Honest Errors, the Rendulic Rule, and Modern Combat Decision-Making


| Oct 24, 2023


In 1948, in a judgment commonly known as the Hostage case, a U.S. military tribunal at Nuremberg acquitted Lothar Rendulic, commander of the German 20th Mountain Army, of offences related to the devastation of the Norwegian counties of Finnmark and Troms between late 1944 and early 1945. At trial, Rendulic argued that military necessity imperatively demanded scorched earth tactics and the forced evacuation of the entire civilian population if feared Soviet advances were to be impeded although, in the event, no such advances materialised. The judges found it inappropriate to question Rendulic’s decision with the benefit of hindsight. Instead, they approached the situation “as it appeared to the defendant at the time” and ruled that Rendulic’s conclusion might have been faulty but was not criminal. The judgment gave rise to the so-called “Rendulic Rule” against second-guessing battlefield decisions made under difficult circumstances and with limited information.

In Honest Errors? Combat Decision-Making 75 Years after the Hostage Case, an international team of military historians, lawyers, and ethicists assembled by co-editors Nobuo Hayashi and Carola Lingaas presents four key findings:

1. Rendulic did not consider it militarily necessary to devastate northern Norway in its entirety or to forcibly evacuate its residents;

2. Significant deficiencies in Rendulic’s trial shed doubt on its quality and render his acquittal contentious;

3. The admonition against using hindsight, while well-established in international humanitarian law (IHL) and international criminal law (ICL) today, nevertheless requires erroneous combat decisions to be both subjectively honest and objectively reasonable;

4. War’s human conditions, diminishing room for battlefield empathy, and institutional biases make assessing the reasonableness of combat errors challenging.

Rendulic Did Not Consider Devastation and Forcible Evacuation Militarily Necessary

After the Battle of Narvik in June 1940, Norway under German occupation saw no major military operations on its territory. This began to change over the summer of 1944 when Finland, then a German ally, sued for peace with the Soviet Union and agreed to expel or intern all German forces still present on Finnish territory. In response, Rendulic was ordered to extract his 20th Mountain Army from Finland to Norway. On 7 October 1944, elements of Rendulic’s 20th Mountain Army still in Finland’s far north came under Soviet attack. Red Army soldiers entered Norway on 18 October, reached Kirkenes on 25 October and took defensive positions near the Tana River, where they engaged German forces for the last time, in early November. Meanwhile, on 28 October, Adolf Hitler ordered Rendulic to devastate Finnmark and Troms. The following day, Rendulic instructed subordinate units to destroy all houses and forcibly evacuate local inhabitants, ostensibly on account of imperative military necessity.

Historians Sven G. Holtsmark, Gunnar Åselius, Stian Bones and Gunnar D. Hatlehol show how military necessity explains only part of Rendulic’s decisions. From late 1944 until early 1945, Rendulic perceived military threats to his 20th Mountain Army retreating through northern Norway that were diverse, uncertain, and variable. He contemplated several major possibilities, some involving Soviet land invasions and others involving Soviet-UK amphibious operations. Any combination of these eventualities could materialise at different moments, a prospect exacerbated by Rendulic’s limited capacity to gather good intelligence. Although, by late November 1944, it had become increasingly clear to Rendulic that a Soviet invasion from the east area was unlikely, other threats remained.

Countering them might have necessitated pre-emptive property destruction in areas that could prove useful to the invading forces. However, the question arises whether Rendulic kept the timing, manner, and scale of such measures in proportion to that need.

Contemporaneous material portrays a man convinced that some destruction of property was necessary, but not the wholesale devastation of northern Norway and the evacuation of all residents by force. The latter ideas appear to have originated elsewhere. Finland’s exit from the war in September 1944, and the prospect of northern Norway coming under Allied control, prompted existential angst within the occupation apparatus. The London-based government-in-exile’s likely return would bring down Vidkun Quisling’s collaborator government and Joseph Terboven’s Reichskommissariat for Norway. Lessening the impact of a liberated northern Norway by removing its local population became a priority.

By early October, the destruction of military installations and evacuation of armed and labour-capable locals was already underway as part of the 20th Mountain Army’s withdrawal. Initially, Terboven failed to secure Rendulic’s support for a voluntary evacuation. On 26 October, Terboven opined that Soviet forces would invade northern Norway and that the area should be completely destroyed. Terboven then set in motion a proposed devastation of northern Norway that would reach Hitler and receive his approval on 28 October.

Rendulic and his senior officers had different priorities. They were mainly concerned with the 20th Mountain Army’s successful retreat and regarded the devastation and forcible evacuation of the region as a distraction. Once in receipt of Hitler’s 28 October order, however, the implementation of Terboven’s political manoeuvre became a foregone conclusion. It is unlikely that devastating northern Norway and forcibly evacuating its residents was a military solution to which Rendulic’s own assessment of the danger imperilling the retreating 20th Mountain Army had led him.

Deficiencies in Rendulic’s Trial

At Nuremberg, Rendulic was charged with four counts of war crimes and crimes against humanity. Count Two read, in part:

For no compelling military reasons, and in literal execution of instructions to show no sympathy to the civilian population, the evacuated residents were made to witness the burning of their homes and possessions and the destruction of churches, public buildings, food supplies, barns, livestock, bridges, transport facilities, and natural resources of an area in which they and their families had lived for generations.

Article 23(g) of the 1907 Hague Regulations forbids the destruction and seizure of enemy property unless “imperatively demanded by military necessity.” Though not specified in the indictment, this article became the focal point of Rendulic’s devastation charges and the prosecution’s case against him collapsed on a particular interpretation of military necessity.

The judges held that Rendulic had honestly, albeit erroneously, considered the region’s devastation militarily necessary in view of feared Soviet advances. They did so, however, without accounting for the geographical and temporal variability of Rendulic’s evolving threat perceptions. Instead, the judges held:

The course of a military operation by the enemy is loaded with uncertainties, such as the numerical strength of the enemy, the quality of his equipment, his fighting spirit, the efficiency and daring of his commanders, and the uncertainty of his intentions . . . . It is our considered opinion that the conditions, as they appeared to the defendant at the time were sufficient upon which he could reasonably conclude that urgent military necessity warranted the decision made.

Crucial Situational Nuances Missing From the Judgment

In their chapter on trial preparations, Stian Bones and Gunnar D. Hatlehol expose how dissimilar approaches and haphazard interactions among Norway’s local authorities, its government-in-exile in London, and the Nuremberg-based U.S. prosecutors meant that neither the indictment nor the evidence fully reflected the material Norway had gathered.

Emily Crawford notes how significant testimonial inconsistencies between key witnesses, including Rendulic himself, went unchallenged. The prosecutors did not ask why Rendulic did not adjust the severity of property destruction and forcible evacuation in various parts of northern Norway to the threats he had perceived. Nor did the prosecutors vigorously argue the existence of multiple motives, some plainly political rather than military, that had shaped Rendulic’s decisions. There is no indication that the prosecution pre-empted or countered assertions of an honest error and accusations of second-guessing, either. The defence, for its part, skilfully exploited the prosecution case’s weaknesses. Rendulic’s attorneys highlighted that no evidence revealed the absence of military necessity for property destruction and that the testimony of prosecution witnesses was either inexpert or in fact favourable to the defendant. His lawyers also utilised arguments likely to resonate with the panel, such as Allied practices and opinions, and the importance of judging the situation contemporaneously, to good effect.

The tribunal itself did not scrutinise Rendulic’s factual claims. On the contrary, the bench appears to have accepted his stated interpretations of the military situation at face value. While the judgment notes the “uncertainties” with which “the course of a military operation by the enemy is loaded,” it does not delineate the contours of such uncertainties vis-à-vis Rendulic by assessing what was and was not known or reasonably knowable to him in relation to different locations and moments. Nor, when ruling that “the conditions, as they appeared to the defendant at the time were sufficient upon which he could honestly conclude that urgent military necessity warranted the decision made,” did the judges articulate what made these conditions “sufficient.”

Sean Watts also notes the thin legal basis on which the judges acquitted Rendulic. It is unclear where in international law the judges located the admonition against second-guessing battlefield decisions. The judgment does not go through the usual list of sources, now found in Article 38 of the Statute of the International Court of Justice. Rather, having referenced “customs and practices accepted by civilized nations generally”—possibly a mixture of custom and general principles of law—the judges cautioned against “adding rigidity where resiliency is essential” and “plac[ing] the principles of international law in a formalistic strait-jacket.” Their plea for flexibility does not explain their omission to discuss the origin of a no second-guessing rule.

Erroneous Combat Decisions Must Be Subjectively Honest and Objectively Reasonable

Sean Watts also observes that, despite its obscure provenance, the Rendulic Rule seems well-settled. Nowadays, treaty provisions governing the methods and means of warfare as well as individual criminal responsibility often make implicit allowances for honest errors. Domestic enactments and influential IHL manuals exhibit States’ belief that only contemporaneously available information should be used to assess battlefield decisions. Though born of a contentious factual setting and with meagre legal reasoning, the Rendulic Rule is now firmly entrenched in the fabric of IHL.

The Hostage case is also a precursor to the so-called reasonable commander test in modern IHL and the mistake of fact defence under contemporary ICL. Yasmin Naqvi synopsises today’s law as follows. Where a commander acts reasonably given the information available at the time, an IHL violation cannot be established even if the commander errs in his or her assessment of the situation. Nor, where acting reasonably in the sense just described negates culpability or the mental element of an offence, can individual criminal responsibility be established.

The reasonable commander test becomes important in areas where IHL grants commanders a degree of discretion. Examples include military necessity—such was the case in Hostage—as well as precaution and proportionality in attacks. Actively seeking and verifying information about the status of a target and incidental civilian harm form integral parts of precautionary measures codified in Article 57 of Additional Protocol I. If attackers err in their decisions because of a failure to check the accuracy of available information, they act unreasonably and in bad faith.

Key passages of the Hostage judgment reveal that Rendulic’s acquittal was partly to do with the absence of culpability and a guilty mind. We can find similar signs in several other post-Second World War trials. It appears that these cases assessed both the subjective genuineness of the defendants’ beliefs and their objective reasonableness. Article 32(1) of the Rome Statute of the International Criminal Court (ICC) provides for a mistake of fact. The ICC should require any alleged mistake to be objectively reasonable by applying the reasonable commander test. Imposing such a requirement would enable the court to exclude wilful blindness from Article 32(1). This would also help deter reckless use of faulty information in battlefield decisions and strengthen civilian protection as a result.

The Challenge of Assessing the Reasonableness of Combat Errors

The Hostage case’s significance goes beyond the no second-guessing rule, the reasonable commander test, and the mistake of fact defence. The trial raises questions about battlefield errors that continue to preoccupy us today.

Technology and Human Element

Rendulic would have been less limited in the exercise of his honest judgment had he had a clearer picture of the battlefield. According to Tae Hoon Kim, proponents of the so-called revolution in military affairs argue that this is precisely what today’s extraordinary advances in information communications technology (ICT) bring. Improvements in intelligence-gathering and processing, coupled with modern capabilities to execute missions more accurately and precisely, equip soldiers with dominant battlefield knowledge.

Recent wars in Afghanistan, Iraq, and Ukraine present a mixed picture. Initial overall successes of ICT-driven U.S. forces in the conventional battlefield gave way to close-quarter combat with insurgents. Abundant and accurate intelligence alone did little to mitigate elementary mistakes arising from negligence, poor judgment, and fatigue. Nor did such intelligence clear away the soldier’s pervasive sense of uncertainty and insecurity. Russia’s invasion of Ukraine has exposed crippling deficiencies in communications security, imagery intelligence, and army-air force coordination. ICT advantage is, contrary to what its advocates argue, perhaps not the panacea to war’s human and error-prone ways, that is, the psychological, physical, and cognitive strains to which soldiers remain vulnerable.


Rendulic’s judges concluded that he had acted within the limits of his honest judgment. It is unclear whether they did so by comparing Rendulic’s actions to what a reasonable commander would have done in his situation. Valentin Jeutner argues that the reasonable commander test in IHL closely echoes the reasonable person test in common law. The latter test engages the actors’ imagination about their conduct from the standpoint of another, hypothetical person. This test also engages the actor’s empathy.

Portraying the reasonable commander test as an empathy-based perspective-taking technique invites tricky questions. Is the reasonable commander an ordinary person or a military person? A test that accounts for the experience of all persons affected by war may broaden the reservoir of relevant perspective-holders and offer more relatable and predictable decision-making guidance. Also, whereas acting as a State representative might justify prioritising the welfare of that State’s civilians, acting as an agent of humanity would entail treating all civilians equally.

Studies show that the more power a person possesses, the less able he or she is to empathise. This raises the possibility that the reasonable military commander—who may be assumed to hold great power to alter the state of those subjected to it—is already hampered in his or her ability to empathise. Meanwhile, modern militaries expand physical and psychological distances from their adversaries and actively promote de-empathetic modes of fighting. One might wonder whether the reasonable commander test is really fit for purpose when assessing today’s battlefield decisions.

Institutional Bias

Rendulic argued that forcible evacuation had saved the locals from the hardship of devastation and that, had the Soviets invaded, it would have saved them from Bolshevism. This reveals Rendulic’s self-serving and ideologically coloured view of his own actions and their consequences, and his lack of imagination about the magnitude of harm inflicted on the region’s residents. Matthew Talbert and Jessica Wolfendale compare views like these with portrayals of civilian harm caused by U.S. drone strikes as “tragic mistakes” and “awful but lawful.” Such reportage perpetuates and amplifies assumptions about the U.S. military’s moral superiority as an organisation dedicated to protecting civilians with precision weapons and about the blameless inevitability of civilian deaths in modern combat.

A harmful act’s blameworthiness depends on its moral impropriety seen from the victim’s point of view. It is the lack of moral regard the actors owe their victims that makes blame appropriate. Knowledge also plays an important role. A military distributes, suppresses, and withholds knowledge amongst its personnel. Drone operators would be hard-pressed to go out of their way and authority to check all accessible as well as hidden target data. Nevertheless, the objective reasonableness of their belief might depend on how judgments about legitimate and illegitimate targets had been formed higher up in their chain of command. There are worrying signs that the U.S. military denies its soldiers true knowledge about the impact of drone strikes on civilians and, consequently, true belief about the sincerity of its stated commitment to minimising incidental harm. The U.S. military would do well to make good on its professed concern for civilians by treating all civilians, American or non-American, as morally equal.

Food for Thought

The Hostage case’s shortcomings need not invalidate the admonition against second-guessing as an idea. At the end of the day, all battlefield decisions are limited in their honesty and reasonableness, and law demands their meaningful assessment in one way or another. If not from Rendulic’s acquittal, his eponymous IHL rule would have eventually emerged elsewhere.

Our findings point to new research areas. Should second-guessing be permitted where it benefits the decider? Rendulic most likely invoked military necessity less as his honest and reasonable belief of the situation prevailing at the time than as a post factum rationalisation for Terboven’s political project. Had Rendulic’s perceived threats materialised, however, devastating northern Norway and forcibly evacuating its inhabitants would have slowed enemy advances. If measures generated the right results, why should anyone protest afterwards that it had been conceived for the wrong purpose?

Can a sound legal rule come out of a factually unsound case? The Hostage tribunal could have upheld the idea of no second-guessing in principle and rejected the alleged honesty of Rendulic’s errors, but it did not. The prosecution’s lackadaisical work and the tribunal’s readiness to take Rendulic at face value may explain why, despite its central importance to IHL and ICL, the Rendulic Rule has remained relatively obscure and its precise content difficult to articulate.

Will tomorrow’s warfare compel us to re-imagine the reasonable commander test and the mistake of fact defence? Full combat decision autonomy will not eliminate room for harmful errors. When they occur, questions such as what it means for relevant information to have been available to the decision-maker (human or otherwise), what counts as its honest and reasonable assessment, and when accountability is engaged, will need answers.


Nobuo Hayashi is an Associate Senior Lecturer at the Centre for International and Operational Law, Swedish Defence University.


Photo credit: Pexels