there is no doubt that a conditional defence, consistent with Position C in Brownlee’s framework, exists. At this juncture, English law appeared to have moved away from absolute liability.

More recent authorities, however, suggest a return to the former position with the complete denial of a defence. In Abbot v R,22 a Privy Council decision, Lord Salmon maintained that no reliance on superior orders could be made, at least with regards to extraordinary crimes committed at a time of war:

“In the trials of those responsible for wartime atrocities such as mass killings of men, women or children, inhuman experiments on human beings, often resulting in death, and like crimes, it was invariably argued for the defence that these atrocities should be excused on the ground that they resulted from superior orders and duress: if the accused had refused to do these dreadful things, they would have been shot and therefore they should be acquitted and allowed to go free. This argument has always been universally rejected. Their Lordships would be sorry indeed to see it accepted by the common law of England.”23

Perhaps one should not to give this statement too wide an interpretation. Not only could this be seen as consistent with Position B — no defence or absolute liability — but, in light of the types of crime mentioned, it would appear equally consistent with Position C. The crimes in question are obviously malum in se and the orders manifestly illegal. Lord Salmon draws a useful parallel with the defence of duress, to which this article will return later, but if there is validity in this comparison, it would in fact lend support to Position C. Assuming a number of requirements are met, duress is an available defence in English law unless the offence is murder or attempted murder. As we shall see though, this restriction is justifi ed on narrower grounds relating not to a distinction between malum in se and mala prohibita or the manifest illegality of the act (both of which relate to broad conceptions of harm) but to the sanctity of human life (a particular and distinct value which, if violated, results in a unique harm).

21 Trainer (n.6).22 [1977] AC 755.23 Ibid., [469].



The decision in R v Clegg confi rmed that English law has no defence of superior orders.24 The defendant was stationed with the British army in Northern Ireland. One night whilst on patrol, a colleague stopped a vehicle at a checkpoint. The car, which was stolen, accelerated away down the middle of the road with its headlights on full beam towards Clegg and three other soldiers. Someone on the checkpoint shouted at them to stop the vehicle and so they opened fi re killing the driver and a passenger. Clegg was charged with murdering the passenger who, crucially, had been shot in the back once the car had passed. The case was primarily about the use of excessive force by a soldier on duty and whether a plea of self-defence was capable of reducing culpable homicide from murder to manslaughter in this situation. The argument failed on the basis that the House of Lords felt that it was a matter for Parliament to determine.25 Superior orders did not arise directly, but Lord Lloyd of Berwick noted the following:

“It should be noticed that the point at issue here is not whether Pte Clegg was entitled to be acquitted altogether on the ground that he was acting in obedience to superior orders. There is no such general defence known to English law, nor was any such defence raised at the trial …. The point is rather whether the offence in such a case should, because of the strong mitigating circumstances, be regarded as manslaughter rather than murder. But so to hold would … be to make entirely new law. I regret that under existing law, on the facts found by the trial judge, he had no alternative but to convict of murder.”26

III. Superior Orders in Comparative Law

As English law has fl uctuated between an absolute prohibition on a defence and a suggestion that a defence exists where the illegal activity is not malum in se or clearly unlawful, there is virtue in reviewing the jurisprudence from a limited number of common law and civil law jurisdictions: primarily Australia, France, Israel and the United States. What emerges is an inconsistency of approach, ranging from a complete denial of a superior order defence to a willingness to allow a defence except where manifest illegality is concerned. Practice tends then to alternate between Brownlee’s Position B and Position C. Particularly in civil jurisdictions where reliance is made on a criminal code, the position in most jurisdictions is settled. Clarity though has the potential to disguise complexity: the underlying

24 [1995] 1 AC 482.25 Ibid., 500.26 Ibid., 498.



justifi cations for having a superior orders defence are every bit as contestable in a jurisdiction with a criminal code. Even if the vast majority of countries adopted an identical stance in their codes, and there was no case law to challenge the position taken, that cannot be taken as suffi cient justifi cation for allowing, or refusing, a superior orders defence. That task requires a theoretical assessment which will be found in subsequent sections. However, reviewing how diverse countries have responded to superior order claims provides valuable material that can inform this exercise. The jurisdictions have been grouped here according to Brownlee’s classifi cations.

A. Position B: absolute liability and the refusal of a defenceAustralian law refuses to recognise a defence of superior orders. The English courts27 and the Privy Council28 have both cited, with approval, the ruling in A v Hayden (No 2):29

“It is fundamental to our legal system that the executive has no power to authorize a breach of the law and that it is no excuse for an offender to say that he acted under the authority of a superior order.”30

As Mason J comments in the judgment, there is “an air of unreality” about the facts of the case such that it “has the appearance of a law school moot based on an episode taken from the adventures of Maxwell Smart”.31 The Commonwealth sought to recruit counter-espionage offi cers and assured the individuals that their identities would remain protected for personal and national security reasons. They were then instructed to take part in a training exercise in the Sheraton Hotel in Melbourne which involved the risk of offences being committed. Subsequently, the security forces argued that they were entitled to disclose the identities to the Chief Commissioner of Police for Victoria so that he could investigate whether the individuals had committed offences with a view to prosecuting anyone who had engaged in illegal activity.32 Superior orders, it was concluded, “are not and never have been a defence in our law — though it is conceivable that the plaintiffs may have had some such belief”.33 As the next section will show, some other jurisdictions, notably the United States, have moved away from this absolute prohibition.

27 Ibid.28 Yip Chiu Cheung v R [1995] 1 AC 111, 114.29 (1984) 156 CLR 532.30 Ibid., 540.31 Ibid., 550.32 Ibid.33 Ibid.



B. Position C: a defence unless the order is manifestly unlawfulMost civil law jurisdictions adopt Brownlee’s Position C.34 Articles 122–124 of the French Criminal Code provides an example: “A person who carries out an act ordered by a legitimate authority is not criminally liable, except if this act is obviously illegal.” Elliott explains how the defence operates:

“Where an order was obviously illegal, people who carried it out have no defence on the grounds of superior orders, though they may be able to argue they acted under a constraint. A police offi cer who carries out an arrest under a mandate issued by an investigating judge, which is subsequently declared to be illegal, would have a defence if the illegality was not obvious. But if the police offi cer knew the order was illegal then no defence would be available. The more serious the conduct the more likely a court will conclude it was obviously illegal. Article 213-4 of the Criminal Code expressly states that a person who commits a crime against humanity cannot rely on this defence, such an argument only being available as a cause of mitigation when the court is determining what sentence to impose. If it can be shown that the subordinate knew the order was illegal, even though this illegality was not obvious, then the courts are not likely to allow this defence.”35

These comments are illuminating. The use of the phrase “obviously illegal” is potentially misleading as Elliott makes it clear that a defence will not be available where there was subjective knowledge that the order was unlawful even if the illegality was not objectively obvious. This suggests a more complex defence than at fi rst appears. It would appear to be unavailable where (1) the act is obviously illegal or (2) the act is not obviously illegal but the defendant knows that it is unlawful. Scenario (1), which is what is described in arts.122–124 of the Criminal Code, demands an objective assessment of the quality of the order. Elliott indicates that the seriousness of the conduct will inform this judgment. More interesting is scenario (2) because it does not refl ect the wording of the Code. Here subjective awareness of unlawfulness, regardless of its objective severity, removes the possibility of a defence. Put another way, a defence is only available where the act is not obviously illegal and where the defendant does not realise that the conduct was unlawful. The Code suggests a wider defence than operates in practice. A second point which is worth making briefl y in the context of this article is that, even with regards to crimes against humanity, superior orders may be considered

34 For other examples, see the Dutch Criminal Code, art.43 and the Italian Criminal Code, art.51, para.4.35 Catherine Elliott, “France” in Reed and Bohlander with Wake and Smith (eds), General Defences in

Criminal Law (n.1) 225.



as mitigation. This refl ects a view that culpability is to some extent reduced even though there is a reluctance to allow a defence.

Early American case law was clear that superior orders afforded no defence.36 The earliest authority, Little v Barreme,37 concerned a military instruction issued by President John Adams during the United States–French war authorising the seizure of vessels travelling to or from French ports. The instruction though was wider than the Act of Congress which only permitted the seizure and forfeiture of vessels travelling to French ports. George Little, a Navy Captain, seized a Danish ship which had left a French port and escorted it to Boston. The ship’s owners sued for trespass in a maritime court. Chief Justice Marshall expressed sympathy for Little’s position:

“I confess the fi rst bias of my mind was very strong in favor of the opinion that though the instructions of the executive could not give a right, they might yet excuse from damages. I was much inclined to think that a distinction ought to be taken between acts of civil and those of military offi cers, and between proceedings within the body of the country and those of the high seas. That implicit obedience which military men usually pay to the orders of their superiors, which indeed is indispensably necessary to every military system, appeared to me strongly to imply the principle that those orders, if not to perform a prohibited act, ought to justify the person whose general duty it is to obey them, and who is placed by the laws of his country in a situation which in general requires that he should obey them. I was strongly inclined to think that where, in consequence of orders from the legitimate authority, a vessel is seized with pure intention, the claim of the injured party for damages would be against that government from which the orders proceeded, and would have been a proper subject for negotiation. But I have been convinced that I was mistaken, and I have receded from this fi rst opinion. I acquiesce in that of my brethren, which is that the instructions cannot change the nature of the transaction or legalize an act which without those instructions would have been a plain trespass.”38

This ruling was followed in United States v Bright39 and in Mitchell v Harmony.40 Having an absolute bar on the defence meant that service personnel could fi nd

36 This section on the United States position draws considerably on the sources cited in James B Insco, “Defense of Superior Orders before Military Commissions” (2003) 13 Duke Journal of Comparative & International Law 389; Gary D Solis, “Obedience of Orders and the Law of War: Judicial Application in American Forums” (1999) 15(2) American University International Law Review 481.

37 6 US 170 (1804).38 Ibid., 179.39 24 F Cas 1232 (CCD Pa 1809).40 54 US 115 (1851). Although superior orders were rejected as a defence in this case, it was the earliest

American authority to state that it was a relevant factor when determining sentence.



themselves in a horrendous position: not only were individuals strictly liable for acts committed in compliance with an unlawful command, but they were also strictly liable if they failed to comply. This was illustrated graphically in Wilkes v Dinsman,41 where the Supreme Court upheld the legality of disciplinary action (including fl ogging and the use of irons) against a marine who challenged the lawfulness of his orders.

American law began to change after the decision in McCall v McDowell.42 The defendant was sued for false imprisonment having imprisoned the plaintiff in an attempt to stop rioting following the assassination of President Lincoln. Assuming that the order is not “palpably atrocious as well as illegal”,43 a defence would be granted. This represents the fi rst authority in American law to advance Brownlee’s Position C.44 James Insco documents a brief later shift in the law, however, this was not the abandonment of a “manifest illegality” position in a reversion to the absolute liability position (Brownlee’s Position B), but the expansion of the defence to cover all orders (a rare example of Brownlee’s Position A).45 This protected American service personnel during the World Wars but equally afforded a defence to enemy combatants who could claim that they had acted on the instructions of their superiors. The impediment that such a defence posed to punishing Axis soldiers after the Second World War was the primary motivator to restrict the defence and, once again, disallow it in cases where there was manifest illegality.46 The Korean and Vietnam Wars demonstrate that American personnel who were involved in gross abuse were afforded no defence.

The Korean confl ict gave rise to the notorious case United States v Kinder.47 Kinder captured a Korean intruder whilst on sentry duty. The intruder was passed to a colleague who attacked him until he lost consciousness. Kinder was then instructed by the commanding offi cer to shoot the individual, which he did. He was subsequently convicted of premeditated murder and conspiracy to murder and, on appeal, it was submitted on his behalf that complying with superior orders provided a defence regardless of whether the order was manifestly illegal. It was a simple argument: namely that the law provided an absolute defence (Brownlee’s Position A) as opposed to the earlier authorities which suggested that the defence did not apply when manifest illegality was present (Brownlee’s Position C). Kinder’s appeal was rejected by the Air Force Board of Review due to the “manifest and unmistakable illegality of the order”.48

41 48 US 89 (1849).42 15 F Cas 1235 (CCD Cal 1867).43 Ibid., 1241.44 See also the early authorities of Riggs v State 91 Am Dec 272 (Tenn 1866) and State v Sparks 27 Tex 627

(1864).45 See Insco, “Defense of Superior Orders before Military Commissions” (n.36) pp.403–404. 46 Ibid., pp.404–405.47 (1953) 14 CMR 742.48 Ibid., 774. The commanding offi cer was also convicted of premeditated murder: United States v Schreiber

(1955) 5 CMA 602.



Similar facts arose in the Vietnam War case United States v Keenan,49 which concerned the assassination of an elderly civilian. The defence of superior orders failed as the order was one which a man of “ordinary sense and understanding” would know was illegal. United States v Calley50 attracted considerable public attention and led to a sense of injustice which is diffi cult to comprehend half a century after the My Lai Massacre.51 Calley, who admitted his involvement, argued that he was merely complying with superior orders. Unsurprisingly, given the established legal position by this time, the defence failed.52 Such was the public outrage that President Nixon ordered his release the day after he was sentenced to life imprisonment.53

The current formulation in the United States is that “a subordinate must obey all orders except those that are palpably illegal”.54 Palpable and manifest illegality would presumably be interpreted identically. What is useful about this formulation is that it highlights the general duty that service personnel have to comply with orders. The position would appear to differ from France in so far as a defence would be available to someone who follows an illegal command knowing it to be unlawful, provided it is not palpably unlawful. The desirability of protecting individuals who obey orders outweighs other concerns until the illegality is such that it is appropriate to prosecute and punish. For Brownlee, the United States is an exemplar of Position C.55 Following the decision in R v Finta, this also represents the position in Canadian law.56

Israeli law has also moved from a denial of the defence to a variant of the “manifestly illegal” approach. The trial of Adolf Eichmann in the District Court of Jerusalem for crimes against the Jewish people, crimes against humanity and war crimes centred on the availability of a superior orders defence.57 The District Court accepted that he was following orders but rejected the defence on the basis that he was “not a puppet in the hands of others”; rather he was “amongst those who pulled the strings”.58 The Supreme Court also stated that there was no defence of superior orders in Israeli law, but questioned whether someone so senior in the Nazi

49 (1969) 18 CMA 108.50 48 CMR 19 (CMA 1973).51 For historical accounts, see William Thomas Allison, My Lai: An American Atrocity in the Vietnam War

(Baltimore, MD: John Hopkins University Press, 2012); Michael Bilton and Kevin Sim, Four Hours in My Lai: A War Crime and Its Aftermath (London: Penguin Books, 2003); Kendrick Oliver, The My Lai Massacre in American History and Memory (Manchester: Manchester University Press, 2006).

52 Calley (n.50), 27. 53 See Insco, “Defense of Superior Orders before Military Commissions” (n.36) p.406.54 United States v Pacheco 56 MJ 1, 7 (CAAF 2001).55 See Brownlee, “Superior Orders – Time for a New Realism” (n.12) 407.56 [1994] 1 SCR 701. For the earlier Canadian position, see Geoffrey Creighton, “Superior Orders and

Command Responsibility in Canadian Criminal Law” (1980) 38(1) University of Toronto Faculty of Law Review 1; James M Simpson, “The Defence of Superior Orders in Canada: A Review of Superior Orders in National and International Law” (1977) 15 Canadian Yearbook of International Law 306.

57 A-G of the Government of Israel v Eichmann 36 ILR 18 (Israeli District Court, 1961).58 Ibid.



hierarchy could actually be said to have been following orders in any event.59 With regards to their own armed forces, Israeli law now provides a defence for those obeying superiors unless the order was “blatantly illegal”. One supposes that there is no material difference between manifest illegality and blatant illegality.

IV. Superior Orders in International Law

Often it is after confl ict that defendants seek to argue that they were acting under superior orders and the forum in which they are facing trial is not a domestic court but an international tribunal. This has meant that international lawyers have also debated the virtue of allowing a superior orders defence.60 As with domestic law, the legal position has not been settled although there has been a trajectory from an absolute denial of a defence (Brownlee’s Position B) to a “manifest illegality” approach (Brownlee’s Position C) akin to the jurisdictions in the previous section.61 This section is subdivided into two parts and will consider the early denial of a defence and then the more recent shift to a defence save where “manifest illegality” is present.

A. Th e denial of a superior orders defence in international lawIt will be recalled that during both World Wars, American service personnel operated with the impunity of a total defence of superior orders (Brownlee’s Position A).62 This was inconsistent with both earlier American law and the domestic law in other jurisdictions. After the conclusion of the World War II, the Allies produced the London Agreement establishing the terms of the Nuremberg Tribunal. Article 8 provided the following with regard to superior orders:

“The fact that the defendant acted pursuant to order of his Government or of a superior shall not free him from responsibility, but may be considered in mitigation of punishment if the Tribunal determines that justice so requires.”

Defence counsel in one case advanced an argument, based on the position found in the American military manual, that there was an absolute superior orders defence

59 A-G of the Government of Israel v Eichmann 36 ILR 277 (Supreme Court of Israel, 1962).60 Recent analyses include Hiromi Sato, “The Defense of Superior Orders in International Law: Some

Implications for the Codifi cation of International Criminal Law” (2009) 9(1) International Criminal Law Review 117; Bohrer Ziv, “The Superior Orders Defense: A Principal-Agent Analysis” (2012) 41(1) Georgia Journal of International and Comparative Law 1.

61 For a more detailed account, see Hilaire McCoubrey, “From Nuremberg to Rome: Restoring the Defence of Superior Orders” (2001) 50(2) The International and Comparative Law Quarterly 386.

62 See Insco, “Defense of Superior Orders before Military Commissions” (n.36) pp.403–404.



at the time of the war, but this was rejected outright.63 However, the initial opening speeches from the prosecutors might be taken to show that some awareness should be made of the situation in which combatants found themselves. Take this extract from the American prosecutor’s opening submission:

“[We] do not argue that the circumstances under which one commits an act should be disregarded in judging its legal effect. A conscripted private on a fi ring squad cannot expect to hold an inquest on the validity of the execution. The Charter implies common sense limits to liability just as it places common sense limits upon immunity.”64

The British prosecutor’s opening statements similarly raise questions about the scope of the defence in customary international law. By using the phrase “manifestly contrary”, is there an implication that Brownlee’s Position C represents the position rather than Brownlee’s Position B? It would appear that the discussion does not relate specifi cally to art.8 of the London Agreement but rather to the customary international law position at the time: “[There] is no rule of international law which provides immunity for those who obey orders which … are manifestly contrary to the very law of nature from which international law has grown”.65

What is evident from the preceding sections is that there was no uniformity in domestic approaches at the time of the Nuremberg trial. Moreover, different military manuals appeared to offer diverse accounts of the law for service personnel. Against this background it would be diffi cult to fi nd agreement on a customary rule of international law. This issue would resurface when the international community drafted the Rome Statute establishing the International Criminal Court, which will be considered in the next section. However, there was no doubt regarding the London Agreement — no superior order defence was available regardless of the position in customary international law. By the end of the Nuremberg trial, the prosecutors no longer expressed sympathy for the plight of the defendants. Arguably, the most famous extract from the summing up reads as follows:

“These men … destroyed free government in Germany and now plead to be excused from the responsibility because they became slaves. They are

63 Adam Roberts and Richard Guelff, Documents on the Laws of War (Oxford: Oxford University Press, 3rd ed., 2000) p.13. The authors also take this as evidence that:

“[although] national military manuals have an important function in providing evidence of the law, in certain instances they may need to be viewed with some caution …. The Tribunal did not consider the statements as conclusive on this point”.

64 Trial of the Major War Criminals before the International Military Tribunal (Nuremberg Trial), Opening Speech for the Prosecution, Vol 2, pp.150–151.

65 Ibid., Vol 19, pp.465–466.



in the position of the fi ctional boy who murdered his father and mother and then pleaded for leniency because he was an orphan.”66

The judgment also picked up on the lack of a defence and, of particular interest, provided a rationale for denial:

“[The] provisions of [art.8] are in conformity with the law of all nations. That a soldier was ordered to kill or torture in violation of international law of war has never been recognized as a defence to such acts of brutality, though, as the Charter here provides, the order may be urged in mitigation of punishment. The true test, which is found in varying degrees in the criminal law of most nations is not the existence of the order, but whether moral choice was in fact possible.”67

Article 6 of the 1946 Charter of the International Military Tribunal for the Far East took a similar position:

“Neither the offi cial position, at any time, of an accused, nor the fact that an accused acted pursuant to order of his government or of a superior shall, of itself, be suffi cient to free such accused from responsibility for any crime with which he is charged, but such circumstances may be considered in mitigation of punishment if the Tribunal determines that justice so requires.”

There is a possible argument that this is slightly more permissive than the London Agreement as it does not completely rule out a defence, rather it provides that it cannot provide a defence in itself. Superior orders could perhaps then have bolstered another defence, however, there is no authority on the issue. Nonetheless, the overall stance is clear. The post-World War II tribunals adopted an absolute approach (Brownlee’s Position B). The position seemed settled and was enshrined in the Statute of the International Criminal Tribunal for the former Yugoslavia68 and the Statute of the International Criminal Tribunal for Rwanda.69 The potential impact

66 Cited in Geoffrey Best, War & Law since 1945 (Oxford: Clarendon Press, 2002) p.190.67 Trial of the Major War Criminals before the International Military Tribunal (1948), Vol 12, pp.411, 466.68 UN Security Council, Statute of the International Criminal Tribunal for the Former Yugoslavia, art.7,

para.4 states that:

“The fact that an accused person acted pursuant to an order of a Government or of a superior shall not relieve him of criminal responsibility, but may be considered in mitigation of punishment if the International Tribunal determines that justice so requires.”

69 UN Security Council, Statute of the International Criminal Tribunal for Rwanda, art.6, para.4 states that:

“The fact that an accused person acted pursuant to an order of a Government or of a superior shall not relieve him or her from responsibility, but may be considered in mitigation of punishment if the International Tribunal for Rwanda determines that justice so requires.”



of the rule will be considered later in the article when a case from the International Criminal Tribunal for the former Yugoslavia is discussed in the context of duress and superior orders.

B. Adopting a (partial) defence in international lawThat there was in fact considerable international disagreement about whether it was appropriate to provide a superior orders defence in international criminal justice emerged in the discussions leading to the establishment of the International Criminal Court. In part, this debate could have been predicted given the divergence of approach at the domestic level, but there was a deeper tension: How could states ensure that their own military personnel received adequate protection whilst at the same time making sure that human rights violators in other armies could not rely on the same defence? The debate centred around whether there should be absolute liability (Brownlee’s Position B) or whether there should be a qualifi ed defence (Brownlee’s Position C). The idea of an absolute defence (Brownlee’s Position A) was consigned to history. In the event, Gaeta describes how an uneasy compromise was reached:

“In essence, the two opposing positions were reconciled simply by juxtaposing them in the same rule. The absolute liability approach was adopted for genocide and crimes against humanity, while the other approach was chosen with regard to war crimes and possibly the crime of aggression.”70

The relevant provisions read as follows:

“Article 33

1. The fact that a crime within the jurisdiction of the Court has been committed by a person 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 the order was unlawful; and (c) The order was not manifestly unlawful.2. For the purposes of this article, orders to commit genocide or crimes

against humanity are manifestly unlawful”.

70 Paola Gaeta, “The Defense of Superior Orders: The Statute of the International Criminal Court versus Customary International Law” (1999) 10 European Journal of International Law 172, 189.



The decision to specify that genocide and crimes against humanity are manifestly unlawful in art.33(2) but to omit war crimes seems perverse when it is diffi cult to conceive of a situation where a war crime would not be manifestly unlawful. Those facing trial at the International Criminal Court though do have a defence of superior orders provided the conditions of art.33(1) are met. However, given the offences that come within the International Criminal Court’s jurisdiction,71 it is unlikely to envisage the defence proving successful.

Having provided a brief survey of comparative and international approaches towards allowing a superior orders defence, the article will now consider the normative basis behind the claim that a defence should be allowed, at least in some situations. To start this task, consideration will be given to the widely-accepted defence of duress because in both cases the defendant seeks to excuse voluntary conduct on the basis that unreasonable pressure was brought to bear on them.

V. The Relationship between Duress and Superior Orders

Implicit in a claim of superior orders is a sense of compulsion or lack of autonomy; although the individual still had the capacity to choose how to respond, his decision-making was compromised at the time. What is the basis for this assertion? An initial concern is that the decision was restricted because unacceptable pressure was brought to bear on the individual to act in a particular way. The act may be intended but it is not desired. This pressure distinguishes someone who kills after being instructed to do so by a superior offi cer, who arguably carries less culpability than someone who kills of his own volition. The potentially severe ramifi cations of non-compliance impact on the person’s decision-making and, whilst few decisions are taken in isolation from other factors, they are of suffi cient severity to challenge acceptable notions of culpability. A secondary, and often linked, concern is that there is inadequate time for meaningful refl ection given the manner and context in which a command is issued. These issues are fundamental to the case for having a superior order defence but, as the previous sections have demonstrated, the case law in domestic and international courts and tribunals has not grappled properly with notions of freedom and culpability. Where reasons are provided, they tend to be policy-based and relate to the need for military personnel to obey orders without question except where the nature of the order would be of evident illegality. However, some of these issues have been debated with regards to the accepted defence of duress.

The scope of duress differs according to jurisdiction, but in essence, it excuses illegal conduct where an individual was threatened with death or serious injury if

71 UN General Assembly, Rome Statute of the International Criminal Court, art.5.



he refused to offend.72 Aspects of duress are similar to a claim of superior orders. In both cases, the conduct is voluntary in the strict sense of the word and, again speaking strictly, the defendant chose to accede to the threat or to the order. But, with regards to duress, it is recognised universally that the pressure exerted on the defendant reduces his culpability to the point that it would be unjust to convict and punish.73 The next section will attempt to construct a similar argument regarding superior orders.

It was seen in the analysis of the existing law on superior orders that distinctions have sometimes been drawn depending on the gravity of the offence committed. The same is true with duress and the restriction is narrower and, rather than relying on a generic test of obvious unlawfulness or manifest illegality, is offence-specifi c and denies the defence in domestic law to murder74 and attempted murder and to war crimes, crimes against humanity and genocide in international criminal justice. As legal systems share a belief that duress should provide the basis for a defence, this section will focus on how the defence has been limited and how the courts have justifi ed removing the defence when these most serious of crimes have been committed. Duress remains available to many other offences which are manifestly illegal. What is distinctive about the offences where the defence is unavailable? By mapping out the limits of the defence, we can gain a sense of the competing policy objectives. The leading English case and the most important authority in international criminal justice will be considered in turn.

The principle that duress cannot operate as a defence to murder in English law was established in Howe.75 Lord Hailsham LC advanced a distinction between cases where the instant offence was murder and cases which did not involve the loss of life on the following basis:

“Other considerations necessarily arise where the choice is between the threat of death or a fortieri of serious injury and deliberately taking an innocent life. In such a case a reasonable man might refl ect that one

72 In England and Wales, Lord Bingham outlined the seven requirements in R v Hasan [2005] 2 AC 467:

“(i) there must be a threat of death or serious injury; (ii) that threat must have been made to D or D’s immediate family or someone for whom D

would reasonably regard as responsible; (iii) D’s perception of the threat and conduct in response are to be assessed objectively; (iv) D’s conduct must have been directly caused by the threats; (v) there must have been no evasive action for D to reasonably take; (vi) D cannot rely on threats to which D has voluntarily laid open; (vii) defence is unavailable to murder, attempted murder or treason”.

73 See Joshua Dressler, “Exegesis of the Law of Duress: Justifying the Excuse and Searching for Its Proper Limits” (1989) 62 Southern California Law Review 1331; Claire O Finkelstein, “Duress: A Philosophical Account of the Defense in Law” (1995) 37 Arizona Law Review 251; Stephen J Morse, “Culpability and Control” (1994) 142(5) University of Pennsylvania Law Review 1587.

74 R v Howe [1987] AC 417.75 Ibid.



innocent human life is at least as valuable as his own or that of his loved one. In such a case a man cannot claim that he is choosing the lesser of two evils. Instead he is embracing the cognate but morally disreputable principle that the end justifi es the means.”76

At fi rst glance, Lord Hailsham’s argument appears wholly concerned with sanctity of life, but in truth, it amounts to a claim that a defendant who kills in this situation is blameworthy (and hence deserving of conviction and punishment), whereas someone who commits a lesser offence lacks culpability. In the case of death, the defendant “is embracing the cognate but morally disreputable principle that the end justifi es the means”. But, by invoking “the lesser of two evils” argument to justify the availability of a defence for more minor offences, Lord Hailsham opens up an objection to his refusal to countenance a defence to murder. One can envisage scenarios where a killing under duress results in a net saving of life rather than a net loss of life — an example would be where a gang took the defendant’s family hostage and threatened to kill them all unless the defendant killed one individual. Suddenly the moral case for restricting the defence appears suspect on a utilitarian basis. Lord Hailsham was mindful of this possibility but argued that it could be remedied by determining that prosecution was not in the public interest or by mitigating sentence. The possibility of treating duress as mitigation allows fl exibility, which may be useful if the threat was ambiguous or where the alleged consequence(s) of non-compliance are not suffi ciently serious to warrant a full defence (say where the threat relates to property or reputational damage). Decisions about whether it is in the public interest to prosecute where several people have been killed places a heavy onus on a prosecutor, although, again, the residual discretion could protect those placed in such an invidious position.

Domestic criminal law scholars have been able to treat multiple-victim cases as ones of theoretical interest. In international criminal law though, the issue has arisen in a practical context, most (in)famously in Prosecutor v Erdemović.77 During the ethnic confl ict in the former Yugoslavia, Erdemović, a 19-year-old conscript to the Bosnian Serb army, was instructed to take part in the assassination of about 70 men and boys. After initially refusing, he was told that, if felt sympathy with the victims, he could stand alongside them. He took part in the massacre. His involvement only came to light when he described the events in detail to a foreign

76 Ibid., 433.77 [1997] 2 ICTY Jud Rep 1610 (International Criminal Tribunal for the former Yugoslavia). For detailed

analysis, see Aaron Fichtelberg, “Liberal Values in International Criminal Law: A Critique of Erdemović” (2008) 6(1) Journal of International Criminal Justice 3; Peter Rowe, “Duress as a Defence to War Crimes after Erdemović: A Laboratory for a Permanent Court?” (1998) 1 Yearbook of International Humanitarian Law 210; Olivia Swaak-Goldman, “Prosecutor v Erdemović, Judgement Case No IT-96-22-A” (1998) 92(2) The American Journal of International Law 282; David Turns, “The International Criminal Tribunal for the Former Yugoslavia: The Erdemović Case” (1998) 47(2) International and Comparative Law Quarterly 461; Thomas Weigend, “Kill or Be Killed: Another Look at Erdemović” (2012) 10(5) Journal of International Criminal Justice 1219.



television crew. He pleaded guilty but stated that he had committed the killings under duress (a claim which would by implication mean that he had been acting under superior orders).

A majority of the Appeal Chamber held that duress was not available as a defence to war crimes or crimes against humanity. Few cases expose the dilemmas facing criminal courts so starkly. No one can sensibly deny the magnitude of the harm committed or the acute pressure Erdemović faced at the time, but attempting to reconcile this fairly appears futile. Even if it is accepted that quantifying the harm is comparatively straightforward, quantifying his culpability is fraught with diffi culty. It would not be unreasonable to advance an argument that the exceptionalism of the situation removes any notion of blameworthiness. Alternatively, and here the reasoning is similar to Lord Hailsham’s in Howe, one could say that, by prioritising his own survival over that of his victims, his actions were “morally disreputable”, culpable and deserving of punishment. There are good grounds why justice should not allow people to save their own lives by killing others. As in Howe, Judges McDonald and Vohrah accepted that the basis for a defence of duress was a lack of blameworthiness but found that the sanctity of life was such that the defence should not be available with respect to war crimes or crimes against humanity.78 This view was not shared universally. Judge Stephen found no basis for restricting the defence in this way if the ultimate justifi cation for duress related to blame.79

To paraphrase Brownlee’s schema fairly liberally, what has emerged both in English law and in international criminal justice with regards to duress, is a variant of Position C. Duress exists as a potential defence save where the offence is particularly serious. This is defi ned narrowly so as to include only a limited number of specifi ed offences. The general principle behind the defence — that the threats made reduce the defendant’s culpability to the vanishing point — appears compelling. What needs to be decided is whether the grounds for limiting the defence are sound. The relevance of this to an assessment of the desirability of a superior orders defence is that it may assist in determining the scope of such a defence if it is concluded that one should be available. It would seem incongruous to have an absolute defence of superior orders if duress were to be limited in terms of the offences to which it could apply. Moreover, cases involving military personnel may well involve the use of unlawful fatal force.

78 Erdemović (n.77), [75] states that:

“[If] national law denies recognition of duress as a defence in respect of the killing of innocent persons, international criminal law can do no less than match that policy since it deals with murders often of far greater magnitude. If national law denies duress as a defence even in a case where a single innocent life is extinguished due to action under duress, international law, in our view, cannot admit duress in cases which involve the slaughter of innocent human beings on a large scale.”

79 Ibid., separate judgment of Judge Stephen, [25]–[64].



VI. The Case for a Defence of Superior Orders

It will be seen from the earlier survey of English, comparative and international criminal law that there is a lack of commonality in regards to how claims of superior orders are dealt with. All systems, however, accept that it should either provide a defence, albeit usually in restricted circumstances, or should act as mitigation. This suggests that there is no fundamental dispute that following superior orders impacts on individual culpability. The debate is how to accommodate this and whether it is possible to argue that the pressure exerted on the individual to comply is such that a full defence should be available. If not, there seems acceptance that complying with superior orders amounts to mitigation. Both approaches will be considered in turn as it should not be assumed that mitigation is a default position if the case for a defence is not made out. This section will consider the case for a defence which could apply either in all situations where illegal orders are followed (Brownlee’s Position A) or which could apply save where the order was manifestly illegal (Brownlee’s Position C).

Simester et al. use Kelly v Ministry of Defence80 to illustrate the pressure that can face military personnel.81 A soldier serving in Northern Ireland mistook a joy-rider for a terrorist and shot and killed him. Altering the facts slightly, they ask what should happen if the soldier had been ordered to shoot because a superior had mistaken the victim for a terrorist? It would appear that the soldier cannot win:

“if he refused the order he might well be court-martialled for disobeying a lawful order, whereas obedience to an unlawful order may lead to a mandatory life sentence …[a] junior soldier is surely in no position to decide these issues”.82

The law regarding the use of force, and particularly fatal force, is far from clear and junior soldiers will be in no position to access legal information.83 Decision-making must often be instantaneous. Kelly also shows how incidents frequently occur in charged political contexts which make it especially diffi cult to achieve an objectively just outcome.84

James B Insco foresees practical ramifi cations if there is absolute liability (Brownlee’s Position B):

“[In the absence of a defence], it is likely that hesitation in carrying out orders will increase; there will be more instances of insubordination, and

80 [1989] NI 341.81 Andrew Simester, John Spencer, Robert Sullivan and Graham Virgo, Simester and Sullivan’s Criminal

Law: Theory and Doctrine (Oxford: Hart Publishing, 5th ed., 2013) p.756.82 Ibid.83 See Insco, “Defense of Superior Orders before Military Committees” (n.36) p.391.84 Simester, Spencer, Sullivan and Virgo, Simester and Sullivan’s Criminal Law (n.81) p.756. See also the

reaction to Calley (n.50).



in volunteer armed forces, recruiting may be adversely affected. Indeed, as soldiers begin questioning every order, military preparedness erodes, which has devastating effects for a nation’s security.”85

He also raises a point of principle:

“The absolute liability approach is unsatisfactory because it seeks to categorically eliminate a defense of superior orders by mere linguistic manipulation. Since the absolute liability approach asserts that soldiers are only bound to follow legal orders, it eliminates the need for a defense which by defi nition is available only when soldiers follow illegal orders. When a soldier is given an illegal order, the demands of military discipline, which are enforced by legal sanctions, confl ict with the necessity to preserve the supremacy of the law by identifying and punishing criminal violators. The absolute liability approach promotes the supremacy of the law, but only by impaling itself on the military disciplinary horn of the dilemma.”86

An initial observation is that jurisdictions which provide a defence remove it in cases of manifest illegality. This suggests that whatever justifi cations there might be for recognising a defence are outweighed once a particular threshold of harm has been met. The way in which the courts have restricted the defence of duress reveals something of how competing values are reconciled. The basis for having the defence is recognised — this is a recognition that someone who offends under duress does so with greatly reduced culpability — but the emphasis given to this becomes subservient to the perceived need to protect the sanctity of life in murder cases. It should be emphasised that the courts do not claim that there is no reduction in culpability even where someone kills under duress. Two things happen. Most obviously, sanctity of life overrides the acknowledged reduction in culpability. The second point is far more suspect as the courts transfer culpability to the decision to value one’s own life above another’s. What we do know from the law on duress is that culpability can be used (however convincingly) as a justifi cation for a defence and as a justifi cation to limit a defence. An argument could be advanced that adopting Position C and allowing a defence unless the order is manifestly illegal recognises culpability because it presumes that no culpability attaches to a failure to know the law except where the illegality was palpably obvious:

“The manifest illegality principle allows the subordinate to presume that his orders are legal, and obedience to those orders is a defense unless the illegality of the orders is obvious to any person of ordinary understanding.

85 See Insco, “Defense of Superior Orders before Military Committees” (n.36) p.392.86 Ibid.



The presumption that orders are legal helps maintain and promote good order and discipline. Since subordinates do not risk incurring liability in most situations, the presumption effectively compensates for the subordinate’s lack of information and eliminates the possibility of hesitation and delay in carrying out orders.”87

Like duress, an argument for adoption of Position C on the basis of culpability can be made; again, culpability can be used as a justifi cation for a defence and as a justifi cation to limit a defence. Military personnel are under a duty to follow instructions and those who fail to comply are culpable as insubordination threatens the capability of the armed forces. It is for this reason that military law provides sanctions for those who do not follow orders. The onus on personnel to comply allied to the evident social benefi t in having forces with a clear and effi cient command structure means that someone who obeys an unlawful command should ordinarily be provided with a defence as greater blame would attach to a failure to respect the order. So far, culpability has been used as the basis for a defence (and up until this point it could justify either Brownlee’s Position A or Position C). Culpability though can limit the defence in order to favour Position C. Once the illegality is palpable and manifest, someone who complies with the order is more culpable than someone who does not. The social benefi t argument also becomes more problematic. Whilst there is evident utility in protecting soldiers who obey unlawful superior orders in most cases, it would seem impossible to justify this in cases where, for example, civilians were killed as in the Korean and Vietnam War cases.

The uncomfortable reality though is that public sentiment may be more permissive. We have already seen how there was mass outrage in the United States after a participant in the My Lai Massacre was refused a superior orders defence and sentenced to life imprisonment.88 A telephone poll conducted for Newsweek a month after Calley’s conviction found that 79 per cent of the American public disagreed with the verdict and 81 per cent believed that the sentence was too harsh.89 The objection presumably centred on Calley’s perceived culpability; a life sentence was deemed disproportionate even though he admitted to participating in a massacre. It is dangerous to read too much into one example, particularly one occurring in a unique historical and political context, but this may give some support to a claim that popular notions of culpability would support a complete defence (Brownlee’s

87 Ibid., p.393.88 Calley (n.50).89 Cited in Claude Cookman, “An American Atrocity: The My Lai Massacre Concretized in a Victim’s

Face” (2007) 94(1) Journal of American History 154. Sympathy for Calley went beyond the United States. An Australian study conducted after the massacre found that 59 per cent of the sample thought that the soldiers should be let off and 30 per cent reported that they would shoot civilians if instructed to do so; Leon Mann, “Attitudes towards My Lai and Obedience to Orders: An Australian Survey” (1973) 25(1) Australian Journal of Psychology 11.



Position A). Perhaps, and this is speculation, the public reaction to the mistreatment of Iraqi prisoners by American and British forces shows growing unwillingness to pardon such behaviour.90 The point though is that a model of culpability supporting Position C does not necessarily refl ect popular notions of blame.

The need to deter unlawful conduct is also an important concern. Removing the defence of superior orders in cases of manifest illegality (Brownlee’s Position C) may cause service personnel to refl ect on their actions and potentially desist from committing atrocities. Unfortunately, experience should make us wary of believing that the existence of sanctions in international criminal justice will affect behaviour.91 It is simply naïve to believe that the punishments imposed are likely to have any impact on later confl icts. This will come as no surprise to anyone with knowledge of the criminological research into how punishment impacts on rates of future offending. The consensus of opinion is that likelihood of conviction has some limited effect on crime rates, but the imposition of a particular penalty has no discernible effect.92 Having no defence (Brownlee’s Position B) maximises the potential for a conviction but, due to the number of individuals who have followed unlawful superior orders, prosecution is necessarily highly selective. Most combatants, it is suggested, would not entertain the thought that they might be tried for manifestly illegal conduct at a later date.

VII. The Case for Mitigation

Sentencing is arguably the primary business of the criminal court given the proportion of offenders who admit guilt prior to trial. Despite this, it is striking how little academic attention has been devoted to the sentencing process. There is a vast literature surrounding the purposes of punishment but comparatively little is known about how this translates to sentencing practice. A search for work on mitigation in England and Wales found that the defi nitive (and only sole authored)

90 The abuse at one notorious institution is documented in Mohamad G Alkadry and Matthew T Witt, “Abu Ghraib and the Normalization of Torture and Hate” (2009) 11(2) Public Integrity 135; Sean D Murphy, “U.S. Abuse of Iraqi Detainees at Abu Ghraib Prison” (2004) 98(3) The American Journal of International Law 591. See also Sunita Patel, “Superior Orders and Detainee Abuse in Iraq” (2007–2008) 5 New Zealand Yearbook of International Law 91.

91 See Payam Akhavan, “Beyond Impunity: Can International Criminal Justice Prevent Future Atrocities?” (2001) 95(1) The American Journal of International Law 7; Hunjoon Kim and Kathryn Sikkink, “Explaining the Deterrence Effect of Human Rights Prosecutions for Transitional Countries” (2010) 54(4) International Studies Quarterly 939; Julian Ku and Jide Nzelibe, “Do International Criminal Tribunals Deter or Exacerbate Humanitarian Atrocities?” (2006) 84(4) Washington University Law Review 777.

92 See Daniel S Nagin, “Deterrence in the Twenty-First Century” (2013) 42(1) Crime and Justice 199; Michael Tonry, “Learning from the Limitations of Deterrence Research” (2008) 37(1) Crime and Justice 279; Andrew von Hirsch, Anthony E Bottoms, Elizabeth Burney and P-O Wikstrom, Criminal Deterrence and Sentence Severity: An Analysis of Recent Research (Oxford: Hart Publishing, 1999).



text was 35 years old93 and that there has been one recent edited collection which considered the role of mitigation and aggravation.94 There is a body of American literature which has reviewed the impact of mitigation, but almost exclusively in the particular context of capital cases.95 Broader reviews of sentencing law do discuss mitigating factors in some depth usually by reviewing how courts have responded to particular pleas in reported cases.96 None of these works, though, touch upon how the courts should treat superior orders if it was ever raised in mitigation.

In order to understand the role that superior orders should play at the sentencing stage it is necessary to start with the overall aims of a particular sentencing system. To illustrate how this might work, the example of England and Wales will be taken because that is the jurisdiction best known to the author. According to Statute, sentencers must consider a diverse and potentially confl icting list of aims: the punishment of offenders; the reduction of crime (including its reduction by deterrence); the reform and rehabilitation of offenders; the protection of the public; and the making of reparation by offenders to persons affected by their offences.97 Commentators have despaired at the lack of coherence and failure to specify which aim(s) should take precedence.98 (The Government at the time saw no confl ict and stated that sentencers could balance the objectives as they saw fi t.)99 However, when the Act is scrutinised more carefully, one aim appears paramount. All of the key disposition thresholds which determine when it is appropriate to impose a custodial

93 Joanna Shapland, Between Conviction and Sentence – The Process of Mitigation (London: Routledge and Kegan Paul, 1981).

94 Julian V Roberts (ed), Mitigation and Aggravation at Sentencing (Cambridge: Cambridge University Press, 2011).

95 See Michelle E Barnett, Stanley L Brodsky and Cali Manning Davis, “When Mitigation Evidence Makes a Difference: Effects of Psychological Mitigating Evidence on Sentencing Decisions in Capital Trials” (2004) 22(6) Behavioral Sciences and the Law 751; Stephen P Garvey, “Aggravation and Mitigation in Capital Cases: What Do Jurors Think?” (1998) 98(6) Columbia Law Review; Jeffrey L Kirchmeier, “A Tear in the Eye of the Law: Mitigating Factors and the Progression toward a Disease Theory of Criminal Justice” (2004) 83(2) Oregon Law Review 631.

96 See Andrew Ashworth, Sentencing and Criminal Justice (Cambridge: Cambridge University Press, 6th ed., 2015) pp.176–204; Susan Easton and Christine Piper, Sentencing and Punishment: The Quest for Justice (Oxford: Oxford University Press, 3rd ed., 2012) pp.86–89.

97 Criminal Justice Act 2003, s.142. Similar provisions relating to offenders under 18 are listed in s.142A, the key distinction being the omission of the reduction of crime (including its reduction by deterrence) from the latter. It goes without saying that “the punishment of offenders” cannot be a justifi cation for punishment despite it being a stated aim. It is suggested that it is an implicit reference to retribution.

98 See Andrew von Hirsch and Julian V Roberts, “Legislating Sentencing Principles: The Provisions of the Criminal Justice Act 2003 Relating to Sentencing Purposes and the Role of Previous Convictions” [2004] Crim LR 639; Laurence Koffman, “The Rise and Fall of Proportionality: The Failure of the Criminal Justice Act 1991” [2006] Crim LR 281.

99 HC Deb 20 May 2003, vol 405, col 868 states that: “[the aims] are not in the least bit contradictory and that is why we were careful to ensure that they were not contradictory in the provisions on the purposes of sentencing”.



sentence,100 a community sentence101 or a fi ne102 rest on offence seriousness. The sentence has to be a proportionate response to the gravity of the offending and this suggests that retributivism is the dominant concern.103

Retribution is a natural response to wrongdoing and has been used as a “justifi cation” for the most extreme forms of punishment. However, it is simplistic to assert that retribution is by defi nition harsh and reactionary. In the 1970s and 1980s, liberal criminal lawyers advocated a system of “just deserts” as a mechanism for ensuring that offenders did not receive disproportionate sentences when their instant offences were minor. The concern was twofold: a focus on rehabilitation meant that the sentence was determined with primary regard to the offender’s perceived needs and this did not necessarily refl ect the gravity of the offence for which he was being sentenced; and an increasing realisation that, for all the good intentions, schemes designed to rehabilitate were failing to demonstrate success. This version of retributivism was designed to limit punishment to that which was deserved. To quote two leading proponents:

“Punishing someone consists of infl icting a deprivation on him, because he has purportedly committed a wrong, under circumstances and in a manner that conveys disapprobation of the offender for his wrong. The visitation of censure … is a central justifying feature of punishment. If one asks why penalties ought to be proportioned to the gravity of the criminal conduct, therefore, the answer is not that this would create the optimum deterrent or inhibition-reinforcing infl uence, for it may or may not do so. The requirement of proportionate punishment is, instead, derived directly from the censuring implications of the criminal sanction. Once one has created an institution with the condemnatory implications that punishment has, then it is a requirement of justice, not merely of effi cient crime prevention, to punish offenders according to the degree of the reprehensibleness of their conduct. Disproportionate punishments are unjust not because they possibly may be ineffective or counterproductive, but because they purport

100 According to Criminal Justice Act 2003, s.152(2):

“The court must not pass a custodial sentence unless it is of the opinion that the offence, or the combination of the offence and one or more offences associated with it, was so serious that neither a fi ne alone nor a community sentence can be justifi ed for the offence.”

101 Ibid., s.148(1) states that:

“A court must not pass a community sentence on an offender unless it is of the opinion that the offence, or the combination of the offence and one or more offences associated with it, was serious enough to warrant such a sentence.”

102 Ibid., s.164(2) states that: “The amount of any fi ne fi xed by a court must be such as, in the opinion of the court, refl ects the seriousness of the offence.”

103 See Gavin Dingwall, “Deserting Desert? Locating the Present Role of Retributivism in the Sentencing of Adult Offenders” 47(4) Howard Journal of Criminal Justice 400.



to condemn the actor for his conduct and yet visit more or less censure on him than the degree of blameworthiness of that conduct would warrant.”104

Personal characteristics remain relevant but for a different purpose. Rehabilitation is what is termed a consequentialist justifi cation for punishment, along with deterrence and public protection. In each case, the harm of punishment is justifi ed with reference to the benefi t to society of a reduction or cessation of offending by the offender in the future. Each consequentialist justifi cation seeks the same end but by different means. But, in each case, personal characteristics of the offender are highly relevant. In order to justify a sentence on the grounds of rehabilitation, for example, a study would have to be made of the factors that had led to the offending behaviour and consideration given as to how meaningful remedial action could take place. This personalised approach was one of the main concerns of retributive critics because those who had committed identical crimes could legitimately receive radically different sentences on this basis.

Why are individual traits important if a proportionate punishment is sought? It is widely recognised that the gravity of an offence cannot be measured solely in terms of the harm committed. Different individuals could commit an identical illegal harm but with different degrees of culpability, hence the residual, though distinct, relevance of personal characteristics. Some distinctions are so marked that they are taken as a given: young offenders and mentally disordered offenders are dealt with routinely in a separate manner for this reason.105 Culpability was considered by the Sentencing Guidelines Council (SGC) in their fi rst guidance on offence seriousness.106 Four general levels of culpability, based loosely on variants of mens rea in substantive criminal law, were identifi ed:

“(i) [Where the offender] has the intention to cause harm, with the highest culpability when an offence is planned. The worse the harm intended, the greater the seriousness.

(ii) Is reckless as to whether harm is caused, that is, where the offender appreciates at least some harm would be caused but proceeds giving no thought to the consequences even though the extent of the risk would be obvious to most people.

(iii) Has knowledge of the specifi c risks entailed by his actions even though he does not intend to cause the harm that results.

(iv) Is guilty of negligence”.107 (Emphasis in original.)

104 Andrew von Hirsch and Andrew Ashworth, Proportionate Sentencing: Exploring the Principles (Oxford: Oxford University Press, 2005) p.134.

105 Although it is worth bearing in mind that despite this, the stated aims of punishment in ss.142 and 142A of the Criminal Justice Act 2003 are nearly identical.

106 Sentencing Guidelines Council (SGC), Overarching Principles – Seriousness: Defi nitive Guideline (London: SGC, 2004).

107 Ibid., para.1.7.



In the superior orders cases reviewed earlier, the harm has always been intentional and so, according to this framework, the individual’s culpability is high. Few, though, involved planning, so the highest category of culpability is avoided. The guidance then turns to aggravating and mitigating factors. It is immediately obvious that the list of aggravating factors is far longer than the list of mitigating factors. Only four factors indicating “signifi cantly lower culpability” are mentioned:

(1) A greater degree of provocation than normally expected. (2) Mental illness or disability. (3) Youth or age, where it affects the responsibility of the individual defendant. (4) The fact that the offender played only a minor role in the offence.108

At fi rst glance, there would be little here to suggest that the fact the offender followed the orders of a superior would be looked at favourably. It is perhaps possible to argue that a military command could be analogous to a form of provocation, but that would be to stretch the meaning and is clearly not what the SGC envisaged. Age may be more meaningful if a young recruit followed an illegal command issued by a senior offi cer as that would “impact on the responsibility of the individual defendant”. More signifi cantly, the Council recognises that sentencers still have a statutory right to take account of any matters that “in the opinion of the court, are relevant in mitigation of sentence”.109 This would give the courts licence to mitigate the sentence of someone following an unlawful order. It is to be hoped that the courts would recognise that this is appropriate due to the decrease in personal culpability. A remedy, therefore, is available, although neither the Act nor the Guidelines compel a sentencer to conclude that following the orders of a superior affects culpability in all cases. In the majority of cases, however, one would expect sentencers to draw the conclusion that, as the offender was complying with superior orders, his culpability was reduced.

VIII. Conclusion

The crux of this article is the extent to which the culpability of an individual who follows unlawful superior orders is diminished and how the legal system should make allowance for this. If it is accepted that someone who is obeying the orders of a superior is less blameworthy, the primary task is to determine the most appropriate remedy. A failure to acknowledge that the actor’s culpability is reduced by ignoring the need for either a defence or mitigation would be profoundly unjust. An appropriate criminal justice response has to be proportionate and, crucially, this

108 Ibid., para.1.25.109 Criminal Justice Act 2003, s.166(1).



must refl ect both the harm infl icted and individual culpability.110 Quantifying harm and culpability can be exceptionally diffi cult even though instinctive judgments are made by courts on a daily basis.

Believing that there was a defi ciency in criminal justice scholarship, the present author in conjunction with Tim Hillier attempted to provide a holistic account of the role blame plays in the attribution of criminal liability.111 It was evident that assessments of culpability were central to the enterprise, yet these vital calculations appeared often to be arbitrary and subjective leading us to conclude that a “far greater understanding of blame attribution and blame quantifi cation is required if sentencing is to become a truly scientifi c and objective process”.112 The process becomes especially profound when there is a signifi cant disparity between the individual’s culpability and the harm that resulted. Cases where superior orders were observed may well fall into this category, as a review of the earlier cases demonstrates, as objectively very serious harm was commonplace but, arguably, the perpetrator had severely limited culpability. The facts surrounding Erdemović provide the most vivid illustration. Denying a superior orders defence in such cases effectively prioritises one part of the harm/culpability equation by removing the other. Or, more precisely, culpability retains relevance but its role is diminished because harm determines whether a defence exists and it is only at the sentencing stage that culpability enters the equation as a relevant factor in determining the seriousness of the offence and a sentence of corresponding severity.

In order to arrive at a fair assessment of culpability, it must be asked what the law can and should honestly expect of an individual faced with the prospect of implementing unlawful superior orders? Removing a defence and/or saying that following orders should have no bearing on sentence, would mean that only the heroic resister was not a criminal. This would be a signifi cant conclusion, although one with which all legal systems adopting Brownlee’s Position B appear comfortable. Accepting human frailty, which is central to the function of defences and mitigation, is both more realistic and a more morally credible alternative. This is not a denial of agency; a failure to acknowledge this misses the point even though, as Ormerod and Laird recognise, service personnel are trained to obey commands to the extent that it becomes “almost a refl ex action”.113 But the option to ignore a command exists; in many cases such a course of action may be viewed

110 The Criminal Justice Act 2003 in England and Wales makes this clear in s.143(1):

“In considering the seriousness of any offence, the court must consider the offender’s culpability in committing the offence and any harm which the offence caused, was intended to cause or might foreseeably have caused”.

111 Gavin Dingwall and Tim Hillier, Blamestorming, Blamemongers and Scapegoats: Allocating Blame in the Criminal Justice Process (Bristol: Policy Press, 2015).

112 Ibid., p.165.113 David Ormerod and Laird, Karl, Smith and Hogan’s Criminal Law (Oxford: Oxford University Press,

14th ed., 2015) p.426.



as insubordination and should be punished under military law but in other cases non-compliance may be justifi ed and even, potentially, heroic.

What is at issue with superior orders is meaningful choice and the freedom necessary to make such a choice. Agency, choice and freedom are complex and ambiguous philosophical concepts.114 Complexity only emphasises the need for care. It is emphatically not a reason to abandon the search for a just solution. Drawing an initial conclusion that someone with agency and choice but with less freedom to make that choice is less culpable than someone exercising that choice without constraint is self-evident. Determining how to respond is less obvious.

When the focus is on freedom of choice, there is no need to adopt a binary position. Freedom would appear to be a continuum — just as total freedom of action may be illusory, a total lack of freedom may be exceptional. The degree of freedom (and the corresponding culpability) may be diffi cult to measure objectively,115 but allowing for a continuum better refl ects this reality. Viewing superior orders as potential mitigation allows for this calibration unlike a defence which, by defi nition, succeeds or fails.

A second consideration, and again one that may tend to favour mitigation as an approach, is that questions of freedom of choice are fact-specifi c. Duress depends upon a clear threat (amongst other requirements) if it is to succeed.116 Superior orders will also be clear in many cases but it is possible to envisage situations where a soldier may act in the mistaken belief that he was being instructed to do so or that his actions would be consistent with earlier instructions or broader military objectives.117 The fact that courts could face problems applying a defence is not suffi cient grounds for abolishing it because many cases will raise few factual concerns and the courts are asked routinely to make factual determinations on the basis of partial evidence. The cases reviewed in earlier sections suggest that there was little factual dispute about what had happened. The question was whether — on the agreed facts — a defence should be available.

When freedom of choice emerges as an issue in criminal law, prior fault often surfaces as a potential problem. This would not appear to be the case with superior orders. For there to be a plausible claim of prior fault, there has to be fault, or moral culpability, associated with the earlier decision.118 It may be possible to develop

114 See Ian Carter, A Measure of Freedom (Oxford: Oxford University Press, 1999); Ian Carter, “Choice, Freedom, and Freedom of Choice” (2004) 22(1) Social Choice and Welfare 61; Meir Dan-Cohen, “Conceptions of Choice and Conceptions of Autonomy” (1992) 102(2) Ethics 221; William Glasser, Choice Theory: A New Psychology of Personal Freedom (New York: Harper Collins, 2010); Serena Olsaretti, “Freedom, Force and Choice: Against the Rights-Based Defi nition of Voluntariness” (1998) 6(1) Journal of Political Philosophy 53.

115 Dingwall and Hillier, Blamestorming, Blamemongers and Scapegoats (n.111) p.165.116 Hasan (n.72). For an argument that in international criminal law implicit duress should be recognised,

see Jennifer Bond and Meghan Fougere, “Omnipresent Threats: A Comment on the Defence of Duress in International Criminal Law” (2014) 14(3) International Criminal Law Review 471.

117 See eg Thomas (n.18).118 See Larry Alexander, “Causing the Conditions of One’s Own Defence: A Theoretical Non-Problem” (2013)

7(3) Criminal Law and Philosophy 623; John J Child, “Prior Fault: Blocking Defences or Constructing Crimes” in Reed and Bohlander with Wake and Smith (eds), General Defences in Criminal Law (n.1).



an argument which distinguishes between enlisting in an army or joining a militia or a paramilitary organisation.119 However, the difference between legitimate and illegitimate forces is not always clear and can be subjective (consider Northern Ireland) and there is a danger that outcome can determine status (a variant of victor’s justice). Unless one adopts a pacifi st stance and argues that anyone who engages voluntarily in the military is acting in a morally reprehensible manner, no prior fault should attach to a decision to enlist. Indeed, the common view would most likely be that joining the armed forces is laudable.

It is necessary now to determine a remedy. So far, the article has presented the potential remedies as alternatives but it is worth considering whether the approaches could co-exist. Could superior orders be available both as a defence and as mitigation? There are precedents in domestic law where a partial defence co-exists with a presumption that the same factor can be considered as mitigation. In England and Wales, loss of control provides a partial defence in limited situations. Critically, it is only available as a potential defence to murder. A defence is not available in cases of non-fatal violence but provocation remains relevant when determining sentence.120 Limiting the defence in this way, though, is best seen as a mechanism to allow the imposition of an appropriate penalty which supports the earlier point that the quantum of blame is reduced and, in order to refl ect this, the criminal justice system needs room to manoeuvre. As murder carries a mandatory sentence of life imprisonment,121 the only way in which this can be achieved is by providing a partial defence which reduces the charge to manslaughter and removes the mandatory penalty.122 The same problem explains the partial defence of diminished responsibility, which again is available only to murder. However, there is an obvious difference with superior orders and duress because, regardless of the approach adopted, the courts have consistently denied that these defences can apply in murder cases. Conviction is still seen as deserved where provocation is present or in cases of diminished responsibility: the partial defence exists to achieve justice at the sentencing stage. Rather than provide a partial defence for those who follow superior orders to kill, it would be preferable to remedy the injustice associated with having a mandatory sentence for murder. This explains why there is no need for a partial defence in international criminal law as there are no mandatory penalties even for genocide or crimes against humanity.

119 Hasan (n.72).120 SGC, Overarching Principles (n.106) para.1.24 states that a greater degree of provocation than normally

expected signifi cantly reduces culpability. This principle is reiterated in the Sentencing Council (SC), Assault: Defi nitive Guidelines (London: SC, 2011).

121 Murder (Abolition of Death Penalty) Act 1965, s.1(1). Schedule 21 of the Criminal Justice Act 2003 provides the minimum periods that can be served prior to parole but these are recommendatory.

122 Sentencers do not have complete discretion in sentencing manslaughter by reason of provocation as they have a statutory duty to follow the relevant guidelines contained in SGC, Manslaughter by Reason of Provocation: Defi nitive Guideline (London: SGC, 2005).



There is, however, no realistic possibility that the mandatory sentence for murder will be abolished for political reasons.123 Would an additional partial defence in cases of murder not grant sentencers the necessary discretion to arrive at a just penalty? There is not space here to develop the argument fully — it would necessitate a full exposition of loss of control and diminished responsibility — but the author would argue that there is a normative difference behind a claim that one was acting under a superior’s orders and a claim based on lack of control or diminished responsibility. In essence, there appears to be greater agency in the former and hence more culpability.124

It will probably come as no surprise that this author does not believe that the case for having a defence of superior orders has been made out. Regardless of whether the defence would apply to all offences (Brownlee’s Position A) or would exclude crimes of manifest illegality (Brownlee’s Position C), providing a defence would have practical ramifi cations and would rely on a weak philosophical justifi cation. If a defence was provided, criminal liability could only be attributed to a limited number of individuals, restricting the ability of the criminal law to respond robustly to illegal activity committed by service personnel. The criminal law co-exists with military law as a means of controlling delinquent behaviour but serious offences should be heard in a criminal court to ensure public confi dence.

The philosophical reasons for denying a defence are various. The fi rst is the need to recognise that the individual had agency, acted voluntarily and retained choice. This is not to minimise his plight but rather to show precision in locating the basis for a possible defence. Here the locus is culpability. Evidently it has been reduced due to the pressure to comply, but it would overstate the case to claim that no culpability exists. One might object to this reasoning on the basis that a defence is provided for duress and it is well-accepted. However, the nature of the threat is different and it is not unreasonable to conclude that someone acting in fear of their life is more deserving of a defence than someone fearing disciplinary action on the basis of non-compliance. In some situations, such as Erdemović’s, the threat may appear identical but that is a case involving both superior orders and duress and, as was recognised in that case, these are separate defences.125 Duress should be available as a defence in such a scenario. This would require duress to become

123 Although empirical work has cast doubt on whether the public believes that life imprisonment should be mandatory in all murder cases. See Barry Mitchell and Julian V Roberts, “Sentencing for Murder” (2012) 52(1) British Journal of Criminology 141.

124 See Steven Yannoulidis, “Excusing Fleeting Mental States: Provocation, Involuntariness and Normative Practice” (2005) 12(1) Psychiatry, Psychology and Law 23.

125 Erdemović (n.77), [34] states that:

“Superior orders and duress are conceptually distinct and separate issues and often the same factual circumstances engage both notions, particularly in armed confl ict situations. We subscribe to the view that obedience to superior orders does not amount to a defence per se but is a factual element which may be taken into consideration in conjunction with other circumstances of the case in assessing whether the defence of duress or mistake of fact are made out.”



available in murder cases and, although it is beyond the scope of this article to fully articulate my reasons, the rationales used by the courts for restricting the defence would, at least, seem worthy of revisiting.

Justice is most appropriately served by taking full account of the fact that the individual was following superior orders at the sentencing stage. This factor can be considered along with any other mitigating and aggravating factors in assessing the seriousness of the offence based both on the gravity of the harm and of the individual’s culpability. It is suggested that the fact that the individual was following the order of a superior at the time should be viewed as highly mitigating although it would be inappropriate to prescribe the exact effect that it should have on sentence. This is an argument in favour of Brownlee’s Position B. Although this still represents the position in some legal systems, notably England and Wales, it is recognised that it is a more restrictive approach to that taken in many jurisdictions which provide a defence unless the orders were manifestly illegal. Similarly, it is acknowledged that it runs counter to the trajectory in international criminal law, which has increasingly shifted from an absolute position to Brownlee’s Position C. However, the earlier arguments by the courts prior to these shifts appear apposite and accord with the recognition that choice, though restricted, remains. This is an argument that superior orders should not be relevant to criminal liability but should be relevant to punishment.


Gavin Dingwall - Journal of International and Comparative Law

Documents

Transcript of Gavin Dingwall - Journal of International and Comparative Law