Chapter 6The Dilemma of Inadvertent Expansion Japan and Italy

Historical Background

On the eve of the invasion of Manchuria in September 1931, Japan had had a continuous presence on the Chinese mainland since its surprise victory in the Russo-Japanese War in 1905. Under the Treaty of Portsmouth, the Russian Empire ceded to Japan its lease of the Liaodong Peninsula (which would be renamed the Kwantung Leased Territory) as well as the South Manchuria Railway, the 1,129 km rail system with lines running from Port Arthur to Changchun and from Mukden to Antung.3 Manchuria as a whole totaled roughly 985,000 km2 in China’s northeast region, bordering the Korean Peninsula to the south and the Soviet Far East to the north and east.4 It had a population of 30 million, approximately 220,000 of which were Japanese migrants who had traveled there to work for the South Manchuria Railway or to pursue other opportunities.5 This was shortly after the “Warlord Era” in China (1916–28), when Manchuria had been ruled by the influential warlord Zhang Zuolin. While China would be weakly unified after Chiang Kai-shek’s “Northern Expedition” in 1928, Manchuria still had a great deal of autonomy from the Nationalist regime in Nanjing.6

Stationed in the Kwantung Leased Territory, and all along the South Manchuria Railway, was the Kwantung Army, a colonial branch of the Imperial Japanese Army, headquartered in Port Arthur.7 Totaling just 10,400 personnel, the Kwantung Army was commanded by a series of generals on two-year rotations and staffed by a few hundred mid-ranking officers.8 Among these was a forty-two-year-old lieutenant colonel by the name of Ishiwara Kanji.9 A career army officer, Ishiwara was intense, idealistic, and intellectually gifted.10 Yet, he was also strident, impulsive, and contemptuous of Japan’s Taisho-era political leadership.11 After converting to Nichiren Buddhism in his early thirties, Ishiwara developed radical, apocalyptic views of a future world-altering clash between Japan and its enemies in the West, particularly the United States.12 Preparing Japan for such an eventuality became his life-defining mission, the first step of which was to take place in Manchuria. Ishiwara was joined on the Kwantung Army staff in June 1929 by his childhood friend and classmate, Itagaki Seishiro. Despite many personal differences, what the two shared in common were a deep conviction that Manchuria presented threats and opportunities for Japan and that outright occupation was the necessary response.13 Ishiwara and Itagaki were the actors on Japan’s imperial periphery who played a crucial role in planning and executing of the invasion of Manchuria.14

The prime minister of Japan at this time was Wakatsuki Reijiro. His second stint at the premiership, he and many members of his cabinet tended to take a dovish view of Japan’s relations with China and Manchuria. Chief among these was the famed diplomat and foreign minister, Shidehara Kijuro, whose very name came to be associated with the liberal views that defined Japanese foreign policy in the 1920s. Standing somewhat outside this more liberal consensus was War Minister Minami Jiro, an army general whose foreign policy views, naturally, hewed more closely to those of the Imperial Japanese Army. Finally, there was the head of state, the Showa Emperor Hirohito, who was only a few years into his reign but was proving to be a more politically active emperor than his recent predecessors.15 These were the leaders in the capital Tokyo who would be dragged unwittingly, and mostly unwillingly, into further territorial acquisitions on the Chinese mainland.16

Manchuria, on the eve of the Japanese invasion, was ruled by Zhang Xueliang, the son and successor of the warlord Zhang Zuolin. While the father, Zuolin, had been an ally of the Japanese in Manchuria against the government in Nanjing, his unauthorized 1928 assassination by a member of the Kwantung Army would push his son, Xueliang, to cooperate more closely with Nanjing against the Japanese.17 Zhang Xueliang was the head of what was known as the Fengtien Army, a force of approximately 250,000 personnel.18 To the south, on the other side of the Great Wall, was the recently established Nationalist Kuomintang regime of Chiang Kai-shek, with its capital in Nanjing. And to the north of Manchuria was the Soviet Union, which had been established less than a decade earlier at the conclusion of the Russian Civil War. In any case of armed conflict involving the Kwantung Army, the key questions were what kind of resistance Zhang’s forces would put up, and whether the Kuomintang or the Soviet Red Army would intervene.

Tokyo and Manchuria

Leaders in Tokyo faced severe principal-agent problems vis-à-vis the Kwantung Army in Manchuria as a result of information asymmetries favoring the army and a divergence of preferences between capital and periphery. First, there were stark information asymmetries in favor of the Kwantung Army due to its substantial institutional autonomy and unusually weak civilian oversight over the Japanese military.19 According to articles 11 and 12 of the Meiji Constitution of 1889, the army and navy were overseen by the emperor himself, not the cabinet or the Japanese Diet (parliament). These same constitutional provisions formally institutionalized the military’s traditional “right of supreme command” (dokudan senko), the principle allowing staff officers of field armies autonomy from civilian control in the areas of operational planning and execution.20 Japanese law also mandated that the military had to approve ministers of the army and navy for appointment and that the resignation of either of these officers could lead to the dissolution of the cabinet.21 These institutional realities were not helped by the fact that the Kwantung Army itself was the primary source of information coming out of Manchuria. While there were Japanese Foreign Ministry consulates in Mukden, Antung, Dairen, and other major cities in the region, they were lightly staffed and often had to rely on army sources themselves.22 There was also the South Manchuria Railway; however, its management was generally sympathetic to the views of the Kwantung Army and, therefore, willing to put a similar “spin” on information it sent to Tokyo.23 These constitutional and informational conditions gave the Kwantung Army a high degree of independence, greatly hampering civilian and even central military control.

Second, there was a sharp divergence of preferences between the leaders in Tokyo and the Kwantung Army in Manchuria. Prime Minister Wakatsuki and much of his cabinet were highly cautious when it came to China policy. This was the era of “Shidehara Diplomacy,” whose core tenets were international cooperation, economic diplomacy, and nonintervention in China’s domestic political affairs.24 Wakatsuki and Foreign Minister Shidehara, therefore, advocated for maintaining the territorial status quo in China and for dealing with any existing problems through negotiation. War Minister Minami’s policy views on Manchuria were more hawkish than the views of his cabinet colleagues, though he did want to tread carefully, was concerned with discipline within the Kwantung Army, and was willing to restrain its more radical elements.25 And while Emperor Hirohito’s specific policy views with respect to Manchuria were not always clear or consistent, he did favor a more conciliatory approach to China and repeatedly argued for the need to maintain discipline within the army.26 In short, leaders in Tokyo embodied what I referred to in chapter 1 as “the view from the capital.” Their responsibilities were broad and weighty, being concerned not only with policy in and around Manchuria but with the well-being and defense of Japan, the interests of the empire as a whole, and relations with other regional states and global great powers. And being cloistered away in the capital Tokyo, they felt few of the daily effects of the nationalist, anti-Japanese upheaval in Manchuria and saw little urgency to act.27

The Kwantung Army, in contrast, was much less cautious in its China policy. They were far less concerned about the possibility of a Soviet intervention or of the reactions of the other great powers. Being away from the main islands, the Kwantung Army was largely isolated from the domestic politics of Japan.28 With little knowledge of, or experience dealing with, international trade and finance, they were far less concerned about the risk of economic sanctions.29 And many Kwantung Army officers were largely indifferent to the opinions of the other great powers and their publics.30 Thus, many members of the Kwantung Army and, most particularly, officers, such as Ishiwara and Itagaki, advocated for the complete annexation of Manchuria.31 Manchuria was a rich source of valuable natural resources, which would be necessary in the approaching war of attrition with the West.32 And invading Manchuria would forestall what Ishiwara saw as an inevitable Soviet occupation, preventing, in his terms, the “communization of Asia.”33 Ishiwara and many of his Kwantung Army colleagues, in sum, clearly embodied the “view from the frontier.” Their responsibilities were relatively narrow, being concerned with defending the roughly 3,700 km2 that Japan possessed in Manchuria, rather than the empire as a whole.34 And their sense of urgency to take action in Manchuria was great, as they faced Chinese nationalist upheaval daily and directly. As the historian and Ishiwara biographer Mark Peattie puts it, “Ishiwara in a very real sense was stationed on a sort of Japanese ‘imperial frontier,’ a semi-colonial environment in which the proximity of danger and opportunity served to reinforce the conviction that the clearest solution to national problems lay close at hand.”35

The “Mukden Incident”

While the idea of separating Manchuria from China by force had emerged within the Kwantung Army as early as 1916, planning in earnest for an invasion began in July 1929.36 The Kwantung Army was numerically and materially inferior to its adversaries, lacking mechanized forces and aircraft and being lightly equipped in artillery, engineering, and transport.37 Thus, it was essential that the plans be meticulously organized, stressing the importance of intelligence, rigorous training, and the use of surprise, speed, and the concentration of force.38 The idea was to devise a series of tightly interlinked operational plans that would trigger one another in a sequential fashion, creating a process that, once set in motion, would be very difficult to stop or reverse. Then, all that would be needed was a crisis of a sufficient magnitude to light the fuse, which would be easy to orchestrate. As Ishiwara put it in May 1931, if needed, the Kwantung Army could “create the occasion for this with a plot and force the nation to go along.”39 Operational plans were finalized by the summer of 1931.

Despite the information asymmetries favoring the army in Manchuria, it would prove difficult to keep the conspiracy a secret for long. In August and early September 1931, rumors began to circulate in the capital that trouble was brewing in Manchuria, prompting reporters to regularly press the prime minister and foreign minister for more information.40 On August 18, a top adviser to Emperor Hirohito said to an aide, “I can’t help but think the imperial army is cooking something up in Manchuria, Mongolia, and China.”41 On September 4, the Foreign Ministry received a telegram from Manchuria warning that “a plot is afoot among young officers in the Kwantung Army to thrash the Chinese army.”42 These rumors were taken so seriously that, on September 11, the emperor himself summoned War Minister Minami to question him on the state of military discipline.43 While Minami assured him that things were under control, the emperor admonished him to “be even more cautious.”44

On September 15, Foreign Minister Shidehara received a telegram from the consul general in Mukden, informing him that the “Kwantung Army [is] assembling troops and bringing out munitions[;] seem likely to start action in the near future.”45 That same day, the War Ministry dispatched the General Staff intelligence section chief Major General Tatekawa Yoshistugu to Manchuria to remind the Kwantung Army of the cabinet’s policy of nonintervention in China and to put a stop to any impending plots. Yet, news of Tatekawa’s trip was cabled from an accomplice of Itagaki’s in Army Headquarters in Tokyo, warning: “Plot discovered. Tatekawa coming; strike first to avoid implicating him.”46 Ishiwara and Itagaki took this advice to heart. While the invasion of Manchuria had been planned for September 27, they moved it up to the evening of the eighteenth, the day that Tatekawa was supposed to arrive.47 Upon his arrival, Itagaki then had Tatekawa whisked away to be wined and dined at a local restaurant, where he would ultimately pass out.48 Just a few hours later, the invasion of Manchuria was launched.

At 10:20 p.m. on September 18, 1931, there was an explosion on the southbound track of the South Manchuria Railway at Liutiaokou, just north of Mukden. The charge had been set by a lieutenant in the Kwantung Army, with the aim of framing Zhang Xueliang’s army with the sabotage. Local Kwantung Army conspirators then rushed to the scene, claimed they were fired on by Chinese soldiers and returned fire, and pursued the enemy while calling for reinforcements.49 In accordance with Kwantung Army plans established by Ishiwara and Itagaki, a local battalion commander then ordered an attack on the Fengtien Army barracks at Mukden, which housed as many as ten thousand personnel. As Zhang had recently ordered his troops to under no circumstances resort to force in any confrontation with the Japanese, the barracks were overrun within a few hours and at minimal cost to the Kwantung Army.50 Just four hours after the initial explosion, the Imperial Japanese Army in Korea received a request from the Kwantung Army to dispatch reinforcements, which began to mobilize immediately.51 By 1:00 p.m. the next day, Mukden as a whole was under Kwantung Army control, and by 3:00 p.m., the South Manchuria Railway’s terminal city of Changchun, to the north, was occupied. Within less than twenty-four hours, the invasion of Manchuria was well underway.

Tokyo Reacts

The first meeting of Prime Minister Wakatsuki’s cabinet to deal with what became known as the “Mukden Incident” was held in Tokyo on September 19 at 8:00 a.m., the morning after the explosion. It was agreed, in line with Wakatsuki’s, Foreign Minister Shidehara’s, and Emperor Hirohito’s preferences, that the crisis should be localized, the spread of hostilities contained, and the dispute settled as expeditiously as possible.52 As Wakatsuki put it that morning, the plan was to “immediately instruct the commanding officer of the Kwantung Army not to enlarge the theater of conflict nor to bombard and occupy government buildings and fortifications.”53 War Minister Minami was more sympathetic to the Kwantung Army’s position but felt bound by the cabinet’s, and especially the emperor’s, wishes, and so he relayed these orders to the Kwantung Army at 6:00 p.m. that evening.54 The desire to promptly settle the crisis, and to contain and even reverse the Kwantung Army’s expansion to the greatest extent possible, would be the Wakatsuki cabinet’s position for the remainder of its tenure.55

The cabinet’s primary concern was the geopolitical risk associated with expanding further into Manchuria. To start, leaders in Tokyo were concerned about the hundreds of thousands of Manchurian forces led by Zhang Xueliang.56 Then there was the Soviet Union to worry about. As noted earlier, the Soviet Union shared a lengthy border with Manchuria and had interests in northern Manchuria, operating the Chinese Eastern Railway there. While the Soviets had only approximately one hundred thousand military personnel east of the Ural Mountains, it was industrializing rapidly, being midway through its first five-year plan.57 In 1929, just two years earlier, the Red Army had intervened and routed the Fengtien Army when it threatened the Chinese Eastern Railway.58 Thus, the Soviet Union seemed to have both the capabilities and the will to intervene effectively when its interests in the region were threatened. While many military planners saw the risks of Soviet intervention to be relatively low, Tokyo’s position was one of caution regarding this possibility.59 This was especially the case when it came to the possible extension of hostilities north of the South Manchuria Railway.60 To try to head off this potential, the cabinet issued a resolution on September 23, ordering the Kwantung Army to stay out of the north.61 These concerns were enunciated repeatedly by Prime Minister Wakatsuki, Foreign Minister Shidehara, and even War Minister Minami over the course of the crisis.62

The cabinet was also deeply concerned with the reaction of the other great powers, the members of the Nine Power Treaty and the League of Nations at large. In the cabinet’s first meeting dealing with the crisis, on the morning of September 19, Prime Minister Wakatsuki rhetorically queried War Minister Minami: if the Mukden incident turned out to be “a conspiracy of the Japanese army, what will Japan’s position be in the eyes of the world?”63 In the decade running up to the invasion of Manchuria, Japan’s trade as a percentage of its gross domestic product amounted to an average of approximately 35 percent, and, thus, the threat of sanctions loomed particularly large.64 In an October 1 cabinet meeting, Prime Minister Wakatsuki warned that “if Japan does not act with due consideration of her international position, Japan in the end will be isolated, and this will bring an unexpected misfortune upon the nation.”65 The following week, in response to the suggestion of setting up an autonomous regime in Manchuria, Wakatsuki said doing so would be “a violation of the Nine Power Treaty, and then the whole world will be our enemy. In view of the present economic situation, we are practically forced to be isolated.”66 Similar concerns of economic sanctions and diplomatic isolation were shared by Foreign Minister Shidehara, Emperor Hirohito, and central army authorities as well.67

Thus, the geopolitical risk associated with further acquisitions in Manchuria led the government in Tokyo to oppose the invasion of Manchuria and to try to rein in the Kwantung Army. However, the very fact of the Kwantung Army’s unauthorized expansion constrained the cabinet in two important ways that would ultimately make withdrawal impossible to achieve. First, the Kwantung Army’s quick successes in its operations in Manchuria firmly sunk the costs of the acquisition. By September 21, the Kwantung Army had secured all major centers along the South Manchuria Railway and had occupied Kirin, a city roughly 100 km east of Changchun. In mid-November, the army moved on Tsitsihar in north Manchuria, and, in early January 1932, it took Chinchow and Shanhaikwan, completing the occupation of the south right up to the Great Wall.68 On February 5, Harbin was occupied, effectively completing the conquest of Manchuria.69 These territories had been acquired at remarkably little cost. In taking the barracks at Mukden, the Kwantung Army suffered only twenty-four casualties in an assault on as many as 10,000 personnel.70 The Kwantung Army suffered 155 casualties in its occupation of Changchun just days later.71 Kirin was then occupied without firing a shot.72 And Tsitsihar was taken over the course of less than two days, and the entry into the city itself was bloodless.73 Overall, the Imperial Japanese Army suffered 2,530 killed in its conquest of Manchuria, a fairly low figure given the area of the territory acquired, the size of the forces it faced, and the kinds of casualties it had suffered in past wars.74 (See figure 6.1.)

These successes were facilitated not only by the rigor of the Kwantung Army’s training and the detail of its planning but also by the Fengtien Army’s policy of nonresistance and the lack of intervention by the Soviet Union and the other great powers. The Soviet Union was far more concerned with domestic political issues, and with its European flank, than with events in Manchuria at this time, and Japan’s leaders came to understand this.75 In a strong signal of their defensive intentions, the Soviets proposed a nonaggression pact with Japan in December 1931 as the invasion was ongoing, an offer Tokyo ultimately passed up.76 And the reactions of the United States and the other great powers were similarly muted.77 This was, in part, because they were still preoccupied with the continuing fallout of the Great Depression of 1929. However, it is also clear that within a few days of the invasion, both the United States and the United Kingdom were aware that the central government had lost control of the Kwantung Army, and this, too, likely tempered their responses.78 This absence of effective resistance severely weakened the arguments for restraint from leaders in the capital.79 And it made the hawks, who had doubted the great powers’ willingness to intervene, look prescient. Thus, the reservations of leaders in the capital would dissipate as the conquest progressed, as, one by one, their greatest fears failed to materialize.80 And with every act of defiance committed by the Kwantung Army, the cabinet, Diet, and imperial court was rendered progressively weaker.81

A second important constraint that would crop up and make retrenchment difficult was the overwhelming support among the public and the press that the Kwantung Army’s actions received.82 In a severely economically depressed Japan, the idea of a resource-rich Manchuria as an “economic lifeline” came to be widely accepted among the populace, particularly in rural areas.83 In fact, Ishiwara, Itagaki, and other conspirators had deliberately sought to shape elite and public opinion at home and in Manchuria in the months and weeks leading up to the invasion.84 They did so by producing and distributing pamphlets, magazines, and books, and organizing speaking tours throughout Japan, with the support of some members of the General Staff in Tokyo.85 They were greatly aided in this by the Manchurian Youth League, a nationalist organization formed with Kwantung Army backing in 1928. The league traveled widely throughout Japan in the months preceding the invasion, arguing for a stronger policy in Manchuria.86

Once the invasion was underway, the Kwantung Army continued to foster this support, setting up a propaganda office, holding regular briefings, distributing pamphlets, and broadcasting patriotic songs and messages over the radio.87 All of this public support generated significant pressure on the cabinet to protect and defend Japanese nationals and soldiers in Manchuria. And the Kwantung Army would exploit this, using “false flag” operations as an excuse to occupy Manchurian cities. In Kirin, for instance, Kwantung Army agents were dispatched to foment unrest, which was then used as an excuse to invade, in September 1931, to “protect” Japanese property and nationals.88 A similar strategy was attempted in Harbin that same month, though in this case the cabinet stood firm for the time being.89 As time went by, the public and the press became increasingly unified behind the Kwantung Army’s invasion of Manchuria. In the words of one scholar, the Japanese press had voluntarily turned itself into a “propaganda machine for the army.”90

The overwhelming public support for the army’s actions created easy avenues of attack for the cabinet’s opponents in and out of government, severely raising the risk of domestic political punishment. The first attacks on the Minseito Party cabinet of Prime Minister Wakatsuki came from the opposition Seiyukai Party and its leaders.91 Already prone to see the ruling cabinet’s Shidehara Diplomacy in China as “weak-kneed,” the opposition was quick to capitalize on the opportunity of a popular war in Manchuria to attack.92 And these attacks ultimately paid off at the ballot box. While Wakatsuki came to office with his Minseito Party holding a large majority in the Diet, the party would be absolutely trounced in the February 1932 elections, losing 127 seats (and their majority) to the opposition Seiyukai.93 This was a strong endorsement of the Kwantung Army’s actions in Manchuria, and a vote against the Minseito cabinet’s cautious policy in China.94

Other forms of potential domestic political punishment faced by the cabinet were far more severe. For instance, the Sakurakai (Cherry Blossom Society), a secret ultranationalist organization within the Imperial Japanese Army, staged coup attempts against Minseito cabinets in March and October 1931.95 Since their formation in September 1930, they had advocated for a much more forceful policy in Manchuria and the establishment of a totalitarian government in Japan.96 These plots were a grave concern for leaders in Tokyo. In late September 1931, a close adviser to the emperor pointed out that the “fact that their plot has succeeded in Manchuria will surely give a certain element of the Army the confidence that they can also do the same in Japan, and there lies the real danger.”97 While in neither case did the coup succeed, these were clear expressions of opposition to Shidehara Diplomacy in China. Coup rumors and threats continued to swirl among leaders in Tokyo through the early months of 1932 and would have a chilling effect on the cabinet, leading the foreign minister and others to soften their resistance to the Kwantung Army’s insubordination.98

Besides coup plots, there were also assassination attempts. In November 1930, Prime Minister Hamaguchi Osachi was shot by an ultranationalist who was opposed to his signing of the London Naval Treaty, reducing Japan’s naval armaments.99 While Hamaguchi would survive the initial attempt, he would never recover his health and died from related complications less than a year later.100 It would also come to light that the abortive October 1931 coup included planned assassinations of both Prime Minister Wakatsuki and Foreign Minister Shidehara, among others.101 And Inukai Tsuyoshi, the Seiyukai party leader who would succeed Wakatsuki as prime minister in December 1931, would himself be killed in office in May 1932. His residence was stormed by young officers of the Imperial Japanese Army and Navy who, despite the Inukai cabinet’s more forward-leaning policy in Manchuria, were opposed to his attempts to subject the military to stricter civilian control.102 Attempted and successful assassinations on other business and political elites in these years had profound effects on Japanese leaders and contributed to a general climate of fear among Tokyo’s leadership.103 As Shimada notes, “The spectacle of army terrorism was reducing the cabinet, and even the supreme command, to impotence.”104

One final additional means by which the Kwantung Army manipulated the cabinet in Tokyo and enhanced their relative influence was through veiled threats of secession from Japan. In mid-October 1931, rumors began to circulate in government and military circles in Tokyo of the possibility of the Kwantung Army seceding and independently establishing itself in Manchuria.105 While the origin of these rumors is a matter of dispute, during the invasion of Manchuria, Ishiwara Kanji is believed to have sent a telegram to Tokyo stating that “if the Japanese government constantly interferes . . . [then] we will have to break the glorious history of the Imperial army and separate ourselves from the empire.”106 The purpose of these threats was to put pressure on the government in Tokyo to support the Kwantung Army’s expansionist policy in Manchuria.107 And whatever doubts may have existed in Tokyo as to the credibility of these threats, central army authorities took them seriously enough to investigate carefully.108 Thus, a combination of the Kwantung Army’s sinking the costs of expansion and fervent support for their actions among the press and the public would effectively bind the hands of leaders in Tokyo, making restraining or withdrawing the Kwantung Army incredibly difficult.

Wakatsuki’s Dilemma

Thus, the cabinet in Tokyo faced conflicting pressures in Manchuria. On the one hand, the early expected risk of Russian intervention and the potentially adverse reactions of the other great powers pushed them to tighten the reins on the Kwantung Army. On the other hand, the Kwantung Army’s initial successes and the overwhelming support their exploits received from the public and the media pressed leadership to swim with the tide and accept their faits accomplis. The cross pressures created by these conflicting pressures presented the cabinet with a deeply distressing dilemma. The prime minister himself was discouraged, even despairing. As he told a secretary to the court the day following the invasion, “Under these circumstances I am quite powerless to restrain the military.

How can his majesty’s military act without his sanction? What can I do? . . . I am in serious trouble.”109 Almost a month later, the prime minister was even more exasperated. As he told this same secretary on October 12:

The present day situation is exceedingly unfavorable. I have endeavored, wholeheartedly, to better the international relations of Japan. At Cabinet meetings, I have frequently requested the War Minister to carefully keep under control the actions of Japanese troops in Manchuria . . . However, Army officials act counter to this order whenever they find it convenient to do so. This sort of thing has its immediate repercussions in the League of Nations and the outcome is that the League feels it has been completely betrayed . . . I don’t know what to do about the matter. Under this sort of conditions, I can’t continue in my office indefinitely, but I can’t resign right now. It is an exceedingly difficult situation.110

The dilemma facing the cabinet weighed heavily on Foreign Minister Shidehara as well. The Mukden consul general met with Shidehara on November 16 and described him as follows: “His demeanor seemed discouraged, disappointed, and dejected. One could recognize without words how much he was suffering in this unprecedented emergency. I had unbound sympathy for him in his predicament.”111

Tokyo Decides

These pressures were ultimately more than the leadership could bear. On December 12, 1931, having lost the confidence of the imperial court, the Wakatsuki cabinet fell.112 It was replaced the following day, as noted earlier, by the opposition Seiyukai cabinet of Inukai Tsuyoshi. By this point it had become clear that the geopolitical risk involved in acquiring Manchuria was far less severe than initially supposed, allowing the Inukai cabinet to adopt a more forward-leaning Manchuria policy than its predecessors. The rise of Inukai effectively ended serious resistance by the central government to the Kwantung Army’s conquest of Manchuria.113

Yet, the Inukai cabinet faced their own struggles with the Kwantung Army. Inukai, a hawk on Manchuria policy, was nonetheless concerned about the reactions of the great powers and was intent on restoring discipline within the army.114 This was not just his preference; he was under orders to do so. As Emperor Hirohito admonished him upon his appointment as prime minister, “the Army’s meddling in domestic and foreign politics, endeavoring to have its way, is a situation which we must view with apprehension for the good of the nation. Be mindful of my anxiety.”115 The emperor again warned Inukai in late December to “maintain international trust” and to be aware of the impact the Kwantung Army’s actions were having on international affairs.116 Inukai would try, and he would worry. As he wrote to a senior army official in a February 15, 1932, letter:

What is most worrisome is that the will of the senior officers is not thoroughly observed by their subordinates. For example, the action in Manchuria seems to have been brought about by the united power of the field-grade officers, who made their superiors acquiesce automatically. . . . It is feared that it might become customary to act single-mindedly upon the belief that should those who hold direct command over regiments unite and cause a disturbance, the superiors would finally give ex post facto approval to all matters, and that [such a trend] might create a major change in military control and discipline. . . . Therefore I wish the elders of the army to take remedial measures now, when the malady has not yet spread widely.117

But Inukai, too, was only minimally in control of events in Manchuria. With popular opinion surging behind the Kwantung Army, and with an incredibly hawkish cabinet advising him, in the long run Inukai had few options but to swim with the tide.118 The invasion and occupation of Manchuria was an accomplished fact.

The Japanese puppet state of Manchukuo was proclaimed on March 1, 1932, marking the end of Japan’s conquest of Manchuria. It was nominally independent but was, in fact, under the strict control of the Kwantung Army. As an internal Kwantung Army document from January 1932 put it, Manchukuo would adopt “the external form of a constitutional, republican government . . . but maintain the internal reality of a centralized dictatorship imbued with the political authority of our empire.”119 The Inukai cabinet would initially hold off on formally recognizing Manchukuo; notably, out of concern for the reactions of the other great powers.120 Though, this, too, would occur in September 1932 with the same sense of inevitability that had characterized the entire affair. Inukai’s cabinet would be the last party-led government in prewar Japan, and his May 1932 assassination was an important milestone in Japan’s turn toward military dictatorship.

Despite their insubordination, Itagaki Seishiro and Ishiwara Kanji would be generously rewarded for their actions in Manchuria and would continue to rise through the ranks of the Imperial Japanese Army. Itagaki would be promoted to the rank of general and would serve as chief of staff of the Kwantung Army and the China Expeditionary Army, as well as a short stint as war minister. Ishiwara, for his part, was given the Order of the Golden Kite, third class, for the invasion of Manchuria, and was promoted to full colonel ahead of most of his Army Staff College classmates. He retired from the Imperial Japanese Army at the rank of lieutenant general in March 1941, just months before the outbreak of the Pacific War.121 No individual had been as important to the Japanese conquest of Manchuria as Ishiwara. With the aid of Itagaki and others, he had been deeply involved in nearly all aspects of the invasion, from the Mukden Incident in September 1931 to the establishment of Manchukuo in March 1932. As Itagaki had told a friend a few weeks into the invasion, it was not Japan’s war, or even the Kwantung Army’s war, it was “Ishiwara’s war.”122

Historical Background

Fiume in the early twentieth century was a bustling small port city of approximately fifty thousand in the northeast corner of the Adriatic Sea, where the Dalmatian coast meets the Istrian Peninsula. As a port city nestled between Austrian and Hungarian territories within the Dual Monarchy’s multiethnic empire and sitting just 120 km from Italy’s pre–World War I border, Fiume’s political and cultural identity had long been diverse and cosmopolitan. A 1910 census recorded the population as 49 percent Italian, 31 percent Slav, and 13 percent Magyar (Hungarian), with a smattering of Germans and other ethnicities.123 Fiume was strategically and economically important as a regional economic hub with rail lines connecting Belgrade, Prague, Budapest, and Zagreb to the coast.124

With the collapse of the Austro-Hungarian Empire in the aftermath of World War I (1914–18), the Allies established in the city a joint occupation consisting of American, British, French, and Italian forces.125 The Italians, for their part, were intent on ultimately annexing the city. While Fiume was not promised to Italy in the 1915 Treaty of London, which had conditioned Italy’s entry into the war, many surrounding territories had been promised to it, and the city would have added to Italy’s growing dominance of the Adriatic. After 462,000 dead, 954,000 wounded, and three-and-a-half years of fighting, many in Italy only hoped to receive what they saw as their due.126 Yet, the status of Fiume would have to await the negotiations of the Paris Peace Conference, which were to begin in January 1919. And, as it would turn out, this port city of roughly 30 km2 would be a major stumbling block in Italy’s negotiations with its allies at Versailles.127

Italy’s territorial ambitions in Fiume, and beyond, ran headlong into the Allies’—and particularly, the American president Woodrow Wilson’s—interest in what was known as “national self-determination,” the idea that nationalities should have the right to freely choose their sovereignty. Wilson flatly refused to accept the Treaty of London, arguing in the first of his Fourteen Points that only “open covenants . . . openly arrived at” should be recognized in the postwar international order. The president also claimed, for Italy specifically, that adjustment of its borders “should be effected along clearly recognizable lines of nationality,” a clause that weakened many of Italy’s territorial claims, Fiume included.128 While, as noted earlier, Fiume proper had a plurality of Italians in its population, if the adjacent and deeply interconnected suburb of Suzak was added, the plurality went to the Slavs. And estimates based on political party affiliation in Fiume suggest that a narrow plurality favored annexation to the newly formed kingdom of Yugoslavia, rather than to Italy.129 In short, Italy’s territorial claims rested on shaky ground, and the American president simply would not budge.

The conflict over Fiume at Versailles would spark protest and unrest both in Italy and in Fiume itself. Italy faced severe economic hardship in the aftermath of the war. It owed the Allies the equivalent of $3.5 billion in wartime loans, saw greater inflation than anywhere in Europe with the exception of Russia, and had widespread unemployment resulting from demobilization.130 This was the beginning of Italy’s Biennio Rosso (Two Red Years), a period marked by mass strikes, land seizures, factory occupations, and violent conflict between socialist, anarchist, and nationalist political organizations.131 Fiume, as well, saw roving nationalist gangs, deadly riots, and armed clashes between Italian and Allied forces in the spring and summer of 1919.132

Amid the political turmoil of these months, Fiume emerged in Italy as a potent symbol of Italian pride and honor; and Italy’s failure to acquire it became a symbol of national humiliation. As the Italian prime minister Vittorio Orlando told his colleagues at Versailles in April 1919, “Italian public opinion is very excitable. I am doing what I can to calm it; but the consequences of disappointment of this kind would be very grave.”133 The nationalist press pushed for the annexation of Fiume, while placards were posted and walls painted throughout the country with similar demands. When Orlando eventually withdrew from the conference over disagreements regarding Fiume, he was greeted in Rome with cries of “Viva Fiume!”134

Prominent among those agitating for the annexation of Fiume was the Italian poet, playwright, novelist, and philosopher, Gabriele D’Annunzio. D’Annunzio was Italy’s most esteemed literary figure of the time—so broad was his fame and success that most Italians simply referred to him as Il Vate, or “The Poet.”135 He was notoriously eccentric and famously promiscuous, carrying on dozens of affairs throughout his adult life.136 He was also an ardent nationalist, had an immensely inflated sense of his own importance, and had a great deal of contempt for the political class in Rome.137 Besides writing and womanizing, D’Annunzio also had a short parliamentary career and a roving commission between Italy’s Army, Navy, and Air Force during World War I.138 Back in Italy after the war, at the age of fifty-five, he became engaged in nationalist causes and organizations and was unmatched in his ability to whip crowds up into a frenzy with rousing speeches and pungent language. D’Annunzio was the actor on the periphery who would aim to drag Italian leaders in Rome into acquiring Fiume.

Fiume

Within a week of the armistice on November 11, 1918, D’Annunzio was contacted by local Italian authorities in Fiume seeking his aid in facilitating its acquisition.139 Planning for some sort of march on the port city began in earnest between the winter of 1918 and the early spring of 1919.140 Financing for the operation, largely through nationalist organizations and Italian industrialists, was secured in the spring of that year, as were the forces necessary to carry it out. By May 1919, D’Annunzio was enlisted to lead the march. Rumors began to circulate in Rome that summer that a conspiracy involving Fiume was being cooked up.141 The Italian prime minister Francesco Nitti repeatedly reminded his generals in the region that the government’s policy was to avoid any precipitous action in Fiume at all costs.142 Naturally, D’Annunzio did nothing to calm nerves in Rome, asking, in a public address on May 7, “Down there, on the roads of Istria, on the roads of Dalmatia, do you not hear the footsteps of a marching army?”143

Prime Minister Nitti’s grasp of the Italian military was not exactly firm. Civilian governments in Italy had relatively weak control of the military in these years for several reasons.144 First, for most of its existence, the Italian military had been large and central to its political system. The final stages of Italian unification—known as the Risorgimento—had been accomplished through a series of wars between 1861 and 1870, which started off the unified Italy with a large military. In the years that followed, political leaders relied on the military, rather than on a large civilian police force, for domestic security and the maintenance of order, which further entrenched its position.145 Second, as a constitutional matter, the Italian military was under the control of King Victor Emmanuel III, not the prime minister or cabinet.146 While the Italian parliament voted on military budgets and conscription requirements, the military itself was loyal to, and was effectively run by, the monarchy.147 And third, Italy’s significant losses in World War I, perceptions that Italy had been subjected to a “mutilated peace” at Versailles, and the rapid demobilization of the Italian military that occurred in the aftermath of the war created significant resentment toward civilian leaders among military officers and the rank and file. The Italian Army went from 3.8 million personnel in December 1918 to 1.6 million just seven months later, and it would continue to be rapidly cut in the months that followed. This was accompanied by a suspension of promotions, lower pay rates, and increased requirements for pensions.148 All of these factors—historical and constitutional—greatly hampered the cabinet’s control of the military and made unauthorized peripheral expansion particularly likely in this case.

The conquest of Fiume ultimately was set for September 11, 1919.149 D’Annunzio was to travel from Ronchi in Italy’s northeast corner, across the Istrian Peninsula, arriving in Fiume in the morning, a distance of about 100 km. He began with just 186 members of the Italian military but had a Fiuman militia that was to join his forces upon arrival.150 He and his grenadiers set off from Ronchi at midnight on the eleventh, D’Annunzio leading the column in a bright red Fiat 501.151 Before departing, D’Annunzio sent then-nationalist journalist Benito Mussolini the first of hundreds of letters the two would exchange in the months that followed: “My dear companion, the die is cast. I depart. Tomorrow morning I will conquer Fiume. May the God of Italy help us.”152

Along the way, D’Annunzio and his forces were met by numerous Italian soldiers who were under orders to stop, and even fire on, the Poet if he tried to pass. However, most were sympathetic to his cause and instead cheered him as he passed, with many even abandoning their posts to join his column.153 By the time he arrived at the outskirts of Fiume on the morning of the twelfth, D’Annunzio was at the head of between two thousand and twenty-five hundred personnel, with dozens of trucks and armored vehicles.154 D’Annunzio was met outside of the city by the commander of Italian forces in Fiume, who implored him to turn back. When D’Annunzio refused, the general saw no other option and let D’Annunzio and his forces into the city.155 By noon the conquest of Fiume was complete. With characteristic grandiosity, D’Annunzio would refer to his acquisition of Fiume as the Sacra Entrada (Sacred Entrance).156 That evening, at 6:00 p.m., D’Annunzio appeared on the balcony of the governor’s palace and addressed the crowd: “Italians of Fiume! . . . I proclaim: I, a soldier, a volunteer, a wounded veteran of the war, believe that I interpret the will of the people of Italy in proclaiming the annexation of Fiume!” a declaration that was met with an eruption of celebration.157

Rome Reacts and Decides

When Prime Minister Nitti learned of events in Fiume, he was visibly shocked and absolutely livid, forcefully pounding his fist on his desk.158 He was not surprised by D’Annunzio’s attempt to take Fiume; he had, after all, been receiving reports on this possibility for months. What surprised him was the Poet’s success—and the defection of thousands of Italian soldiers that it had required. The following day, on September 13, Nitti made a statement before the Italian parliament, expressing publicly his anger and disapproval and assuring his colleagues that “the Government had taken appropriate measures.”159 A few days later, on September 18, Nitti’s cabinet ordered a blockade of Fiume, with the Italian Third Army surrounding the city by land and the Italian Navy blocking the entrance to its harbor.160 Then Nitti took the extraordinary step of requesting that Victor Emmanuel III call a meeting of his privy council, which was held a week later, on September 25. It was attended by leading Italian political figures, top military leaders, the king and his closest advisers, and the prime minister, and they were unanimous in their opposition to D’Annunzio’s unauthorized conquest.161 With this strong backing, the prime minister returned to parliament and called for snap elections to be held in November. The leadership’s position was firm. D’Annunzio’s fait accompli could not be accepted.

The primary reason for the government’s strenuous opposition was the geopolitical risk associated with accepting the city. As noted earlier, D’Annunzio’s fait accompli occurred in the context of a joint Italian occupation of Fiume alongside the United States, Britain, and France. After receiving assurances from Rome that the matter would be dealt with expeditiously, Italy’s great power allies agreed to have their forces make a hasty exit, though no one was pleased with the situation.162 Prime Minister Nitti initially thought that the situation could be used to Italy’s advantage and that D’Annunzio’s escapade might help strengthen his position in negotiations over Fiume.163 Yet, while the British and French were somewhat more sympathetic, President Wilson was absolutely firm: D’Annunzio had to go, and Fiume was to become a free city under League of Nations auspices.164 The president made his view clear to the Italian foreign minister Tommaso Tittoni just days after D’Annunzio’s march on Fiume, and Britain, France, and the United States penned a joint memorandum on December 9, pointing to the “urgent necessity” of creating an independent Fiuman state.165 When he delivered this memorandum to the Italians, the French president Georges Clemenceau noted that “there could be no peace in Europe till this question was settled.”166 It was clear to Nitti that the Allies were in no mood to make concessions.167

And Italy had few other options. For one, it was militarily much weaker than any of its allies, let alone all three of them together, so it could not exactly stand and fight.168 But more importantly, in its dire postwar economic state, it needed its great power allies, and the United States in particular, more than ever. As noted earlier, Italy had borrowed billions from its allies, and the United States was continuing to extend it credit. Rupture with the United States at this point would have meant true economic calamity for Italy, something Prime Minister Nitti, as a trained economist, understood only too well.169

However, the decision was complicated by the fact that D’Annunzio’s conquest of Fiume itself generated constraints that made withdrawal difficult from the perspective of Rome. The first was the simple fact of his success. Fiume had been acquired by D’Annunzio at no cost in human life, sinking the costs of acquisition for leaders in the capital. But second, and more importantly, the conquest of Fiume had the backing of a significant portion of the Italian military, as well as the press and public more broadly. A perceived risk of popular backlash led Prime Minister Nitti to soften the blockade of Fiume after just a few days.170 Its resulting leakiness meant that soldiers, sailors, and air personnel continued to desert to Fiume in droves to enlist in D’Annunzio’s army and join the cause. The Fiuman forces numbered as many as nine thousand at its peak, and, at a certain point, D’Annunzio had to begin turning military defectors away for lack of accommodations.171 The press, too, seized on the march on Fiume, painting D’Annunzio as an Italian folk hero.172 And important sections of the public backed D’Annunzio’s venture as well.

The domestic political threats that leaders in Rome faced were not merely electoral. For instance, former prime minister Vittorio Orlando claimed during the Paris peace negotiations that a secret society had pledged to assassinate him if he returned without Italy’s irredentist claims.173 Furthermore, in June 1919, a nationalist coup plot seeking to overthrow the Italian government was uncovered and broken up.174 And there were rumors circulating about assassins sent from Fiume to kill Prime Minister Nitti and Foreign Minister Tittoni.175 Thus, members of the Nitti cabinet were aware that rising nationalist sentiment in Italy represented a threat not only to their electoral fortunes but to Italian political institutions and even to their lives.176

This combination of strenuous allied opposition and public and military support created a real dilemma for the Nitti cabinet.177 President Wilson was unwilling to give an inch on Fiume and had significant economic leverage over the prime minister. Yet, there were military and nationalist forces pressing him on, threatening not only his prime ministership but possibly his life. It was as if the ground beneath Nitti’s feet, as he put it, “had been mined.”178 Under these trying circumstances, the prime minister adopted a patient and delicate strategy of assuring the Allies that Italy would clean up the Fiuman mess, while negotiating with D’Annunzio to resolve the situation.

D’Annunzian Fiume and Its End

For the fifteen months of its existence, Fiume under Gabriele D’Annunzio reflected all of the eccentricities of its leader. The outlaw city attracted curious visitors from all over Europe—gangsters and prostitutes, politicians and war heroes, famed musicians and Nobel Prize–winning scientists.179 There were parades and political rallies by day and banquets and torchlit processions by night. And the Poet was at the center of it all, addressing throngs of admirers, glad-handing his loyal supporters, and hosting debaucherous soirees at the governor’s palace. Yet, there was a much darker side to it all as well. For D’Annunzio would turn out to be not only a hopeless administrator but also a deeply authoritarian leader.180 The Poet embraced a charismatic form of personalistic rule, in which he was entirely above the law and dissent was made a capital crime.181 Before long, there were security forces on every corner, the prisons in Fiume began to overflow, and extrajudicial expulsions, kidnappings, and killings became commonplace.182 In his raucous public addresses, D’Annunzio used a dialogical style that would later become associated with Italian and German fascism, employing violent and vulgar language and having crowds hurl obscenities at his political enemies in Rome. And Benito Mussolini was watching carefully, visiting the city as a journalist on a number of occasions and corresponding with D’Annunzio regularly.183 Reflecting on this period, the famed Italian diplomat Carlo Sforza would refer to D’Annunzio as the true “inventor of fascism.”184

It did not take long for the D’Annunzian spectacle to begin to lose its luster. In November 1919, Italians went to the polls where Nitti was confirmed in his leadership, the nationalist party won just a handful of seats, and not a single fascist candidate was elected to office.185 In the prime minister’s view, as he communicated to D’Annunzio shortly after the election, the results were a strong indication that Italians were “against any adventurous policy” in Fiume and beyond.186 And while it would be more than a year of on-again, off-again negotiations before the Poet was finally removed, it was clear by early 1920 that his days in Fiume were numbered. Francesco Nitti would resign from office in June 1920 to be replaced by Giovanni Giolitti, a more decisive politician who soon entered into negotiations with the Kingdom of Yugoslavia over the fate of Fiume.187 The resulting Treaty of Rapallo, signed on November 12, 1920, established the Free State of Fiume as an independent city-state between Italy and Yugoslavia. Despite the treaty passing by overwhelming majorities in the Italian parliament, D’Annunzio clung to power in Fiume and continued to call for its annexation to Italy.188

But enough was enough. Prime Minister Giolitti had to take action. On December 20, 1920, he sent D’Annunzio an ultimatum, demanding his exit. In response, the following day D’Annunzio declared war on Italy. On Christmas Eve, the Italian Army and Navy were ordered into action. After a few dozen casualties were taken by both sides, the Italian Navy cruiser Andrea Doria fired two shells on D’Annunzio’s palace on December 26. This was ultimately decisive. On December 28, D’Annunzio left Fiume for good.189

D’Annunzio would not be punished for his open defiance of Italian authorities. Despite his reduced stature, he was still deemed too popular among important segments of the Italian public. This, and the fact that Giolitti chose December 24 for D’Annunzio’s ouster to minimize press and public attention, clearly indicate the popular constraints under which Italian leaders felt they were operating.190 D’Annunzio may have ultimately failed in his greatest ambitions, but his example would play an important role in the success of some of his fascist descendants in Italy, Portugal, Spain, and Germany.191

Less than two years later, Benito Mussolini seized power with his “March on Rome,” establishing a fascist dictatorship in Italy and adopting many of the repertoires of rule he observed in Fiume. D’Annunzio, for his part, effectively retired to his home on Lake Garda at the government’s expense. He evidently still had his hypnotic charm, and Mussolini saw him as a potential political threat in his emerging fascist movement. As Il Duce explained, “When you have a rotten tooth, you have two possibilities open to you: either you extract the tooth or you fill it with gold. With D’Annunzio I have chosen the latter treatment.”192

This chapter has presented comparative cases of inadvertent expansion and nonexpansion in Manchuria and Fiume. Both cases strongly support the theory of inadvertent expansion presented in chapter 1. First, in both cases unauthorized peripheral expansion resulted from inadequate monitoring and control over agents on the periphery—the Kwantung Army in Manchuria and D’Annunzio and important segments of the Italian Army in Fiume. Second, in both cases peripheral expansion generated constraints that made it difficult for leaders in the capital to easily withdraw—in both Japan and Italy, the most important of these were domestic political risks and costs associated with doing so. And third, in both cases the ultimate decision of whether to accept the fait accompli was crucially determined by the geopolitical risk associated with doing so. In the case of Japan in Manchuria, the perceived geopolitical risk pushed Prime Minister Wakatsuki to work strenuously to put a halt to the conquest and to rein in the Kwantung Army. Yet, once it became clear that such risks would not come to fruition, his government was replaced by the more pliant Inukai cabinet, and the invasion of Manchuria moved forward. In the case of Italy in Fiume, the strong stance of Britain, France, and, particularly, the United States in opposition to D’Annunzio’s conquest gave the Italian government little alternative but to roll it back, which it ultimately did by force. In both cases, what I referred to in chapter 1 as the “dilemma of inadvertent expansion” was illustrated powerfully—the agonizing situations in which leaders simultaneously face severe domestic political costs associated with territorial withdrawal as well as significant geopolitical risk associated with territorial acquisition. Showing how leaders navigate these perilous circumstances has been an important aim of this chapter.