The Nanking Atrocity: An Interpretive Overview


October 1, 2007

The Nanking Atrocity: An Interpretive Overview
The Nanking Atrocity: An Interpretive Overview
The Nanking Atrocity: An Interpretive Overview

Fujiwara Akira

Translator’s Introduction

This article is slightly adapted from a chapter by the late Fujiwara Akira, an emeritus professor at Hitotsubashi University until his death in 2003, which first appeared in Bob Tadashi Wakabayashi, ed., The Nanking Atrocity 1937-38: Complicating the Picture (New York and London: Berghahn Books, 2007). Fujiwara wrote one of two introductory chapters to this volume about the Nanking Massacre, the seventieth anniversary of which will be observed in December.

In this essay, Fujiwara provides a concise narrative of Japan’s decision to escalate the “China Incident” into a full-scale war by July 1937. This ultimately led to an assault on China’s wartime capital of Nanking by imperial armed forces, who captured it in December. Fujiwara also gives a trenchant, critical account of the Nanking Massacre (a.k.a. “the Rape of Nanking”), plus an admittedly partisan yet nonetheless fair analysis of right-wing views in Japan today that downplay or deny this atrocity. On this last point, Fujiwara argues that Japanese deniers and nationalistic revisionists seek to build a public consensus that will allow their nation to re-emerge as a military power uninhibited from waging future wars based on putatively unwarranted feelings of guilt about the past.

The English translation of Fujiwara’s chapter, completed in 2002, seems prescient in the light of subsequent events in violation of Article IX of the postwar Constitution: 1) Japan dispatched armed troops to Iraq in January 2004 and extended their mission in December of 2004 until July 2006. 2) Self-Defense Forces have been providing logistical support to US military forces in the form of fuel supplies despite well-founded allegations that these are being redirected to Iraqi battlefields. 3) Earlier this year, former Prime Minister Abe sought to “reinterpret” Article IX into non-existence based on proposals from a panel of advisors hand-picked for precisely that purpose.

One major article of postwar leftist faith is that Japan must never again become a “normal nation”– in the sense of exercising its sovereign right to wage war– because imperial armed forces at Nanking and elsewhere proved that they could not be trusted to behave in a lawful, humane, and responsible manner. The present essay constitutes Fujiwara’s final testament to this article of faith, prepared for an international readership.

Note that, as translated by Wakabayashi for Japan Focus, this essay omits endnotes and macrons over long vowels in Japanese terms. A small number of Chinese terms– such as Nanking, Amoy, Hsiakwan, and Kwantung Army– are romanized in Wade-Giles because they have found their way into the English language in that form. All other Chinese terms are rendered in pinyin. Bob Wakabayashi


Modern Japan’s aggression against China began with the Meiji-Qing or First Sino-Japanese War of 1894-95, and continued with the Twenty-one Demands of 1915, the Shandong Expeditions of 1927-28, and the Manchurian Incident of 1931-3. But an all-out war of aggression began with the 7 July 1937 armed clash at Marco Polo Bridge outside Beijing. Culpability for turning that minor skirmish into an all-out war lay with Japan– primarily the imperial government and central army authorities. Although a local truce settled the affair on 11 July, Prime Minister Konoe Fumimaro’s government expressed “grave resolve” in passing a cabinet resolution to send more troops on that same day. Konoe, an imperial prince, flaunted his regime’s belligerence by inviting the media to his official residence and calling on them to foster national unity. Based on this cabinet resolution, commanders hastily sent two brigades from Manchuria and a division from Korea to northern China, the General Staff prepared to send three divisions from Japan, and the Army Ministry halted all discharges. At that time, two-year recruits received an early discharge in July– before their active duty actually ended– to go home for peak months of farm work when the labor of young men was sorely needed. By rescinding this provision, the government showed that Japan was gearing up for war in earnest.

Prime Minister Konoe Fumimaro in 1937

Japan’s hard line created a sense of crisis in China. Chiang Kai-shek of the Guomindang (GMD) Nationalist government met with Zhou Enlai of the Chinese Communist Party (CCP) on 17 July to discuss stepped-up efforts for a united front, and Chiang made a speech on the need for resolve in resisting Japan. The Chinese people’s will to resist heightened as two more armed clashes broke out in the north China tinderbox. By 27 July, reinforcements from Korea and Manchuria had arrived, as did naval air force units, and Emperor Hirohito issued Army Chief of Staff Order 64. It read: “Along with its present duties, the China Garrison Army (CGA) shall chastise Chinese forces in the Beijing-Tianjin area and pacify [i.e., occupy] strategic points.” The emperor used the term chastise that Prime Minister Konoe later made famous. On 27 July, the government decided to send reinforcements from Japan proper. Chief of Staff Order 65, issued by the emperor, called for sending three divisions and mobilizing another 209,000 men plus 54,000 horses. Real fighting began on 28 July with a general offensive in the north that saw imperial troops occupy Beijing and Tianjin.

This course of events was the converse of that which began the Manchurian Incident. In September 1931, the imperial government and central army authorities had wanted to settle that conflict quickly whereas field armies were intent on expanding it. Now, in July 1937, it was the government in Tokyo that escalated the war by sending massive reinforcements to northern China even though field armies had reached a settlement on 11 July. Ishiwara Kanji, Chief of the General Staff Operations Division, reversed his hawkish views of Manchurian Incident days, and was now an exception among central army authorities in opposing the extension of operations to China. More typical of that group was Army Minister Sugiyama Hajime, who sided with Prime Minister Konoe, Foreign Minister Hirota Koki, and other civil government hawks. Even so, the initiative for future army decision making would ultimately lay with local commanders who zealously pushed for escalation despite their gravely flawed grasp of conditions in China. Blind to the patriotism forging national unity there, they persisted in disparaging the Chinese military and people in the belief that “one telling blow,” or quick decisive victory, would make the enemy sue for peace.

In August 1937, naval marine units took the war to Shanghai on the pretext of protecting Japanese civilians against popular Chinese unrest. Army hawks dismissed opposition from more cautious elements such as Ishiwara Kanji, and boldly extended the scope of operations from northern to central China. Hirohito, as he himself would relate in 1946, sought to expand the war at this time by sending even more units from Manchuria. He berated Ishiwara for weakness and was instrumental in transferring units from Qingdao in northern China to Shanghai. Thus the central government started what became a full-scale war by dispatching huge army units, but offered no justification worthy of the name, saying only that imperial forces would “chastise the unruly Chinese”– a slogan that Konoe issued in lieu of formally declaring war. There were 3 main reasons for pursuing this conflict as an “incident” rather than as a war: (1) Even at this late date, army and government leaders felt convinced that “one telling blow” would end it; they did not dream that a major, long-term conflict would result. (2) Japan had no compelling reason for war. “Chastise the unruly Chinese” was hardly a war aim that would whip up popular support at home. (3) With the premiership of Hirota Koki from March 1936 to February 1937, the army and navy had begun pursuing armament expansion programs that relied on imports of strategic matériel from the United States, a neutral power. Japan could not go on importing these key items easily under international law if it formally became a belligerent state by declaring war on China.

Central army leaders in Tokyo had no plan to attack the capital of Nanking when they dispatched troops to Shanghai in August; in this, they differed from Matsui Iwane and Yanagawa Heisuke, who later led the assault on Nanking. Instead, leaders in Tokyo expected a quick local settlement like that which had ended the Shanghai Incident of January to May 1932. This time, a Shanghai Expeditionary Army (SEA), or Shanghai Expeditionary Force (SEF), was assembled on 15 August 1937 under Matsui’s command. It had strictly limited orders: “to protect imperial subjects by destroying enemy forces in and around Shanghai and occupying strategic points to the north.” Nanking, it bears noting, is roughly 300 kilometers west of Shanghai. The SEA’s initial strength was hastily set at two divisions, and a heavy artillery unit joined within two weeks. Three and a half more divisions joined in September, and one more in October. Thus the SEA comprised the Third, Ninth, Eleventh, Thirteenth, Sixteenth, One Hundred-first, and One Hundred-sixteenth divisions. The Tenth Army was formed under Yanagawa’s command in October. It comprised three and a half divisions: the Sixth, Eighteenth, One Hundred-fourteenth, plus part of the Fifth. This Tenth Army was not supposed to attack Nanking either. Its mission, like that of the SEA, was to destroy Chinese armies and protect Japanese nationals in the Shanghai area– nothing more.

The imperial army’s foremost priority throughout the 1930s was to prepare for war with the Soviet Union. Army leaders had no wish to commit large forces in China for the long term, and most were convinced that this “incident” would end after they scored one major victory. But events at Shanghai shocked them. Shells ran perilously low. By 8 November, casualties had skyrocketed to 9,115 killed and 31,125 wounded. Reinforcements, which had never been anticipated, were sent repeatedly. The Third and Eleventh divisions, for example, had to be totally replenished. Army leaders shifted the war’s main theater from northern to central China in October and the Tenth Army landed behind Chinese lines at Hangzhou Bay on 5 November. Only that daring move broke the bloody stalemate at Shanghai, but Chinese units beat a hasty full retreat to avoid encirclement and annihilation. Japan, then, did not deliver the “one telling blow” to wipe out enemy forces, and thus could not achieve victory in this “incident.”

On 7 November, two days after the Tenth Army landed, it and the SEA combined to form a Central China Area Army (CCAA) under Matsui’s overall command, with Imperial Prince Asaka Yasuhiko taking over the SEA. At its height, this newly-formed CCAA numbered an estimated 160,000 to 200,000 men. The reorganization signified that Japanese forces were not just on an expedition to Shanghai, but would operate in a broader “central China area.” Even so, the CCAA was still an “impromptu amalgamation,” not a formal battle formation, as reflected in its mission. Its orders read: “Destroy enemy forces in the Shanghai area, break their will to fight, and thereby bring an end to the conflict.” The Chief of the General Staff also stipulated a line of demarcation: “in general, east of the Suzhou-Jiaqing line.” In other words, the CCAA was ordered to remain in the area east of Lake Tai; that is why it received no support-and- supply units. Also, six of the CCAA’s ten and a half divisions were “special divisions,” limited in maneuverability, weak in firepower, and manned by second- or third-pool reservists hastily assembled. They were not officers and men on the active list, in the fighting prime of their early twenties. Their abrupt recall to active duty came in their mid- to late-thirties, or even early-forties– long after they felt their military obligations were over and they had returned to civilian life as bread-winners. Hence, morale and amenability to military discipline were often poor.

When the entire Chinese army began to retreat, the CCAA ignored orders and gave chase westward toward Nanking. Eguchi Keiichi cites SEA Chief of Staff Iinuma Mamoru’s diary to show that, as early as 18 August, SEA commander Matsui Iwane already aspired to capture the enemy capital although central army leaders had no such plans, and even before the CCAA came into being. Matsui, disgruntled by the narrow scope of SEA operations, had to be chastised: “orders for military operations are no different from imperial rescripts; it is impudent to criticize these.” But later that same day Matsui openly declared: “We must resolve to order troops into action as needed based on our traditional spirit of ‘instant engagement, instant victory’ by shifting our main forces from northern China to Nanking. We can debate the issue of where best to deliver the knock-out blow, but right now we absolutely must make Nanking our main target.”

Gen Matsui Iwane entering Nanking on December 17, 1937

After the Chinese flight began in November, frontline troops came to share the newly promoted CCAA commander Matsui’s aspirations; they craved the glory of being first to enter Nanking. Their egregious forced marches, exacerbated by the absence of support-and-supply units, meant that the rank-and-file had to rely on plunder to survive en route. On 20 November, Imperial Headquarters (IH) was set up for the first time since the 1904-05 Russo-Japanese War– a decisive step both strategically and symbolically. Belittling the China war as a mere “incident,” yet unable to win it, Japan had no choice. Given de facto wartime conditions of mass troop deployment and naval support, the coordinating of the two services’ chains of command required an IH under the 1889 Imperial Constitution. Unlike the IH in wars before Hirohito’s reign, however, this one was a purely military body in 1937. No civilian cabinet member, not even the prime minister, could join its deliberations. Instead, an ad hoc liaison council handled communications between the government and IH to ensure that cabinet acts of state conformed with the emperor’s supreme command. Thus began the most enormous, expensive, and deadly war in modern Japanese history– one waged without just cause or cogent reason.

The Atrocity Delimited

On 24 November 1937, IH admitted reality and rescinded its first line of demarcation, that from Suzhou to Jiaxing east of Lake Tai; only then did it begin to think seriously about attacking Nanking. IH now set up a second line of demarcation cutting across Lake Tai from Wuxi to Huzhou, behind which forces would regroup before advancing further. But frontline units ignored this line too, and pressed their attack. On 1 December, the emperor’s Army General Staff Order 7 converted the stopgap CCAA from an “impromptu amalgamation” into a formal battle formation. On the same day, the emperor’s Army General Staff Order 8 read: “CCAA commanders shall assault the enemy capital of Nanking with support from the navy.” Thus, formal orders to attack Nanking came down only on 1 December. Based on Army General Staff Order 8, the CCAA commanded that: (1) the SEA launch operations on 5 December with its main force ready to move toward Danyang and Jurong while a subsidiary force attacked the enemy rear on the north shore of the Yangzi; and (2) the Tenth Army start operations on 3 December with its main force ready to move toward Lishui and a subsidiary force, toward Wuhu. In fact, the SEA was already well past Zhangzhou on 29 November. It occupied Danyang on 2 December– something not scheduled to happen until the seventh. Likewise, the Tenth Army had already taken Guangte on 30 November. Its commander, Lt. Gen. Yanagawa Heisuke, proclaimed at the head of his troops, “I will press the attack on Nanking as I deem fit.” Thus, neither the SEA nor the Tenth Army bothered to wait for orders. The aim behind Army General Staff Order 8 was not to make an all-out rush for Nanking. Both CCAA armies were to advance along a broad front, regroup, encircle Chinese defense forces, and annihilate them. But glory-hungry, frontline units lusted to be first in the enemy capital and staged a mad dash for it. Thus the attack on Nanking, like that on Shanghai, was out of control from the start.

In determining the number of Chinese victims in the Nanking Atrocity, we must first define the event’s time span and area. The SEA advancing from Shanghai, and the Tenth Army after landing at Qinhanwei in Hangzhou Bay, repeatedly indulged in rape, arson, plunder, and mass murder. In that sense, Yoshida Yutaka and Honda Katsuichi are correct when they argue that any study of the Atrocity must include these vicious acts en route to Nanking, not just those in and near the city. Massacres took place all the way from Shanghai to Nanking, so in principle, all persons killed en route should enter into the total. But insofar as we call this the Nanking Atrocity– as opposed to those elsewhere– I delimit the event as lasting from 1 December 1937, when IH and the CCAA issued orders to attack Nanking, to 5 January 1938, when the imperial army felt that a measure of order had returned to the city. Area-wise, the Atrocity took place on both sides of the Yangzi, west of the Zhangzhou-Guangde line, where the two armies met on 1 December. The SEA and Tenth Army main units advanced toward the capital parallel to the major roads and rail lines, while subsidiary units forded the Yangzi to march along its far bank under artillery support from the Eleventh Battle Fleet. Having let Chinese armies evade annihilation at Shanghai, the Japanese now sought to exploit geography to preclude a second escape by pushing them toward the Yangzi, which wrapped around and behind the city of Nanking. Ill-advised Chinese plans to defend the capital at all costs, despite detrimental topographic conditions, played into Japanese hands. These flawed defense plans were a factor that contributed to the Atrocity, as did the fact that Nanking had a huge civilian population, which would be trapped alongside soldiers inside city walls.

Japanese Strategic Blunders

The Army General Staff had repeatedly worked out precise, detailed plans for war with the Soviet Union, but never seriously thought about fighting the Republic of China, led by Jiang Jieshi’s GMD regime after 1927, as a national entity. Thus, Japan had no long-range blueprint to conquer all of China. The General Staff revised its “Guidelines for Defense of the Empire,” its tactical handbooks, and its troop strength levels in 1918, when it named “Russia, the U.S., and China” hypothetical enemies. A second revision took place in 1923, when the list read, “the United States, Russia, and China.” However, just placing China on the list did not mean that the Army General Staff made careful plans to wage full-scale war against it as a unified nation. Strategic thinking remained tied to the notion that China was a collection of warlord satrapies; thus, Japan needed only to occupy key areas as tactical needs dictated. A third revision of “Guidelines for Defense of the Empire” appeared in June 1936. The text of its section on “Tactics toward China” has been lost. But according to a later account by Shimanuki Takeharu, this simply called for the army to attack and destroy enemy forces in northern and central China so as to occupy key points there, and for the navy to support army operations by sinking the Chinese fleet on the Yangzi and along the seacoast.

Based on these June 1936 “Guidelines,” the General Staff in August 1936 drew up a plan titled, “Strategy for China: 1937.” This stated that five hypothetical divisions in northern China might be bolstered by three more for use in the five provinces of Suiyuan, Rehe, Shandong, Hebei, and Shaanxi. Three hypothetical Ninth Army divisions in central China might occupy Shanghai, and a Tenth Army might be formed to land at Hangzhou Bay. These Ninth and Tenth armies might march on Nanking and seize and occupy the strategic Shanghai-Hangzhou-Nanking triangle, but in that case, no operations should take place elsewhere in China. In southern China, the plan would be for the deployment of one division, whose main force might occupy Fuzhou and subsidiary forces might occupy Amoy and Swatow. At this time, Japan’s China Garrison Army, stationed in the Beijing-Tianjin area of northern China, had only recently risen in strength from 1771 to 5774 men. This “Strategy for China: 1937” was the Army General Staff’s first specific plan to occupy parts of northern China and to attack Nanking. But, it must be stressed, the “Strategy” was drawn up only in response to the Hirota regimes’s “Second Guidelines for Settling the China Situation” of August 1936. This latter document envisioned a “Detachment of North China” that entailed the army’s help to create an anti-Communist buffer zone of collaborator regimes. Likewise, the Army General Staff drew up plans for sending troops to Shanghai in central China, but only in response to stiffened Chinese defenses. Those plans, too, did not derive from Japanese initiative.

The “Strategy for China: 1937” drafted in August 1936, the creation of a SEA in August 1937, and actual assault on Nanking by the CCAA in December 1937, all stemmed from a grave misreading of affairs in China, which was fast moving toward national consolidation fed by anti-Japanese nationalism. The December 1936 Sian Incident, which saw the kidnapping of Chiang Kai-shek by his own generals, cleared the way for a second round of GMD-CCP cooperation culminating in a united front; it did not split China as the Japanese had hoped. These momentous events spawned a nationwide commitment to resist Japanese aggression after the July 1937 Marco Polo Bridge Incident, which led to full-scale hostilities. In sum, despite naming China a hypothetical enemy, Japanese military leaders neither deemed it a unified nation-state nor did serious strategic planning on that premise. This arrogance stemmed from the outdated notion that Japan need only occupy this or that key area in a strictly tactical fashion. Even in drawing up its “Strategy for China: 1937”– when the Army General Staff finally started thinking about sending troops to central China and Nanking– no one dreamed that this move would lead to a full-scale, long-term war. Japanese strategists could understand the broad political ramifications of attacking a national capital; that is why imperial army units sped lemming-like over the brink after they saw the Chinese army beat a retreat from Shanghai.

Such considerations shed light on three major underlying causes of the Nanking Atrocity. First, contempt for China as a modern nation led to a deficient concern for applying international law toward it. Just as serious fighting in northern China began, an undersecretary in the Army Ministry sent a notice dated 5 August 1937 to the China Garrison Army’s Chief of Staff: “It is inappropriate to act strictly in accordance with various stipulations in ‘Treaties and Practices Governing Land Warfare and Other Laws of War’.” Similar notices went out to other units as well. The message can only be construed as: “there is no need to obey international law.” Second, this overweening attitude diluted concern for protecting Chinese civilians, as well as foreign diplomats and residents, from the horrors of war. The CCAA was formed haphazardly on 7 November 1937. Since it was not supposed to move far west of Shanghai, it had no supply-and-support units to provision troops, who could only rely on plunder to sustain themselves en route to Nanking. This increased their frequency of contacts with, and opportunities for violence toward, civilians. The SEA and the Tenth Army had no liaison staff or units trained in diplomacy; so those armies’ relations with Japanese diplomatic officials in China were bad, to say the least. Troops viewed diplomats as a thorn in their side; diplomats who tried to stop army brutalities exposed themselves to danger. A third and related underlying cause of the Atrocity lay in the CCAA’s disregard for upholding troop discipline and morality. It had no specialized military police (MP) units, and the few individual MPs who were on hand could not possibly maintain order. As one attached to the Tenth Army bewailed, “With less than 100 of us to control 200,000 men in several divisions, what could we do?”

POWs and the Assault

The 1 December order to attack Nanking prescribed a third line of demarcation; troops were to regroup and consolidate a front from the Mobang Hills to Lishui. From there, they would face off against Chinese defenders. This did nothing to deter the SEA’s Ninth and Sixteenth divisions or the Tenth Army’s Sixth and One Hundred-fourteenth divisions, which raced abreast of each other, intent on being the first to scale the walls of the enemy capital. The CCAA laid down a fourth line on 7 December, a “line of readiness,” before the final push for Nanking; and, it issued instructions, “Essentials for Assaulting Nanking,” listing orderly procedures for taking the city. CCAA commander Matsui Iwane advised Chinese commander Tang Shengzhi to surrender the capital on 9 December, but received no reply by noon on 10 December. Matsui therefore ordered the attack to resume at 1:00 p.m. Frontline units ignored last-ditch efforts to contain the conflict, including those by the International Committee (IC), which established the Nanking Safety Zone (NSZ). The SEA’s Sixteenth division began to assault Purple Mountain just outside Nanking to the east on the tenth, and reached its summit by the twelfth. The SEA’s Ninth division rushed toward the city from the southeast. Some of its units reached Guanghua Gate in the early hours on the ninth, but met fierce resistance for several days. The SEA pulled one regiment, the One Hundred-third Brigade, or Yamada Detachment, in the Thirteenth division, then on standby at Zhenjiang, and ordered this brigade to advance through Wulongshan and Mufushan on the right wing of a beefed-up Sixteenth division. On the eleventh, the SEA ordered a regiment in the Third division, then held in reserve, to augment the Ninth division’s left wing as an advance raiding unit. The Tenth Army’s Sixth and One Hundred- fourteenth divisions, advancing in parallel from the south, broke through Chinese front lines on 8 December and attacked the Fukuo encampment at Rain Flower Heights south of the city on the tenth. The Tenth Army then ordered the Hiroshima Fifth Division’s Kunisaki Detachment, on loan from its Ninth Brigade, to ford the Yangzi near Cihechen (Taiping) and to advance on Pukou. The Tenth Army also ordered its Eighteenth division, which had captured Wuhu on the tenth, to concentrate its forces on standby for an assault on Hangzhou.

Thus the Ninth and Sixteenth divisions marched toward Nanking from the east; and the Sixth and One Hundred-fourteenth, from the south. On 13 December, the Thirteenth division’s Yamada Detachment (or One Hundred-third Brigade) arrived from the north, and the Sixth division’s Forty- fifth Regiment, from the south. Both had orders to plug Chinese escape routes between the city’s western wall and the Yangzi. For good measure, the Eleventh Battle Fleet patrolled the Yangzi, and the Fifth division’s Kunisaki Detachment advanced on its far bank to cut off Chinese troops trying to escape across the river. All told, CCAA forces assaulting Nanking numbered 57 infantry battalions or, as Kasahara estimates, between 160,000 and 200,000 men. They received artillery support from the seventeen-ship Eleventh Battle Fleet. Encirclement of the city would be complete by the early hours of 13 December. Across the lines, commander Tang Shengzhi did not order a retreat until 5:00 p.m. on the twelfth. Tang was the very first to flee, crossing the Yangzi at 6 p.m. Tens of thousands of his troops– until then trapped in the city with orders to defend it at all costs– fled in chaos as their command structure totally broke down.

Japanese units learned of this retreat on the morning of the thirteenth. Skirmishes broke out in many areas as small groups of Chinese troops outside the city, now lacking a chain of command, desperately tried to slip past advancing Japanese forces. Then the surrendering began. Most of the Chinese troops still inside Nanking rushed to escape helter-skelter through Bajiang Gate, which led to the Hsiakwan wharf area. From Hsiakwan, they hoped to cross the Yangzi by boat, by raft, or by clinging desperately to scraps of lumber, or they madly ran up and down the riverbank, only to encounter Japanese forces sent to cut them off. Huge numbers of Chinese troops became prisoners of war (POWs) on the thirteenth and fourteenth at Hsiakwan, Mufushan, Jiangdong Gate, and Xiaohua Gate. With no avenue of escape, Chinese soldiers lost all will to fight. Despite trying to surrender in droves, most were killed in the pell-mell of battle. Sixteenth division commander Nakajima Kesago’s diary entries on 13 December describe the confusion:

We see prisoners everywhere, so many that there is no way we can deal with them…. The general policy is: “Accept no prisoners!” So we ended up having to take care of them lot, stock, and barrel. But they came in hordes, in units of thousands or five-thousands; so we couldn’t even disarm them…. Later I heard that the Sasaki Unit [the Thirtieth Brigade] alone disposed of about 1,500. A company commander guarding Taiping Gate took care of another 1,300. Another 7,000 to 8,000 clustered at Xianho Gate are still surrendering. We need a really huge ditch to handle those 7,000 to 8,000, but we can’t find one, so someone suggested this plan: “Divide them up into groups of 100 to 200, and then lure them to some suitable spot for finishing off.”

Thirtieth Brigade commander Maj. Gen. Sasaki Toichi wrote in his diary on 13 December:

The number of abandoned enemy bodies in our area today was ten thousand plus thousands more. If we include those [Chinese] whose escape rafts or boats on the Yangzi were sunk by fire from our armored cars, plus POWs killed by our units, our detachment alone must have taken care of over 20,000. We finished the mop-up and secured our rear at about 2:00 p.m. While regrouping, we advanced to Heping Gate. Later, the enemy surrendered in the thousands. Frenzied troops– rebuffing efforts by superiors to restrain them– finished off these POWs one after another. Even if they aren’t soldiers [e.g., medics or priests], men would yell, “Kill the whole damn lot!” after recalling the past ten days of bloody fighting in which so many buddies had shed so much blood.

Such diary entries by division and brigade commanders allow us to gauge the chaos of battle, and the extent of the slaughter of POWs. Official battle reports exist for some Japanese units that vividly describe how they handled POWs. These sources are housed in the National Institute for Defense Studies, Military History Department Library in Tokyo. Most are included in the source collection, Nankin senshi shiryo shu, published by the Kaikosha, a fraternal society of former imperial army officers and conservative revisionists. For instance, the Thirty-third Regiment’s battle report for 10-14 December has a “Booty List” with an entry for POWs: “fourteen officers plus 3,082 NCOs and troops.” Under the column “Remarks,” it says, “disposed of POWs.” As a rough number of “abandoned enemy corpses,” it lists “220 on the tenth, 370 on the eleventh, 740 on the twelfth, and 5,500 on the thirteenth, for 6,830 all told.” But, this battle report goes on, “the figure for 13 December includes defeated enemy troops whom we executed.” The Thirty-third Regiment took 3,096 POWs on 13 December and “disposed of” them. The Sixty-sixth Regiment’s First Battalion battle report says that it took “1,657” POWs outside Rain Flower Gate from the afternoon of the twelfth to the morning of the thirteenth. The appendix in the “Thirty-eighth Battalion Battle Report 12,” says that its Tenth Company took “7,200” POWs on the morning of the fourteenth near Xiaohua Gate. Other battle reports listing numbers of POWs taken are those for the Kunisaki Detachment’s Forty-first Regiment Twelfth Company, which took “2,350” at Jiangxinzhou from the night of the fourteenth to the morning of the fifteenth; the Thirty-third Regiment Second Battalion, which took “about 200” at Lion Hill on the fourteenth; and the Seventh Regiment, which took “6,670” in the Naking Safety Zone (NSZ), set up by Westerners as a refugee area, from the thirteenth to the twenty-fourth. Battlefield diaries left by Japanese units are another form of official source material. Only a few are extant and are housed in Defence Agency archives. Among these is that for the Twentieth Regiment’s Fourth Company, which says that it took 328 POWs at the eastern side of the NSZ on the morning of the fourteenth, and shot all of them to death.

Thus battle reports and battlefield diaries– official, public, Japanese military sources– supplement and substantiate personal accounts by Westerners about mass executions of Chinese POWs. Japanese sources of this kind refute the Ministry of Education’s claim, formerly used in textbook screening, that such killings were the acts of a few heartless soldiers in the heat of battle and did not take place in an organized way throughout the army as a whole. These sources also expose the falsity of arguments by Japanese conservative revisionists who, with studied ignorance of international law, insist that the killing of POWs was an extension of combat and thus does not constitute a massacre or an Atrocity. Imperial army records show that Japanese soldiers killed Chinese troops who, having lost all desire and ability to fight back, were begging to surrender so that their lives might be spared.

On the other hand, the fact that battle reports for other units are not extant does not mean that they did not take part in mass killings. One typical organized massacre of POWs occurred at Mufushan, northeast of Nanking. There, the Thirteenth division’s Yamada Detachment or One Hundred-third Brigade, which included the Sixty-fifth Regiment (Morozumi Unit), took custody and “disposed of” 14,777 prisoners. Official, public army records do not mention this fact, but other contemporaneous sources do. These include newspapers such as the Asahi shinbun and volume 1 of the Army General Staff’s own wartime official history, the Shina jihen rikugun senshi. And, personal diaries and private notes by surviving Japanese soldiers in that detachment, brigade, and regiment– plus oral interviews with those men– show that more than 14,477 POWs were massacred. Because of such damning sources, even the postwar Defense Agency’s official war history, the Senshi sosho, and the Kaikosha’s Nankin senshi cannot turn a blind eye to this massacre at Mufushan. The Senshi sosho simply regurgitates an earlier account of the incident, now refuted. According to it, the Japanese units in question released half of their 14,777 men and incarcerated the other 8,000 or so, but half escaped. The units then tried to escort the remaining 4,000 POWs across the Yangzi in order to release them to safety, but the POWs attacked their Japanese guards, so the units had no choice but to open fire, kill 1,000 in self-defense, and let the rest escape. This account is totally make-believe. The foot soldiers’ personal diaries and other private sources that Ono Kenji has unearthed conclusively prove that the Japanese units in question massacred all of the POWs held in custody in an organized manner.

Thus we can assume that many other Japanese units must have taken, and “taken care of,” enormous numbers of POWs– even though this assumption cannot be verified irrefutably because so few battle reports, battlefield diaries, or other “official” records remain extant. The Nankin senshi adopts the position that killed Chinese POWs must enter the victim count when “official” records exist or when the testimonies of eyewitnesses abound, but in all other cases, there is no definitive evidence to substantiate the claim that POWs were taken and killed. This argument– which disparages nonofficial, private sources and oral testimonies– stems from a desire to lower the Chinese victim toll. Still another conservative revisionist argument is that numbers found in battle reports and in other official army records cannot be accepted at face value. As the Nankin senshi asserts, “units undoubtedly inflated figures used in reports to superiors so as to exaggerate their ‘battlefield exploits’.” Of course, statistics in these sources are not totally accurate, but we can only suspect ulterior motives when historians discount their validity solely to deflate Chinese victim counts. The key point here is that both official and nonofficial Japanese records– left by men who did the killing– aver that units “took care of,” “dealt with, or “disposed of” POWs; and, in some cases, expressly say, “shot them dead.” The men who left these documents used such expressions openly because they lacked any idea that the killing of POWs was a violation of international law and a grave crime against humanity. In response to conservative minimalists, we can just as easily conjecture that executions took place when the sources say only that units “took” prisoners, and omit explicit reference to “disposing of” them or “shooting them dead.” Thus the claim that “there are no irrefutable records, ergo there were no massacres,” and the counterclaim that “massacres occurred even though there is no express record thereof,” offset each other– although neither is valid in and of itself.

Organized Nature of the Massacres

To repeat, on 13-14 December, Japanese forces encircled the Chinese army and captured Nanking. Chinese soldiers, lacking a command structure after being abandoned by their commander Tang Shengzhi, lost all will to resist and surrendered en masse only to suffer summary execution in an organized fashion. A key issue in the Nanking Atrocity, then, is to explain why these illegal and unjustifiable executions took place. One answer is that the imperial army– at least during the 1937- 45 Sino-Japanese War– lacked any idea that enemy POWs should be treated in a humane fashion. The idea of universal human rights spread in modern Western states after the French Revolution. Laws governing land warfare were created one after another to ensure the humane treatment of prisoners. Those laws were consolidated in the form of internationally accepted conventions at the 1899 and 1907 Hague Peace Conferences. From 1868 to 1912– in the Meiji and Taishô eras before the Shôwa emperor’s reign– Japan craved recognition as a civilized, modern state equal to the advanced Western powers; so, it strove to earn their respect by obeying international laws of war. That is why the Meiji and Taisho emperors included explicit clauses to that effect in rescripts declaring war on the Qing Empire in 1894, on tsarist Russia in 1904, and on imperial Germany in 1914. But Emperor Hirohito issued no rescript declaring war on the GMD Republic of China in 1937, although the scale of that conflict was unprecedented by far. As just noted, there were pragmatic reasons for calling this an “incident”: the lack of a casus belli that the public would find acceptable, and the need to import strategic matériel from the United States. The government and military unwittingly escalated the up-to-then limited China conflict into a de facto war in August 1937, but even with hostilities spreading in northern China, central military officials in Tokyo told the China Garrison Army General Staff on 5 August 1937 that: “it is inappropriate to follow all specific clauses” in international laws of war, and “our empire is not in a full-scale war with China, so we must avoid using terms such as ‘prisoner of war’ or ‘prize of war’ that may imply the intent to start one.” The same message repeatedly went out to other units later. The Army Ministry’s position was: the laws of war do not apply to an “incident,” so do not use words that connote a formal state of war. This was a momentous change from the past, when imperial rescripts formally declared wars with the stern order for officers and men to obey international law.

Imperial army attitudes at this time exuded contempt for the Chinese army and people. A textbook for noncommissioned officers (NCOs) issued in January 1933 draws a telling distinction between Western states and China in a section titled, “Treatment of Prisoners”: “There is no need to send them to the rear for confinement and wait to see how the war situation changes– as we would do with nationals of other [Western] powers. In the absence of special circumstances, it is alright to release them on the spot or to transport them elsewhere for release. The Chinese’ domicile registration system is full of defects, and most Chinese soldiers are the scum of society, so there is little way for anyone to check whether they are alive or where they are. Thus, even if you were to kill them or release them elsewhere, no one will broach the issue.”

In sum, central army officials instructed field armies not to apply international laws of war. Tokyo did not deem this nonapplication in China to be a war crime, so it is natural that local commanders issued orders to “take no POWs” or to “dispose of” them. Many veterans affirm that high ranking army- and division-level commanders gave such orders during the assault on Nanking. Thus Lt. Sawada Masahisa of the Independent Heavy Artillery’s Second Battalion First Company states: “command headquarters ordered us to shoot to death on sight” 8,000 to 10,000 POWs taken at Xianho Gate on 14 December. (Here, Sawada probably means SEA command headquarters; if so, its commander, Imperial Prince Asaka Yasuhiko, would be complicit.) Or, adjutant Kodama Yoshio of the Sixteenth Division’s Thirty-eighth Regiment says that, when his unit got to a point 1 or 2 kilometers outside the Nanking city walls, the division’s adjutant phoned in a command to “accept no Chinese soldiers who try to surrender; dispose of them.” Official, public battlefield diaries and battle reports list formal commands to “take care of” POWs. One for the Thirty-eighth Regiment contains an order from the Thirtieth Brigade dated “14 December, 4:50 a.m.” Clause 7 reads, “all units are forbidden to take POWs until directed by the [Sixteenth] Division.” A Sixty- eighth Regiment, Third Battalion daily camp ledger dated 16 December reads: “Hereafter, make a cursory survey of troops taken prisoner; then units shall sternly dispose of them.” The Sixty-sixth Regiment’s First Battalion took 1,657 Chinese POWs between 10 and 13 December comprising 18 high-ranking officers plus 1,639 (NCOs) and troops. A battle report dated 13 December records in express detail how this unit killed them.

(8) Received the following order from our Regimental commander at 2:00 p.m.
A. Kill all POWs in accordance with [One Hundred Twenty-seventh] Brigade orders. As a method, we suggest tying them up in groups of less than twenty and shooting them one by one.
B. Collect their weapons and guard these until you receive instructions.
C. While the main force of our Regiment mops up inside the city, your
duties are as outlined above.

(9) Based on the above Regimental command, we [in turn] ordered that the First, Third, and Fourth Companies collect, sort, and guard weapons. At 3:30 p.m., we assembled all companies and, after discussing how to deal with the POWs, decided on the following. We divided them up in 3 equal-sized units and assigned each of our 3 companies to oversee one of these. Each company would place POWs in a guard house to be led out in smaller groups of fifty. The First Company led its group to a valley south of its camp; the Third Company, to a hilly area southwest of its camp; and the Fourth Company, to a valley southeast of its camp. Each company was then supposed to execute its POWs by bayonet, but take pains to guard them heavily, so that none would notice [anything suspicious] when they were being led out. All companies finished preparations and began the executions by 5:00 o’clock, so most were over by about 7:30 p.m. The First Company decided to change plans and instead tried to burn down its guard house. This failed. The POWs, resigned to their fate, stuck out their heads before our swords and stood tall before our bayonets with no sign of fear. Some of them, however, wailed and pleaded for mercy, especially when unit commanders came by to make the rounds.

Only a few official battle reports and battlefield diaries are extant, but we have many personal diaries and reminiscences testifying that summary executions took place on command. There is no doubt that these reflected orders from above and took place systematically– not just haphazardly. The organized nature of the Atrocity was a focal point in the late Ienaga Saburo’s third lawsuit against the government in January 1984. The Ministry of Education had tried to deny this fact by forcing him to retract an account of the Nanking Atrocity in his high school Japanese history textbook. He won the suit in October 1993, when the Tokyo District Higher Court ruled in his favor, and the Supreme Court upheld this decision in August 1997. In sum, Japanese law courts and the government now affirm that massacres took place in an organized way. Organized mass slaughters of POWs, such as that by the One Hundred Third Brigade and Sixty-fifth Regiment at Mufushan as substantiated by Ono Kenji in chapter 4, violated international law.

Defeated Stragglers and “Guerrillas”

The attack on Nanking was a classic example of encirclement. The SEA’s Ninth and Sixteenth divisions plus the Thirteenth division’s One Hundred Third Brigade advanced from the east. The Tenth Army’s Sixth and One Hundred-fourteenth divisions advanced from the southeast. The Fifth division’s Kunisaki Detachment advanced along the far bank of the Yangzi supported by the Eleventh Battle Fleet. Chinese commander Tang Shengzhi at first made no preparations to retreat, and indeed, ordered the city to be defended to the last man. But Chiang Kai-shek then ordered the army to escape for future fighting, and Tang had a change of heart as well; so he ordered a retreat at 5:00 p.m. on the twelfth, before Japanese forces fully surrounded the city in the early hours of the thirteenth. In gross dereliction of his duty as defense commander, Tang and his staff were the first to flee across the Yangzi River.

Abandoned without a chain of command, Tang’s troops lost all will to resist. Hordes of them east and south of the city were in chaos, while those trapped inside the city walls fled for Bajiang Gate, their sole escape hatch, only to find it shut. Masses of defeated Chinese troops and refugees had gathered at Hsiakwan wharf on the thirteenth, when the Eleventh Battle Fleet arrived at 5:00 p.m. According to Japanese naval sources, the fleet “fired fiercely on defeated stragglers who hoped to flee to the far shore and cut them to ribbons.” With their escape route by water cut off, masses of Chinese troops began madly rushing up and down the Yangzi river bank, desperately seeking safety. Victory was already decided; virtually all of the defeated Chinese soldiers lacked weapons and any will to resist. But the imperial army and navy fired on these helpless troops and also on civilians. It is clearly wrong to call this a combat operation; it was a slaughter, a massacre. An “annihilation of defeated enemy troops”– plus great numbers of civilians mixed in– was conducted by the Sixteenth Division at Hsiakwan, Bajiang Gate, and Maqun; and by the Sixth Division at Hsiakwan, Xinhechen, and Jiangdong Gate. In fact, each of these actions was simply a turkey shoot of defenseless people.

Defeated remnants of the Chinese army discarded weapons, stripped off their uniforms, and slipped into the city, where the imperial army began mop-up operations on 13 December with orders to round up anyone suspected of being a soldier. The Ninth division handled areas south of Zhongshan Road; the Sixteenth division, those north of it. On the fourteenth, Japanese troops forayed into the Nanking Safety Zone (NSZ), claiming that defeated Chinese stragglers, disguised in civilian clothing; that is, guerrillas, had taken refuge therein. The reasons for this haste lay in the CCAA’s decision to hold a triumphal entry procession into Nanking on the seventeenth– made despite SEA objections that this was too early to ensure safety. Newspapers at home had been playing up the capture of the enemy capital, so the CCAA could not lose face by seeming to dawdle. On top of that, Lt. Gen. Asaka Yasuhiko– an imperial prince and uncle of Emperor Hirohito– was to take a leading part in the ceremony as SEA commander. Thus the CCAA had to take every possible precaution to prevent harm from befalling his royal personage. The SEA’s Ninth Division Seventh Regiment mopped up the Nanking Safety Zone from the thirteenth to the twenty-fourth, and its battle report records “6,670 [Chinese] killed by bullet and bayonet.” This regiment’s immediate superior officer, Brigade commander Maj. Gen. Akiyama Yoshimitsu, stipulated “Points to Note in Mop Up,” dated 13 December. This document expressly said: “View all youths and adult males as defeated stragglers or soldiers disguised in civilian clothes; round up and detain all of them.” This implies that many civilians were likely among the 6,670 killed.

Seventh Regiment commander Col. Isa Kazuo, First Company Private First-Class Mizutani So, and Second Company Lance Cor. Inouie Mataichi left diaries. Isa made simple entries such as “Mopped up from the morning. The [Nanking] Safety Zone is in our area. It is said to hold about 10,000 refugees” on the fourteenth, or “we sternly disposed of about 6,500 over three days of mop up” on the sixteenth. By contrast, Inoie and Mizutani went into more specific detail. Inoie, for example, writes: “We set out in the afternoon as well and came back with 335 young captives. We ferreted out all males among the refugees who looked like defeated stragglers. Man! Some had family members there, and did they ever wail when we tried to take their men folk away! They’d latch on to our arms and bodies, pleading with us…. We took these 335 down near the Yangzi where other troops shot them dead.” Mizutani’s entry for the sixteenth reads:

In the afternoon we went to the [Nanking] Safety Zone for mop up. We placed sentries with bayonets at the intersections, blocked these off, and went about our work rounding up virtually all young men we came across. We roped them off, surrounded them with armed guards, tied them up in rows, and led them away so that they looked like kids playing choo-choo train. Our First Company clearly took less than other units, but we still got a hundred and several dozen. Lots of women, no doubt their mothers or wives, soon caught up with us to cry and beg for their release. Right away, we released those who clearly looked like civilians and shot thirty-six others to death. All of them wailed desperately to be spared, but there was nothing we could do. Even if some unfortunate innocent victims were mixed in (we couldn’t tell for sure), it just couldn’t be helped. Killing some innocent victims was unavoidable. [CCAA] Commander Matsui ordered us to clean out each and every anti-Japanese element and defeated straggler, so we did that in the harshest possible manner.

The SEA Sixteenth Division’s Twentieth Regiment also mopped up in the Safety Zone. Fourth Company Lance Corporal Masuda Rokusuke wrote: “14 December. Mop Up. Entered the [Nanking] Safety Zone. Cleaned out defeated stragglers mixed in with refugees. Our Fourth Company alone took care of no less than 500; we shot them dead next to Xuanwu Gate. I hear that all our units did the same thing.” In an account written on the order of his company commander after entering Nanking, Masuda noted:

On the morning of the fourteenth we went to mop up the [Nanking] Safety Zone– run by some kind of international committee. We surrounded tens of thousands of defeated stragglers who had fiercely resisted us until yesterday. Not a single one would escape now. They all fled into that Safety Zone. But we were determined to go in, search every nook and cranny, flush them all out, and exact revenge for our fallen buddies. Each of our squads looked over all males in those big, complex Chinese houses. In one of these, Lance Corporal Maebara and his troops found a few hundred defeated stragglers changing into civvies. Hearing of this, we went to have a look. What a sight! Next to them were tons of rifles, revolvers, swords, and other weapons. Some of those men were still in uniform. Some were hastily changing into ordinary Chinese clothes. Others wore civilian shirts with army- uniform trousers. All of the clothes were either unsuited to winter or mismatched as to shirts and pants, so the men obviously had grabbed and donned these in a big rush. We led all of them off, stripped them down, checked them out, and tied them up with downed telephone wires…. With dusk approaching, we marched close to 600 of these defeated stragglers over toward Xuanwu Gate and shot them dead.

To execute soldiers lacking the will and means to resist on the pretext that they are “defeated stragglers” or “combatants disguised in civilian clothes” is unjustifiable, illegal, and inhumane. Worse still, it is a downright atrocity to slaughter huge numbers of civilians in the process without making an effort to ascertain if they in fact are military personnel. Even foreign nationals from states friendly to Japan concurred on this point. On 20 January 1938, the German Branch Consulate in Nanking sent this report to its Foreign Ministry:

Our few policemen could not stop vast numbers of Chinese soldiers from fleeing into the [Nanking] Safety Zone. (Some had thrown away their arms, but even when this was not the case, they lacked any means to resist.) On that pretext, the Japanese army began a massive search of houses and hauled away all Chinese suspected of being soldiers. The usual Japanese way of determining whom to seize was to check for abrasions or other tell-tale signs of having worn helmets on their heads, carried rifles on their shoulders, or lugged knapsacks on their backs. Foreign witnesses say that the Japanese tricked the Chinese by promising to give them work or to pardon them, but then led them away to be killed. The Japanese took no steps to declare martial law or anything of the sort. Why should we expect any such pretensions on their part? They flout the conventions of law in wartime as well as the rules of human decency?

Conservative revisionists in Japan today deny the Nanking Atrocity or justify it by claiming that mop-up operations were conducted against Chinese soldiers disguised in civilian clothes. These guerrillas, or would-be guerrillas, it is claimed, pretended to be peace-loving civilians but actually bore concealed weapons waiting for a chance to snipe at Japanese troops. That form of combat violated international laws of war, so those Chinese combatants forfeited all legal rights that POWs enjoy. As this argument goes, it was a justifiable act of self-defense for Japanese units to kill them, and it was also permissible to capture and execute them for committing these acts, which were war crimes. Raids into the Nanking Safety Zone (NSZ), it is held, were legitimate combat operations to wipe out enemy soldiers disguised as civilians. In fact, however, Chinese soldiers who fled into the NSZ lacked the will to fight that was needed to become guerrillas. They had no place of refuge except the NSZ, and they changed into civilian clothes simply to avoid being killed by the invaders. It was indefensible for Japanese troops to kill them on the spot with no effort to ascertain their true status, or to execute them as war criminals without bringing them before military tribunals. Furthermore, it was even less justifiable to kill large numbers of innocent civilians based on arbitrary criteria such as having what seemed to be helmet or shoulder-strap abrasions, which purported “proved” that they were soldiers.

Atrocities against Civilians

The “Rape of Nanking,” as it was first called in 1937-38, became known the world over because of the huge number of rapes and mass murders committed against civilians. These atrocities took center stage at the Tokyo war crimes trials where Chinese victims and foreign witnesses testified, and the People’s Republic of China (PRC) emphasizes this issue today by seeking out ever more victims and witnesses. Thus, crimes against the general civilian population remains the key point in the debate over Nanking. However, conservative revisionists in Japan remain mum on this point, ignoring all testimonies by Chinese victims, neutral foreign witnesses, and even Japanese victimizers as being uncorroborated in bona fide primary sources. These conservatives begrudge at most that: “The Japanese army did commit misdeeds against the civilian population. The International Committee for the Nanking Safety Zone protested against the killing of 47 civilians. Not all of these cases can be substantiated, but even if they could, that’s a total of forty-seven– hardly a big number compared with other armies in history who captured and occupied foreign capitals in wartime.” Thus, conservative revisionists deny that Japanese troops perpetrated large- scale atrocities against Chinese civilians. But the imperial army itself admitted this fact at the time. On 4 January 1938, IH, in the name of Field Marshal Prince Kan’in Kotohito, an uncle of Emperor Hirohito, issued an unprecedented statement to CCAA commander Matsui Iwane: “If we look at actual conditions in the army, we must admit that much is less than blemish-free. Invidious incidents, especially as to troop discipline and morality, have occurred with increasing frequency of late. However much we may wish to disbelieve this fact, we cannot but have doubts.” Though hardly blunt and direct, this imperial prince admonished against ongoing Japanese atrocities.

Other documents clearly show that troops were being disciplined for criminal acts at the time. For example, one source is a Tenth Army legal department daily ledger from 12 October 1937 to 23 February 1938; another is a CCAA battlefield courts martial daily ledger from 4 January to 6 February 1938. The absurdly small number of military police on hand could not control an invading army of well over 100,000; so we can be sure that only those men caught for the most egregious of crimes were court-martialed, and they were punished only as examples to deter others. Still, the following figures in the Tenth Army legal department daily ledger prove that rapes took place– although these represent the tip of an iceberg. Of the 102 men convicted as of 18 February 1938, twenty-two were for rape, twenty-seven, for murder; two for rape-and-murder; and two for causing bodily injury that resulted in death. Of the 16 men awaiting trial on that date, two were charged with rape and one with murder. The occurrence of such heinous crimes is substantiated as well by a directive issued on 20 December 1937 that inveighs against the high incidence of rape in the Tenth Army: “We have told troops numerous times that looting, rape, and arson are forbidden, but judging from the shameful fact that over 100 incidents of rape came to light during the current assault on Nanking, we bring this matter to your attention yet again despite the repetition.”

Commander Matsui Iwane himself noted the reality of rape in the CCAA battlefield courts martial daily ledger. An entry for 20 December reads: “It seems that our troops committed acts of rape and looting (mainly items like furniture) at about 1:00 p.m.; the truth is that some such acts are unavoidable.” Matsui’s entry in the ledger for 26 December reads: “I again heard of looting and rape in and around Nanking and Hangzhou.” In explaining these passages, the Nankin senshi’s (History of the Nanking Campaign) protective authors claim that Matsui “construed the so-called ‘Nanking Incident’ as comprising violations of foreign rights and privileges in China and incidents of violence and looting against the Chinese. He had no idea of an ‘Atrocity’ that would arise as a problem later on [at war crimes trials].” If there was “a problem,” however, it lay precisely in this lack of cognizance by the highest ranking Japanese commander during the Nanking campaign.

Indeed, then, the imperial army’s upper echelons did know that troops were perpetrating rape and violence against Chinese civilians. Lt. Gen. Okamura Yasuji, who took over command of the Tenth Army before its assault on Wuhan in August 1938, later recalled: “I surmised the following based on what I heard from Staff Officer Miyazaki, CCAA Special Service Department Chief Harada, and Hangzhou Special Service Department Chief Hagiwara a day or two after I arrived in Shanghai. First, it is true that tens of thousands of acts of violence, such as looting and rape, took place against civilians during the assault on Nanking. Second, front-line troops indulged in the evil practice of executing POWs on the pretext of [lacking] rations.”

Rapes were especially prevalent at Nanking, and were a focus of the third lawsuit that the late Ienaga Saburo brought against the Japanese government. In its 1983 screening of high school history textbooks, the Ministry of Education ordered him to delete a footnote that read: “There were many officers and men in the Japanese army who violated Chinese women.” The Ministry did admit that women were violated, but insisted that: “This occurred on all battlefields in all periods of human history. His [Ienaga’s] selection of facts is problematic if he makes this point only in the case of the Japanese army.” Rape is an immeasurably traumatic experience for the females involved; it leaves lifelong emotional scars. Given the shame-inducing nature of this crime, victims naturally wish to keep it secret; so written contemporaneous sources that document it are very rare. Nevertheless, rape by the imperial army was a major problem in the early stages of the 1937-45 Sino-Japanese War, especially at Nanking, even when compared with behavior in the Meiji-Qing (or First Sino-Japanese) War and the Russo-Japanese War.

One source that sheds light on this matter is a directive issued in February 1939 by an Army Ministry undersecretary to units returning home from the China front. It sought to ensure that soldiers exercise discretion in talking about their experiences; in other words, they should keep still about what happened. An appendix to the source lists the following specific examples of verbal statements that the army wished to suppress:

(1) At XX, we took four people captive– parents and daughters. We played with the daughters as if they were whores and killed the parents because they kept on telling us to release the daughters. We had our kicks until the unit was ordered to leave; then we killed the daughters. (2) One company commander hinted that rape was OK, saying, “Make sure no problems arise later on; after you’re finished, either pay them off or kill them outright.” (3) Every soldier who fought in the war must be a murderer, armed robber, or rapist. (4) No one cared about rapes at the front; some guys even shot at MPs who caught them in the act. (5) The only skills I picked up after half a year in combat were how to rape and loot.

These are documents left by the perpetrators. Needless to say, the victims left many as well. Foreigners also left testimonies and conducted surveys. Conservative revisionists say that these Western sources lack credibility because Britons and Americans were enemies who hated Japan at the time. But nationals of friendly states such as Germany, which signed the Anti-Comintern Pact with Japan in 1936, also left documents. I have already cited a message sent by the German Branch Consulate in Nanking. There are also reports to the German Foreign Ministry sent by Georg Rosen of the German Consulate staff and by John Rabe– a Nazi Party member, Siemens company employee, and head of the International Committee (IC) that administered the NSZ. Even these nationals friendly to Japan candidly exposed Japanese misdeeds. For example, on 24 December 1937, Rosen reported: “The most disgusting acts by Japanese soldiers against Chinese civilians have come to light; these clearly work against Germany’s policy of thwarting the spread of Communism.” On 15 January 1938, he reported: “Over a month has passed since the Japanese army occupied Nanking, but soldiers are still abducting and raping women and girls. In that sense, the Japanese army is erecting a monument to its own dishonor.” Rosen also noted that Japanese soldiers were breaking into the German Embassy and the ambassador’s official residence demanding women.

John Rabe’s diary describes horrible rapes by Japanese soldiers. For example, on 17 December he writes: “One of the Americans put it this way: ‘The Safety Zone has turned into a public house for the Japanese soldiers.’ That’s very close to the truth. Last night up to 1,000 women and girls are said to have been raped, about 100 girls at Ginling Girls College alone. You hear of nothing but rape. If husbands or brothers intervene, they’re shot. What you hear and see on all sides is the brutality and bestiality of the Japanese soldiery.” On 24 December Rabe noted: “Dr. Wilson used the opportunity to show me a few of his patients. The woman who was admitted because of a miscarriage and had the bayonet cuts all over her face is doing fairly well.” This woman is probably Li Xiuying, one plaintiff in a lawsuit over wartime compensation launched by Chinese victims in the Tokyo District Court. Such objective contemporaneous reports by Westerners constitute undeniable evidence that huge numbers of rapes took place. And, I wish to stress, these reports were tendered by members of a nationality friendly to Japan, not by Chinese victims or by American, Australian, and British “enemies.”

The imperial army responded to those criticisms by creating typically Japanese “comfort stations” staffed by so-called “comfort women.” At an inquiry prior to the Tokyo War Crimes Trials, a former Kwantung Army staff officer, Tanaka Ryukichi, gave this account of how the institution came into being: “[CCAA Staff Officer] Cho Isamu told me that officers and troops were raping too many women, so he set up brothels in Nanking to stop this.” Earlier, in the 1918-22 Siberian Intervention, Japanese troops had used “comfort women” to prevent rapes and the spread of venereal disease. Now, alarmed by the high incidence of rape at Nanking, the imperial army organized groups of such women to accompany troops during the later assault on Wuhan. Thereafter, the army officially recognized and set up comfort stations wherever it went. Because there were not enough Japanese women to meet the increased demand, Korean women were secured by fraud or force and sent to the front. Rapes occurred repeatedly all the way from Shanghai to Nanking and also in the capital after it fell. This was a major war crime perpetrated by the imperial army; indeed, the epithet “Rape of Nanking,” as the event was called at the time, came to stand for the Atrocity as a whole. This war crime not only left deep scars on the Chinese, it also has had major implications for problems that plague Japan’s relations with North and South Korea. Conservative revisionists in Japan deny that a Nanking Atrocity took place by asserting that only this or that many people were victimized, or that empirical evidence for their victimization is not ironclad. But we will never comprehend the true nature of Japanese atrocities at Nanking if we turn a blind eye to the tens of thousands of women reputedly raped there.

Historical Awareness

The Nanking Atrocity symbolized Japan’s war of aggression against China. There were foreign embassies and news agencies in Nanking, then the capital of China, so reports of the Atrocity went out to the entire world. The Japanese people alone, with few exceptions, remained in the dark because of severe wartime censorship. Thus the great majority of Japanese learned about the Atrocity only during the Tokyo War Crimes Trials, which first broached the issue of Japan’s war guilt and culpability. Foreign physicians and missionaries who had lived in Nanking, plus Chinese persons victimized there, testified at Tokyo. Other types of evidence included sociologist Lewis Smythe’s surveys of damages, burial records left by the Nanking branch of the Chinese Red Cross, the Red Swastika Society, and also those by other local organizations. As a result of this evidence, CCAA commander Matsui Iwane received the death sentence. The verdict read: “the total number of civilians and POWs murdered in Nanking and its vicinity during the first six weeks of the Japanese occupation was over 200,000.” Article XI of the San Francisco Peace Treaty, signed in 1951, stipulates: “Japan accepts the verdicts of war crimes trials.” In sum, the postwar Japanese government formally admits that imperial Japan waged a war of aggression and that it massacred over 200,000 people at Nanking.

Matsui Iwane brought in at the Tokyo Trial

But unlike its former ally Germany, Japan did not make an all-out effort to prosecute war crimes or criminals later in the postwar era after the Allied Occupation ended. Former wartime leaders, even some who had been convicted of A-class war crimes, returned to positions of power. Sentiments to affirm or even glorify the war became influential after the San Francisco Peace Treaty took effect and Japan regained sovereignty in April 1952. As seen in the Ienaga lawsuits beginning in 1967, the Ministry of Education censored school textbooks to ban words such as “aggression” or delete mention of the Nanking Atrocity. Moreover, conservative revisionists began to argue that the Atrocity was a fabrication or an illusion. From the 1970s onward, controversies threatening the very basis of historical fact have raged in Japan. In 1982, China and South Korea formally protested against Japan’s government when they learned of conditions surrounding textbook screening. The Suzuki Zenko regime, in power from July 1980 to November 1982, settled this diplomatic rift by having Miyazawa Kiichi, Director of the Cabinet Secretariat, proclaim that Japan “reaffirms the spirit of self-criticism espoused in the [1972] ‘Sino- Japanese Joint Statement’ and [1965] ‘Joint Communiqué’ between Japan and South Korea,” and also that “the Japanese government will take responsibility for correcting textbook passages.” Hawks in the ruling Liberal Democratic Party as well as other right-wing elements lashed out at China and South Korea for allegedly intervening in Japanese internal affairs on this and other issues; and, as a result, debates over history flared up again, with deniers contending that the Nanking Atrocity was a fiction or a falsehood. In response, we formed the Society to Study the Nanking Incident in 1982, and it has continued publishing scholarly books and articles to this day. As more and more of these studies appeared, most aspects of the Atrocity have become clearly known.

The fatal scholarly blow to the conservative revisionist cause came between April 1984 and March 1985 when the Kaikosha elicited testimonies from members who had served at Nanking for publication in its monthly, the Kaiko. Its editors hoped they would settle the controversy once and for all by publishing great numbers of eyewitness testimonies that denied major misdeeds. But contrary to those expectations, many Kaikosha members sent in accounts affirming that massacres, rapes, and other acts of wanton violence took place. To their credit, the Kaiko editors published these materials unaltered, and Chief Editor Katogawa Kotaro ended the series in March 1985 with an article titled, “Summing Up,” in which he admitted the fact of illegal killings, and even of massacres. He cited estimated victim counts of 3,000 to 6,000 tendered by Unemoto Masami, and of 13,000, by Itakura Masaaki; and Katogawa concluded the series by saying:

We Deeply Apologize to the Chinese People.

To repeat: 3,000 to 6,000 is a terrible figure; how much more so is 13,000. When we began compiling our history, we were prepared to accept that Japan was not innocent. Nevertheless, we can only reflect upon such huge numbers with deep sadness. No matter what the conditions of battle were, and no matter how that affected the hearts of men, such large-scale illegal killings cannot be justified. As someone affiliated with the former Japanese army, I can only apologize deeply to the Chinese people. I am truly sorry. We did horrible things to you.

The Kaikosha published a semi-official history, the Nankin senshi, plus a collection of primary sources, the Nankin senshi shiryo shu. These two works list 15,760 civilian casualties and 16,000 POWs summarily killed, but the editors took a reactionary step by insisting that not all of these were illegal or illegitimate killings, and that Chinese counterclaims of 200,000 or 300,000 victims are fabricated. Even so, by publishing primary sources that contain the facts and by admitting the Atrocity’s historicity, the Kaikosha conclusively repudiated false claims that the event never took place or was an illusion. Indeed, several contributors to the present volume, myself included, cite these Kaikosha publications in our chapters. Thus, as a scholarly argument, denial was dead.

However, the 1990s brought new developments. Konoe Fumimaro’s grandson, Prime Minister Hosokawa Morihiro, in 1993 admitted that “aggressive acts”– though not aggression itself– took place in the last war. And, in 1995 Prime Minister Murayama Tomiichi from the Japan Socialist Party expressed self-criticism and sorrow for Japan’s colonial rule and aggression in Asia. Those statements provoked more right-wing efforts to affirm and glorify the war and to deny historicity to the Nanking Atrocity– propositions that had suffered refutation in academic circles. Once more, battles over historical awareness erupted. Today, this is no longer a debate over facts or empirical proof, no longer a matter of scholarship. Conservative revisionists today who insist on denying the Nanking Atrocity– despite all evidence to the contrary– do so for political reasons. They wish to sanctify a road to future wars by glorifying the last one. According to them, Japan is already an economic power; it now should join the ranks of political and military powers too. This means winning a permanent seat on the UN Security Council. To do that, they argue, Japan must make suitable international contributions by sending troops overseas to join in combat roles as part of UN armed forces. But before that can happen, the Japanese must amend or eliminate Article IX of their postwar Constitution so that their nation can once again use military force to settle international disputes. Such constitutional revision can not take place until the Japanese overcome their aversion to wars in general, based on misperceptions of the last one they fought, and until they stop their abject practice of “diplomacy by apology” due to unwarranted feelings of war guilt. In sum, as the conservative revisionist line goes, pacifism and servility are foolish because World War II was a just and glorious one for Japan. This is the nakedly political logic motivating those conservative revisionists who go on denying the Atrocity’s factuality long after such claims have been exposed as scholarly bankrupt.

This article was posted at Japan Focus on October 23, 2007.

Share with a colleague:

Volume 5 | Issue 10

Article ID 2553

About the author:

The Asia-Pacific Journal: Japan Focus is a peer-reviewed publication, providing critical analysis of the forces shaping the Asia-Pacific and the world.

    About the author:


    Our monthly newsletter provides readers with an in-depth analysis of forces shaping the Asia-Pacific and the world.

      Since 2002

      Asia Pacific Journal: Japan Focus has produced critical reporting on geopolitics, economics, history, environment, and international relations.