1. Since our policy is, in principle, to take no prisoners, we attempted to dispose of all of them. However, they continued to surrender in droves, first 1,000, then 5,000, then 10,000. We could not begin to disarm such a large number of soldiers. They had completely lost the will to fight, and simply followed after us. They did not seem to present any threat, but if a riot had erupted, we would not have been able to control them. Therefore, I had additional units brought in by truck, and assigned them to guard and escort the Chinese.

On the evening of the 13th, we were required to make countless trips with the trucks. But since this event occurred immediately after a battle (which we had won), we were not able to act expeditiously. The Operations Section was unbelievably busy because we had to dispose of far more prisoners than we had anticipated.

2. Later, I learned that Sasaki's unit alone had processed approximately 15,000 individuals, that a company commander with the garrison at Taiping Gate had processed approximately 1,300, that there was a concentration of approximately 7,000 near Xianhe Gate, and that enemy soldiers were still surrendering.¹

Readers were shocked by the first sentence in Nakajima's entry, "Since our policy is to take no prisoners, we attempted to dispose of all of them." In this chapter, we shall analyze this sentence, as well as the remainder of the entry, within the context of circumstances at the time.

During the assault on Nanking, division commander Nakajima was responsible for the area north of Zhongshan Gate (East Gate) extending to Taiping, Xuanwu, and Heping gates. The opposing force, the Nanking Defense Corps, had been so hastily augmented that its officers did not know their subordinates, and vice versa.

The moment Tang Shengzhi escaped, the Defense Corps began to collapse. The leaderless Chinese troops, now an unruly mob, attempted to flee Nanking, but the perimeter of the city was now under Japanese control. Having lost the will to fight, the Chinese began surrendering. On December 13, the day the gates were captured, thousands of Chinese soldiers surrendered to the 16th Division.

But a more careful examination raises several doubts. First, let us assume that "our policy is, in principle, to take no prisoners" means "shoot prisoners to death," and that "dispose of" means "execute on the spot."

1. If "our policy is, in principle, to take no prisoners" meant that prisoners were to be executed immediately, orders to that effect would have been issued not only to the 16th Division, but also to all other divisions. However, there is no execution order in the official records of any other division.
2. Division Commander Nakajima might have issued an execution order, at his own discretion, to the 16th Division and only that division. In that case, one would expect to find the order in the records of the 16th Division, but no such order exists.
3. According to 16th Division records, attempts were made to "dispose of all of [the prisoners]." That would mean that they were summarily and indiscriminately shot to death. When the first 10 or 12 were executed, the sound of gunfire would certainly have been audible. Would "1,000, then 5,000, then 10,000" prisoners have surrendered after hearing the gunshots?
4. There would have been corpses everywhere. Having seen those corpses, would hordes of Chinese soldiers have submissively followed their enemies?
5. Nakajima "had additional units brought in by truck, and assigned them to guard and escort the Chinese." If policy had been to execute prisoners on the spot, was it not his duty to do so?
6. Nakajima would have been disobeying orders in taking the action described in (5). Was he so determined to disarm the prisoners that he was willing to suffer the punishment meted out to those who defied orders? And why would he go to the trouble of making himself "unbelievably busy" requisitioning trucks and ordering emergency reinforcements?

If the policy in force at that time had been to shoot prisoners of war to death on the spot, Division Commander Nakajima would surely have made every effort to implement that policy. But he would also have mentioned his frustration at not being able to shoot 1,000, then 5,000, then 10,000 prisoners, because of their sheer numbers.

If we rewrite the sentence, following the sentence structure used in the journal entry, we have:

1. Since our policy is, in principle, to take no prisoners, we attempted to dispose of all of them (by shooting them to death). However, they continued to surrender in droves, first 1,000, then 5,000, then 10,000. We could not possibly shoot them all.

In other words, if we assume that "our policy is to take no prisoners" was an order to execute them, then it would have been logical for Nakajima to write that his men attempted to dispose of all of them by shooting them to death, but could not because there were so many of them. But that is not what he wrote. What he did write was: "We could not begin to disarm such a large number of soldiers."

If we assume an execution order was indeed issued, this sentence no longer makes sense, which means that it is mistaken to conclude that Nakajima was attempting to execute the prisoners. However, if we assume that no execution order was issued, all of the aforementioned doubts disappear.

Then, what was meant by "take no prisoners?" There are three possibilities. The first possibility (and the prevailing interpretation) was that policy dictated the shooting to death of prisoners of war. If that had been the case, Division Commander Nakajima would have written: "Since our policy is to take no prisoners, we attempted to shoot all of them to death. However, they continued to surrender in droves, first 1,000, then 5,000, then 10,000. We could not begin to shoot such a large number of soldiers." However, Nakajima did not write "We could not shoot all of them to death." Therefore, it was not Japanese military policy to execute prisoners on the spot.

In that case, was it policy to take prisoners? The answer is no, since the entry clearly states "our policy is, in principle, to take no prisoners."

The only remaining possibility is that prisoners were to be released, because the only plausible actions to be taken vis a vis soldiers who surrendered on the battlefield were execution, incarceration, or release.

Therefore, "our policy is, in principle, to take no prisoners" means "Our policy is, in principle, to take no prisoners, but to release enemy soldiers who surrender, after having disarmed them."

From the locution "We could not begin to disarm such a large number of soldiers," we sense Nakajima's frustration in not being able to implement policy, i.e., to take no prisoners, but to disarm and release them. Nakajima could not release such a large number of armed soldiers. Neither could he execute them, since he would have been disobeying orders, which stipulated that they be observed and their presence reported to the Operations Section. However, Nakajima lacked sufficient manpower to guard the prisoners, and was forced to dispatch trucks to transport personnel. That is why he was so busy on the evening of December 13, procuring personnel and trucks, and escorting the prisoners, something he had certainly not anticipated.

The gates of Nanking had fallen but, inside the city, Chinese troops had still not surrendered. The Japanese needed to maintain safety inside the city. Outside Nanking, near Zijinshan, fleeing Chinese soldiers were attacking Japanese units. Some of them were embroiled in the worst hostilities they had experienced since Shanghai. It is not surprising that the division commander and the Operations Section were "unbelievably busy."

It is consistent with the pedagogical example dating back to 1933, which states how prisoners of war are to be treated: "With the exception of special cases, prisoners of war may be released where they were captured, or after having been moved to another location."⁵ It is also in keeping with the notice issued by the Vice-Minister of War in October 1937, which states that Japanese military personnel are to honor the "prohibitions" provided by the regulations annexed to the Hague Convention, and to refrain from attacking "an enemy who has ... surrendered at discretion."⁶

"Battle Instructions" from 13th Division Headquarters directed that prisoners were not to be shot, but disarmed and released. It was the policy of the Shanghai Expeditionary Force to release enemy soldiers who had surrendered. Division Commander Nakajima's policy of disarming and releasing prisoners, and Onishi's testimony about releasing prisoners are completely consistent with Shanghai Expeditionary Force policy.

Later, I learned that Sasaki's unit alone had processed approximately 15,000 individuals, a company commander with the garrison at Taiping Gate had processed approximately 1,300, that there was a concentration of approximately 7,000 near Xianhe Gate, and that enemy soldiers continued to surrender.⁷

There are three possible interpretations of "process." The first is the "execute all enemy soldiers who surrender." Let us suppose that 15,000 prisoners were executed between Zijinshan and northeastern Nanking, 1,300 at Taiping Gate, and 7,000-8,000 near Xianhe Gate (more accurately, near Xianhemenzhen). If that had been the case, a burial report or a statement from a witness would have been required. But there were no corpses and, therefore, no executions.

The second is "execute only rebellious prisoners." Article 8 of the Rules Annexed to the Hague Convention Respecting the Laws and Customs of War on Land states: "Any act of insubordination justifies the adoption towards [prisoners of war] of such measures of severity as may be considered necessary."8 Therefore, the execution of such prisoners would have been in keeping with international law. However, even if that had been the case, it is not likely that many prisoners were executed.

The third and last interpretation is "take no prisoners, but disarm and release enemy soldiers who surrender." This interpretation is supported by the war journal of Major-General Iinuma Mamoru, Shanghai Expeditionary Force chief of staff, which contains two crucial entries, both dated December 14. The first is "From an aircraft, I observed two groups of approximately 1,000 prisoners of war being moved from east Nanking toward Xiaguan."

The second reads: "Report received to the effect that four columns of prisoners of war, eight kilometers long, were observed from aircraft, as they were being escorted north of Nanking."9

The second entry probably refers to several thousand members of the Nanking Defense Corps (which had sworn to defend Nanking to the death) who surrendered, "waving white flags," in the vicinity of Xianhemenzhen.10 According to Sasaki Motokatsu's account in Field Post Office Flag and Shanghai Expeditionary Force Sakakibara Kazue's eyewitness account, on or about December 17, a large number of prisoners were escorted to the Central Prison in Nanking (No. 1 Penitentiary) and incarcerated there.11

If their captors had intended to execute them, they would have done so immediately. It was not necessary for the Japanese to move them into Nanking, and it was even dangerous, since the prisoners might have instigated a revolt. Also, when Major-General Iinuma received a report to the effect that the prisoners had been sighted, he would have ordered a halt to the convoy and had them executed.

The fact that he did not demonstrates that Japanese military policy was to disarm and release prisoners of war. It is likely that Chief of Staff Iinuma dispatched a plane to confirm that units under his command were adhering to policy by escorting the prisoners.

It is true, however, that on December 13, during the pitched battle at Zijinshan, quite a few enemy soldiers were killed subsequent to their surrender. On December 14, when hostilities had virtually ended, prisoners of war were transported and interned. A difference of one day was also the difference between life and death.

Thus, the word "process" had two meanings: (1) release prisoners of war after disarming them and (2) execute antagonistic prisoners. But when the Japanese transported soldiers who had surrendered, they really did intend to release them. Prior to the assault on Nanking, the policy was not to release prisoners immediately, and that policy was still in force. The gates had been occupied, but pitched battles were still being fought to the east and southeast of the city. The Japanese did not dare disarm and release soldiers who surrendered on the battlefield because of the risk of their rejoining their comrades. That is probably why prisoners were transported to the north, as Major-General Iinuma wrote in his journal.

Go to Chapter 7
Back to Index