A. Desertion: Its Continued Relevance and (Non)Regulation

Despite international humanitarian law’s growth over the past decades, none of its treaties or instruments expressly addresses desertion or the different facets of desertion. Indeed, the term “desertion” is not used or defined in any of these documents. Also, compared with other conflict-related issues, desertion seems to have attracted relatively less attention from commentators. Desertion is neither rare nor uncommon in times of armed conflict. This was the case in both World War I [WWI] and WWII, and was to be expected, given the natural inclinations of the ordinary fighter involved in those world wars.Footnote ¹¹ Despite this, James Clause noted in 1961 that the lack of international legal scholarship on desertion may be due to the “fact that few controversial problems have been generated in the past”.Footnote ¹²

Military strategists and scholars observe that wars are won not through the number of most killed but the loss of fighting members, which can take place through desertion.Footnote ¹³ Niall Ferguson argues that, based on this rationale, one should pursue the surrender of one’s enemies in war.Footnote ¹⁴ The same rationale would also call for the encouragement of enemy desertion if one is seeking to end or de-escalate conflicts. Desertion may also become of increasing interest given the nature of contemporary conflicts. In a 1985 paper, Sassòli observed that desertion has “taken on increased importance in recent years” as conflicts are increasingly perceived as ideologically driven, and as opposing sides to a conflict try to convert combatants to their ideology.Footnote ¹⁵ The use of militia or economically motivated fighters also impacts desertion.Footnote ¹⁶ Research in political science and sociology has sought to explain why fighters switch sides or stop fighting.Footnote ¹⁷ Despite the continuing relevance of desertion in modern-day conflict, and its strategic consequences, none of today’s main international humanitarian law instruments expressly address desertion.

Existing international humanitarian law instruments do not directly regulate desertion. During WWII, the few international agreements that then applied did not deal with deserters.Footnote ¹⁸ The postwar 1949 Third Geneva Convention relative to the Treatment of Prisoners of WarFootnote ¹⁹ (hereinafter “the Third Geneva Convention”) also does not refer to desertion. However, through their regulation of POW status, these instruments do impact the status and protection of certain deserters who acquire POW status. Deserters recognized as POWs have a detailed set of rights and cannot be enlisted in the armed forces that they “deserted” to. In other words, even if the deserter wishes to serve in the armed forces that he deserts to, the said armed forces will be prohibited from accepting this deserter’s offer if the deserter is recognized as a POW. Whether and when a deserter obtains POW status thus becomes important. Apart from setting out the rights and obligations of POWs, the Third Geneva Convention also states when an individual obtains or loses POW status.Footnote ²⁰ As explained below, by addressing when POW status is obtained, the Third Geneva Convention also decides when and whether deserters can fight on behalf of the armed forces they deserted to. It should be noted that the Third Geneva Convention only applies to international armed conflicts, and there is also need for international instruments to address desertion in non-international armed conflicts.Footnote ²¹

B. BIA Desertions as a Historical Case-Study

The popular story of BIA desertions in WWII speaks of British Indian soldiers deserting the British military in droves when called upon to do so. Historians have criticized this narrative of BIA desertions, arguing that it fails to recognize the diverse response of BIA personnel or the fact that many deserted under coercion.Footnote ²² To get a better understanding of BIA mass desertions in Singapore, it is necessary to go further back in time.

Prior to Japan’s attack on Malaya, the Japanese military established F Kikan, a military intelligence unit headed by Major Fujiwara Iwaichi.Footnote ²³ This unit was responsible for making contact with local groups, such as the Indian Independence League [IIL], and obtaining their support for Japan against the British. Throughout the war, the Japanese threw their support behind a number of anti-colonial pro-independence movements in the region.Footnote ²⁴ When Japan invaded Malaya on 8 December 1941, Fujiwara and F Kikan accompanied the Japanese military on its march down the Malay Peninsula, reaching out to BIA soldiers captured by the Japanese military. Captain Mohan Singh was one of the earliest soldiers approached by Fujiwara.Footnote ²⁵ Singh was captured by the Japanese military early on in Japan’s attack of Malaya, and Fujiwara successfully persuaded Singh to form an army composed of BIA deserters who would then fight the British for India’s independence.Footnote ²⁶

Fujiwara and Mohan Singh’s men followed the Japanese military in their lightning attack down the Malay Peninsula.Footnote ²⁷ Along the way, they persuaded abandoned or surrendered BIA members to join them. They headed to Singapore, and on 15 February 1942, Singapore fell to Japan when the Japanese military inflicted a crushing defeat on the British military. Two days after the fall of Singapore, the Japanese military separated Indian soldiers from other British troops and brought the former to Singapore’s Farrer Park.Footnote ²⁸ There, Fujiwara and Mohan Singh urged the gathered Indian soldiers to renounce their allegiance to the British and join the INA. According to Singh, who spoke at Farrer Park as the head of the new INA, the crowd responded with euphoric cheers and enthusiasm to his call to fight the British: “There was a spontaneous response from all the soldiers. Along with the raising of hands, thousands of turbans and caps were hurled up in the air … soldiers jumped to their feet … with prolonged shouts of ‘Inqilab Zindabad’ (Long Live Revolution).”Footnote ²⁹

According to Alan Warren, out of the 55,000 Indian soldiers captured by the Japanese military in 1941, 20,000 decided to join the INA when called on to do so, with another 20,000 doing so from June to August 1942.Footnote ³⁰ The INA was thus officially born. When other British military personnel in Singapore heard about the Indian soldiers’ change of allegiance, many saw it as a betrayal and reacted in shock and disappointment.Footnote ³¹ There were in fact many reasons why Indian soldiers chose to switch their allegiance. Some truly did believe Japan’s promise to liberate Asia from the shackles of Western colonial rule.Footnote ³² Others were disillusioned and angry at the British military’s racially discriminatory practices.Footnote ³³ Lower-ranking soldiers often followed the decisions of their Indian officers.Footnote ³⁴ It should be noted that not all captured Indian soldiers joined the INA. Many Indian soldiers also ended up as forced labour in the hands of the Japanese military and suffered great hardship.Footnote ³⁵

The fortunes of these Indian soldiers ebbed and flowed throughout the war. The relationship between the INA and the Japanese military was often fraught. In 1942, Mohan Singh fell out with the Japanese military leadership and was imprisoned.Footnote ³⁶ The next year, the charismatic Subhas Chandra Bose arrived in Singapore to take up leadership of the INA.Footnote ³⁷ Bose’s formation of the Provisional Government of Free India in Singapore inspired many more Indians to join the INA. Another 10,000 captured Indian soldiers joined the INA.Footnote ³⁸ By 1945, 18,000 Indian civilians had also joined the INA.Footnote ³⁹

A. Contestations of POW Status in the Singapore Trials

Questions about BIA desertions also arose in war crimes trials conducted by the British after the war. As background, at the end of WWII, the British and other Allied Powers individually conducted war crimes trials in Asia and Europe. In Asia, the bulk of British war crimes trials were conducted in Singapore, as the island served as a base for British war crimes investigations.Footnote ⁴⁷ These British war crimes trials were conducted pursuant to a Royal Warrant, passed on 18 June 1945 by the British executive, that authorized the British military to establish military courts to try violations of “the laws and usages of war”.Footnote ⁴⁸ British or Allied judges presided over these war crimes trials; the prosecutors were usually from the British military; and the majority of defendants were represented by Japanese defence counsel assisted by British defence advisory officers.Footnote ⁴⁹ Trial findings of guilt were subject to “confirmation” by a confirming officer from the British military.Footnote ⁵⁰ Those convicted at trial had the right to submit petitions for the confirming officer’s examination.Footnote ⁵¹ For the Singapore Trials, the Department of the Judge Advocate General [JAG] in Singapore would prepare a report at the confirmation stage which summarized the factual and legal arguments of the case, and this report would be considered by the confirming officer when making his decision.

In several Singapore trials dealing with war crimes committed against BIA Indian victims, the defence argued that these Indian victims had renounced their British allegiance and were, at the time of the crime, members of the Japanese military.Footnote ⁵² It was alleged that these Indian victims had been given the same treatment as Japanese soldiers and, due to the victims’ status, the crimes charged did not amount to war crimes. Not all trials involving Indian victims featured such legal discussions. It was only in two cases, discussed in this paper, that the court heard detailed legal arguments on the contested nature of BIA desertions and the victims’ status.Footnote ⁵³ These two Singapore trials—Ikegami Tomoyuki and others and Takashima Shotaro and another—featured interesting if different legal discussions at their trial and post-trial stages.

The trial of Ikegami Tomoyuki and others started on 19 November 1946 and ended on 2 December 1946.Footnote ⁵⁴ All six defendants belonged to the Ikegami Unit headed by the first accused, Ikegami Tomoyuki. Two charges were brought against the defendants based on two separate incidents dated 27 June 1945 and 18 June 1945. The facts of the case were as follows. The first accused Ikegami was the commander of a unit which included the “Romu Tai”, also referred to by trial actors as “Special Labour Force”, “Special Labour Unit”, or “Special Labour Corps”. For consistency, this paper will use the term “Special Labour Force”. The Indian victims were argued by the defence to belong to this Force. At the time of these crimes, the Ikegami Unit was operating in Lutong in the Malaysian state of Sarawak. The second, third, fourth, and fifth accused—Takahashi Yoichi, Takahashi Tatsuo, Miyoshi Ren, and Hisano Jun—were technical officers working on the Lutong oilfield. Takahashi Takeshi, the last accused, was a quartermaster working on the Lutong oilfield. The first accused held the rank of major, the second and third accused were first lieutenants, the fourth and fifth accused were second lieutenants, and the sixth accused was a corporal.

On 18 June 1945, during the unit’s withdrawal from the area while under Allied attack, Ikegami received a report that five Indians from the Romu Tai had been arrested while attempting to escape. It was claimed that the Indians had with them a secret map that showed the locations of Japanese units. Upon investigating, Ikegami concluded that they were Romu Tai members conspiring to desert to the enemy with secret information. Ikegami then ordered the victims’ execution, and the victims were executed later that day.

On 26 June 1945, eleven more Romu Tai members escaped but were recaptured on the same day. The second accused Takahashi reported this incident to Ikegami, explaining, among other things, that secret documents had been lost and found. Ikegami then conducted investigations and found that the victims had planned to desert and had stolen confidential documents. He decided the victims were to be executed. The next day, on 27 June 1945, the five other accused reported to Ikegami who explained his execution order and its grounds to these accused. The said Romu Tai members were then executed by the defendants that morning. The first charge, relating to the killing on 27 June 1945, was brought against all accused. The second charge, relating to 18 June 1945, was brought only against Ikegami.

There were three judges assigned to the case: Lieutenant Colonel H.E.R. Smith, Major E.N. Hebden, and Captain E.H. Dunsford.Footnote ⁵⁵ The prosecutor was Major S.J. Smith, who was a solicitor.Footnote ⁵⁶ The accused were defended by Japanese defence counsel Tatsuzaki Ei, who was a judge advocate from the Imperial Japanese Navy.Footnote ⁵⁷ A British defence advisory officer, Lieutenant J.H. Ward, was assigned to assist the defence.Footnote ⁵⁸ Altogether, the trial lasted seven days. The first accused Ikegami was sentenced to death by hanging. The second, third, fourth, and fifth accused were sentenced to ten years’ imprisonment each. The sixth accused was sentenced to five years’ imprisonment. The court’s sentences were nevertheless not confirmed at the post-trial stage for reasons explained further below.

The second case discussed in this paper is Takashima Shotaro and another, which was decided by the same panel of judges and which also discussed questions of allegiance. This trial started on 25 January 1947 and ended on 30 January 1947.Footnote ⁵⁹ This trial involved two accused, Takashima Shotaro and Asako Koichi. Both were charged with the ill-treatment of Indian POWs in a camp run by the two accused in Sankakuyama, New Britain. Takashima was a sergeant and in charge of the party of Indian POWs, while Asako was a corporal and a medical orderly in the camp. The prosecutor in that case was Major Sahay.Footnote ⁶⁰ The defence lawyer was Nakamura Takeshi, recorded as a barrister at the Tokyo Civil Court.Footnote ⁶¹ The defence advisory officer was Captain D.F.H. Sinclair.Footnote ⁶² The trial was conducted over six days. Two witnesses testified for the prosecution, and three individuals testified for the defence, including the second accused Asako. At the close of the prosecution’s case, defence counsel argued that the prosecution had not provided enough evidence or made out a case to answer against Takashima.Footnote ⁶³ The court agreed with defence counsel and acquitted Takashima. The trial then continued, and Asako was eventually sentenced to life imprisonment. The court’s finding and sentence against Asako was, however, not confirmed.

The position of the Indian victims in these two cases differed slightly. In Ikegami Tomoyuki and others, the Indian victims had allegedly sworn allegiance to the Japanese after being captured by the Japanese military. They were then caught trying to desert the Japanese military. In Takashima Shotaro and another, the Indian victims were subject to abuse and harsh treatment while under the command of the Japanese military. In both cases, the defence argued that the Indian victims were in fact members of the Japanese military. Specifically, these Indian victims were members of the Japanese Romu Tai or Special Labour Force. These cases were decided by the same panel of judges.

B. Trial Findings: Side-Switching as Legally Impermissible

In Ikegami Tomoyuki and others, the court started having doubts about the victims’ POW status based on facts raised by witnesses at the trial. In his opening address, defence counsel pointed out that the Indian victims were all “members of the Tokushu Romutai” who had “voluntarily sworn an oath of allegiance to the Japanese”.Footnote ⁶⁴ On 25 November 1946, when the court questioned the accused Miyoshi Ren on the nature of the unit, the latter explained that he had been “ordered and instructed by my superior officer” that the unit comprised “Indians who had sworn loyalty to the Japanese”.Footnote ⁶⁵ Miyoshi went on to explain that he always saw the Indians “being treated the same way as Japanese soldiers”.Footnote ⁶⁶ They had their own commander, who was selected from among them. There were Japanese officers employed in the unit for “liaison purposes”, and who were to “lead” and “instruct” the Indians as the Indians “did not know Japanese customs and manners”.Footnote ⁶⁷ The court also addressed the issue of allegiance during its questioning of the accused Takahashi Tatsuo. Takahashi explained that part of the job undertaken by the Special Labour Force was the “guarding of the oil fields”.Footnote ⁶⁸ The court asked Takahashi whether the Indian POWs had their own commanding officer and “a reasonable amount of freedom” within the area that their camp was located. Takahashi answered in the affirmative to both questions.Footnote ⁶⁹

The Special Labour Force’s organization, its relatively important job, and its members’ unregulated movements raised concerns about the court’s jurisdiction over the crime committed against this unit’s members. The court highlighted its concerns to the prosecutor, noting that there was “a certain amount of corroborated evidence” that the victims were not POWs but “persons of Indian birth formerly subjects of His Majesty the King” who “by oath of greater weight than a parole made to the Japanese authorities”, “renounced their rights as British subjects”.Footnote ⁷⁰ The court asked the prosecutor to consider this issue and submit his advice to the court, adjourning until the next day.Footnote ⁷¹

On 26 November 1946, when the court reassembled, it heard the prosecutor’s arguments on the allegiance issue. In the course of his arguments, the prosecutor referred to a mixture of British law and international law. He argued that any divesting of nationality should be governed by the 1914 British Nationality and Status of Aliens Act and existing British case-law, according to which a British national can only divest himself of nationality during war in favour of an enemy state when residing in that state.Footnote ⁷² Furthermore, based on British case-law, the prosecutor argued that as an individual cannot avoid the “military obligations” that arose before the renouncement of nationality; he also cannot avoid the “privileges and benefits” of those undertaking such military obligations, including that accorded under international treaties.Footnote ⁷³ More interestingly, the prosecution argued that there were two classes of rights in law: those granted “for the protection of the Public at large” and those “solely for individual protection”. The prosecution argued that protections guaranteed by international conventions fell within the category of rights that could not be renounced. Any renunciation by individuals would be “contrary to public policy” because it “would weaken the effect of the Conventions generally” and as such would be “null and void”.Footnote ⁷⁴

After hearing the arguments of the defence, the court adjourned for fifteen minutes to consider the submissions of the prosecution and the defence. Upon reconvening, the court explained its opinion that it was not “empowered” to decide on its jurisdiction over the accused and would adjourn to consult the convening authority.Footnote ⁷⁵ When the court met again on 28 November 1946, it announced that it had considered counsels’ submission and the input of legal advisors from the convening authority. Based on “International Law and the Constitutional Laws of the Empire”, the court held that subjects were unable to renounce their allegiance during a state of war. Indeed, the court noted that “seduction of the captive for his allegiance” by a captor would be a war crime.Footnote ⁷⁶ Based on the court’s finding that any renunciation of allegiance was legally impermissible, the court continued with the trial.

A slightly different argument was attempted by the prosecution in the later case of Takashima Shotaro and another. On the trial’s fourth day, 29 January 1947, defence counsel submitted that the Indian victims had taken an oath of allegiance to the Japanese military and were no longer POWs at the time of the alleged crime. The war crime was therefore not within the court’s jurisdiction.Footnote ⁷⁷ The prosecution sought to cut off defence counsel’s argument by referring to Regulation 6 of the 1945 Royal Warrant and Regulations. This regulation prohibited jurisdictional challenges.Footnote ⁷⁸ The prosecutor stated that Regulation 6 provided a “complete answer” to the claims made by the defence.Footnote ⁷⁹ He further argued that the British Manual on Military Law prohibited occupying forces from forcing inhabitants of occupied territories from swearing allegiance to the former. If this was so for the civilian population, the prosecutor argued that POWs “certainly cannot” change allegiance.Footnote ⁸⁰

The court did not foreclose consideration of the allegiance issue based on Regulation 6, but the route chosen was no less satisfying. The court president explained that the present court had been involved in an earlier trial where “an identical submission” had been made by the defence.Footnote ⁸¹ He was in all probability referring to Ikegami Tomoyuki and others. He went on to explain that the convening authority had advised the court, in the earlier case, that the court had jurisdiction over the case and that victims “were not able legally to forswear their oath of allegiance to the British crown”.Footnote ⁸² The court president further noted that a “number of leading cases” had been discussed in that earlier case.Footnote ⁸³ The court later declined the prosecutor’s suggestion to recall a prosecution witness for the purpose of getting more information about the oath of allegiance.Footnote ⁸⁴

C. Post-trial Findings: Focus on the Accused Person’s Belief

As explained earlier in this paper, based on the 1945 Royal Warrant and Regulations framework, the decisions of guilt of these British military courts were not final until endorsed by the confirming authority.Footnote ⁸⁵ At the confirmation stage, the Department of the JAG in Singapore reviewed the cases and prepared advisory reports that were considered by the confirming officer. Both Ikegami Tomoyuki and others and Takashima Shotaro and another were not confirmed. The reports and messages issued by the Department of the JAG are particularly interesting.Footnote ⁸⁶

In the Department of the JAG review report on Ikegami Tomoyuki and others dated 13 February 1947, the Indian victims were described as being “originally Prisoners of War, but who had sworn allegiance to the Japanese”.Footnote ⁸⁷ It drew a distinction between international law and domestic law, acknowledging that, based on the “domestic law of England”, it was not possible for a British subject to renounce his or her nationality while on enemy or enemy-occupied territory during a war.Footnote ⁸⁸ Under this domestic law, an individual who did so would have committed an offence under the “English law of treason”.Footnote ⁸⁹ However, under “international law”, when an individual “voluntarily” joined enemy forces, disciplinary action against him pursuant to the enemy’s military law would not be considered a war crime.Footnote ⁹⁰ The review report went on to note that “there was no evidence” that the Indians had been “suborned” or coerced.Footnote ⁹¹ It is interesting to note that in Takashima Shotaro and another, the prosecution had argued that questions relating to allegiance should be referred to the convening authority, though this should only be done if the court decided that the change in allegiance had taken place “voluntarily”.Footnote ⁹²

In Takashima Shotaro and another, the Department of the JAG review report dated 15 April 1947 observed that the defence had raised the point that the victims were originally POWs but had taken oaths of allegiance to Japan and had thus lost their POW status.Footnote ⁹³ During the trial, the accused were questioned about the army order that had confirmed the non-POW status of these Indian victims. However, the prosecution’s application to recall a witness to give evidence on this point had been “unfortunately rejected” by the trial court “as a matter of law”.Footnote ⁹⁴ The Department of the JAG review report went on to explain that the accused could still argue the additional point that he “honestly believed that he was acting within his rights towards men whom it was his duty to treat as such”.Footnote ⁹⁵ In other words, the accused could raise the defence that he had honestly believed the victims had lost their POW status. The Department of the JAG review report advised non-confirmation of the trial court’s findings.

It should be noted that these questions on desertion and change of status were of sufficient concern for the Department of the JAG in SEALF to request advice from the JAG of the Forces in London on the issue. In a message dated 10 January 1947, F.G.T. Davis stated that he would be “grateful” for the “opinion” of JAG London on Ikegami Tomoyuki and others and another case tried at Jesselton, namely, Kamikura and another.Footnote ⁹⁶ In this message, Davis observed that there was a need to distinguish between those responsible for coercing the victims to join enemy forces and those accused of ill-treating victims. The person accused of ill-treating Indian victims should not be held culpable for any coercion undertaken by another in getting the Indian victims to change their allegiance. Specifically, the message noted that the accused may have been acting on “a bona fide mistake of fact” and may have truly believed that the victims were “persons properly subject to Japanese military law”.Footnote ⁹⁷

In a subsequent message dated 27 February 1947 from Davis to London, Davis referred to cases including Ikegami Tomoyuki and others as well as Takashima Shotaro and another, while highlighting the receipt of further information about the situation of Indian POWs during the war.

This information showed that there was a risk that the oaths of allegiance concerned were extracted from the victims under duress. Would this fact have impacted the trials? Based on the Department of the JAG review report on Takashima Shotaro and another dated 15 April 1947, such coercion by another would not automatically implicate the accused. The accused, who had not been involved in getting the victims to take these oaths of allegiance, could still have been under the mistaken belief that the victims had indeed switched their allegiance to Japan.

To sum up, the court in Ikegami Tomoyuki and others and the same court in Takashima Shotaro and another held that it was legally not possible for the Indian victims to switch their allegiance. In contrast, the post-trial Department of the JAG review reports drew a distinction between domestic law and international law, noting that, while the former prohibits the changing of allegiance during war, the latter does not. These Department of the JAG reports also pointed to the need to examine the accused persons’ state of mind as well as consider the belief of the accused. How then should we explain the different approach taken by the trial authorities in these cases? One possible explanation is that the trial authorities had made a legal error and had overlooked the points stated in the Department of the JAG review reports. Such a legal mistake would not be incomprehensible, as most judges and trial personnel did not hold formal legal qualifications. Another explanation is also possible. By deciding that it was legally not possible for the victims to switch their allegiance, the court in both cases avoided having to address thorny questions about what the Indian victims did, why they did what they did, and whether they had been coerced. These trials’ circumvention of the allegiance issue may be unsatisfying, but it is worth considering whether such judicial circumvention was not only reasonable but preferable, given international law’s lack of guidance on desertion and the politics involved. In many ways, war crimes trials are not the best forums to study and resolve politically charged factual conundrums.

A. The Need to Recognize the Reality of Desertions

The court in Ikegami Tomoyuki and others took an “impossible to renounce approach” to POW status. Based on this impossible to renounce approach to POW status, the Indian victims would not be able to renounce their POW status and the rights that came with it, even if they had voluntarily decided to desert the BIA to join the Japanese military or the Japanese-sponsored INA. The post-trial Department of the JAG review reports took a different approach. These reports focused on the voluntariness of desertion and side-switching. The Department of the JAG review report on Ikegami Tomoyuki and others noted that, under “international law”, when an individual who had “voluntarily” joined enemy forces, disciplinary action against the individual pursuant to the enemy’s military law would not be considered a war crime. This approach to desertion recognized the ability of POWs to give up their POW status through “voluntarily” switching sides during conflict.

The postwar 1949 Geneva Conventions and the 1977 Protocols do not have provisions that specifically deal with deserters. Presently, the Geneva scheme may be said to lean towards the impossible to renounce approach of the courts in Ikegami Tomoyuki and others and Takashima Shotaro and another. Article 7 of the Third Geneva Convention clearly states that POWs “may in no circumstances renounce in part or in entirety the rights secured to them by the present Convention”.Footnote ⁹⁹ This seems to foreclose the possibility of POWs switching sides and renouncing their POW status once they have obtained POW status. Commentators continue to debate how the Third Geneva Convention should be interpreted and applied to the situation of deserters. James Clause argues that the obligation to grant POW status under the Third Geneva Convention should not extend to those who voluntarily give themselves up, without the need to consider their intent.Footnote ¹⁰⁰ Sassòli argues that deserters could avoid POW status if they make known their deserter status to the detaining authorities at the “first opportunity”.Footnote ¹⁰¹ Therefore, a deserter should declare the intention to desert at the “first opportunity”, else he or she will not be able to do so during captivity.

This paternalistic approach taken by the Third Geneva Convention towards POW status seeks to minimize the risk that parties to the conflict will unduly pressure POWs in captivity to switch sides. Assessing the voluntariness of any desertion in times of conflict in captivity is difficult, as demonstrated by the facts of the Singapore Trials. If we were to apply the Third Geneva Convention’s prohibition of POW renunciation to BIA desertions, most BIA soldiers would not have been able to renounce their POW status during WWII. As explained above, when Singapore fell, the Japanese military took all Allied soldiers into captivity, at which point they obtained POW status. Subsequently, the Japanese military separated Indian soldiers from other Allied personnel, exhorting them to join the INA. BIA desertions took place over a long period of time.

The majority of Indian soldiers who switched sides did not voluntarily give themselves up to the enemy and would qualify as POWs. These Indian soldiers would not fulfil Sassòli’s “first opportunity” principle.Footnote ¹⁰² Nevertheless, as explained earlier, research shows that Indian soldiers were motivated by a variety of reasons to desert the BIA. Many Indian soldiers were indeed coerced, but many were also genuinely converted to the INA cause. It would be incongruous and inaccurate to consider the latter as POWs. There are of course many humanitarian advantages to the impossible to renounce approach. For example, it would avoid the situation where a Detaining Power wrongfully denies POW status to detainees on the basis that they are deserters when in fact they are not.Footnote ¹⁰³ Nevertheless, the BIA desertions case-study shows that voluntary desertions can be a reality. It should be noted that even when deserters are denied POW status, they remain protected under the 1949 Fourth Geneva Convention relative to the Protection of Civilian Persons in Time of War, so the loss of POW status does not mean that an individual is completely unprotected.Footnote ¹⁰⁴ Given the fact that voluntary desertions are a reality, it may be useful to rethink the Geneva Convention’s impossible to renounce approach to POW status.

B. Multiple Side-switching and the Military’s Treatment of its Own Troops

The Singapore trials considered in this paper show that desertion can in fact occur more than once and from more than one side during the same conflict. In Ikegami Tomoyuki and others, the Indian victims had deserted the BIA for the Japanese military. The Japanese military then caught the same Indians trying to desert the Japanese military, taking maps and other confidential information with them. It is reasonable to conclude that if true, these facts show that the Indian victims were trying to return to the British military.

The 10 January 1947 Department of the JAG review report discussed earlier also referred to the case of Kamimura and another, heard by a British military court in Jesselton, in former British Borneo or present-day Sabah, Malaysia.Footnote ¹⁰⁵ The accused in Kamimura and another were also accused of war crimes against Indian victims who were part of the Romu Tai or Special Labour Force working alongside Japanese forces in Kuala Belait, which is part of the former British Borneo or present-day Brunei. Similarly, the defence argued that these Indian victims were not POWs. In this case, the accused alleged that the Indian victims had disobeyed orders, had collected weapons, and were planning to fight for the Allies. The accused had launched a night counter-attack, during which fifty to seventy-two Indian victims were killed. The Department of the JAG review report noted the argument by the defence that the Indian victims were not POWs, having sworn allegiance to the Japanese. The report, however, went on to recognize the “further transition in the status of the deceased Indians” because, by arming themselves and declaring their intention to fight for the Allies, the Indian victims had “thereby effected their own liberation and re-acquired their original belligerent status”.Footnote ¹⁰⁶ The attack organized by the accused was therefore “an incident of war” and “not contrary to international law and usage”.Footnote ¹⁰⁷

Deserters can be in a particularly vulnerable position. On the one hand, deserters may be hostilely treated by POWs from their former armed forces, as these POWs had chosen to affirm their allegiance to the said armed forces even in captivity. Sassòli observes that, in the light of possible “tension” between these two groups, detaining powers should house deserters and POWs separately.Footnote ¹⁰⁸ On the other hand, deserters may be treated with suspicion by the forces to which they have deserted, even when the former have pledged their allegiance to the latter. Deserters may be harshly treated compared to other members of the force. This may be particularly so for deserters who have deserted more than once. The loyalty of these deserters may be held in doubt, and they may be suspected of being spies. International humanitarian law needs to be developed to specifically deal with the situation of deserters. Clear recognition and regulation of the status of deserters will minimize the risks they face.

BIA desertions highlight a more fundamental gap in international humanitarian law. This law does not attempt to regulate how states treat their own forces. States are presumed to have an interest in treating their own forces well. The Singapore trials discussed here show that this is not necessarily the case. It should be recalled that the treatment of a party’s own troops was an issue in Ikegami Tomoyuki and others and Takashima Shotaro and another. In both cases, the defendants and defence witnesses claimed that they did not see any difference between the Japanese military’s treatment of the Indian victims and Japanese soldiers. Research has confirmed that ordinary Japanese military personnel were themselves subject to extremely harsh treatment in the Japanese military.Footnote ¹⁰⁹ The Japanese military’s training and disciplinary practices were brutal and often inhumane, aimed at inculcating unquestioning obedience and endurance among military subordinates.

Should international humanitarian law be further developed to address how forces treat their own troops? This would require a fundamental rethinking of the law of armed conflict, which is beyond the scope of this paper. My aim here is to highlight gaps in international humanitarian law, as revealed by the facts and discussions in Ikegami Tomoyuki and others and Takashima Shotaro and another. In many ways, the heart of the issue in these Singapore trials was not really whether the Indian victims were to be considered POWs or persons associated with the Japanese military. Rather, the root problem stemmed from the Japanese military’s inhumane treatment of all those under its control.

C. Status Fluidity and Mistakes by Lower-ranking Military Personnel

These Singapore trials also highlight how difficult it is for lower-ranking individuals to independently assess whether those under their charge are POWs or deserters in chaotic conflict situations. The defendants all alleged that they had been told that the Indian victims were part of the Japanese Special Labour Force or Romu Tai. Among others, it will be recalled that the accused described how the Indian victims had their own commander, were given the relatively important responsibility of “guarding of the oil fields”, and had “a reasonable amount of freedom”.Footnote ¹¹⁰ These facts appeared to show that the Indians were treated as Japanese soldiers.Footnote ¹¹¹

It will often be difficult for mid-ranking or low-ranking military personnel to independently determine whether “deserters” are genuine POWs or fellow combatants. This is especially so in the messy circumstances of armed conflict. Deteriorating war conditions or resource shortages may result in POWs being given relative freedom or being given jobs that would normally be deemed too sensitive. Such situations may be coupled with widespread rumours or beliefs about desertion and side-switching. Language may be an additional barrier to communication. Some mid-ranking or low-ranking Japanese military personnel may truly have believed that these BIA soldiers were members of the Japanese military as opposed to POWs.

The Department of the JAG review reports considered here recognized the possibility that the defendants had genuinely made mistakes. As mentioned above, in Ikegami Tomoyuki and others, the Department of the JAG review report dated 13 February 1947 noted defence counsel’s argument that, even if the victims were POWs, the highest-ranking defendant had “acted in the bona fide belief” that the victims had “the status of Japanese soldiers” rather than POWs.Footnote ¹¹² As for the other defendants, the Department of the JAG review report noted the defendants’ claim that they believed they were implementing an “apparently lawful” order and were therefore not culpable.Footnote ¹¹³ There needs, however, to be a distinction made between the two types of mistakes highlighted in the Department of the JAG review report. The first mistake would commonly be known today as a “mistake of fact”, while the second deals with a mistake relating to superior orders. I will focus on the former in this paper as there have been previous studies on mistakes of laws relating to superior orders in British postwar trials.Footnote ¹¹⁴

A closer scrutiny of the “mistake of fact” arguments made by the defendants shows that at least some of these arguments may be better classified as what Kevin Heller describes as a “mistake of legal element”.Footnote ¹¹⁵ Specifically, defence counsel in Takashima Shotaro argued that the victims were not “pure” POWs, that they were “released” from POW status by the Japanese military, and that they were members of the “Indian Labour Unit”.Footnote ¹¹⁶ If this was so, the mistake made was not a straightforward mistake of fact about whether the Indian victims were POWs or Japanese military personnel. The mistake made by Japanese military personnel in this case is better described as a mistake of legal element or “MLE”.Footnote ¹¹⁷ It is noteworthy that, under the Statute of the International Criminal Court (hereinafter “the ICC Statute”), certain mistakes of fact or law may exclude criminal responsibility.Footnote ¹¹⁸ The exact scope of MLEs recognized at international criminal law is subject to academic debate, and Heller has persuasively argued for the ICC Statute to be clearly amended so that MLEs exclude criminal responsibility only if the mistake made was reasonable.Footnote ¹¹⁹ Reasonable knowledge of international humanitarian law is something that we should cultivate and expect from those engaging in armed conflict. However, such a reasonable understanding of the law is only feasible if international humanitarian law speaks clearly to soldiers of all ranks. By clearly recognizing and regulating desertion, international law may reduce mistakes being made by soldiers of lower ranks when assessing the status of deserters.