Japanese depiction of the Battle of Shiroyama. Saigō Takamori can be seen in red and black uniform directing his troops in the upper right corner.
Although angered by the students’ rash action, upon learning of the raid on the arsenal, Saigō met with his lieutenants and authorized preparations for war. His intention was to topple the ruling oligarchy headed by Okubo. Believing that opposition to the oligarchy was sufficiently great to realize this goal with a minimum of force, he tried with vague public announcements to maintain the appearance of legality. On February 9, Saigō, Shinohara, and Kirino formally notified Governor Ōyama that they would “shortly leave the prefecture,” taking with them “a number of former troops.” Ōyama obliged by announcing that Saigō was proceeding to Tokyo in order to investigate the plot against his life, that a large force of former government troops would accompany him, and that the emperor had been fully informed. Although clearly in rebellion, Saigō and his officers wore their old Imperial army uniforms, and Saigō issued orders as commander in chief. At first he even refused to enlist volunteers from neighboring prefectures in order to avoid charges of having entered into a confederacy.
By using the shigakkō system, mobilization was carried out with great speed. By the end of the first week in February, even before Saigō had notified Ōyama of his intentions, armed men from the schools had begun to assemble in Kagoshima. Within a week, the vanguard and a four-thousand-strong First Division completed mobilization and departed the capital; soon after, the Second Infantry Division, rear guard, artillery, and finally Saigō’s bodyguard started the march north. On February 20 the Satsuma army crossed into Higo Prefecture. Defeating an advance party of troops from the Kumamoto garrison, they marched into the city and laid siege to the former domain castle that now served as the army headquarters. Almost immediately, two local bands of shizoku – the Gakkōto and the Kyōdōtai – came over to Saigō’s side. It was an auspicious beginning.
After a final attempt to dissuade Saigō, the government mobilized for war. Three thousand troops from the Tokyo garrison were immediately transported by ship to Kobe, and the Osaka and Hiroshima garrisons proceeded directly to Fukuoka in northern Kyushu. Prince Arisugawa assumed command of the hastily assembled army and immediately dispatched two divisions to block further advance by the rebel force. Thereafter, the Imperial army steadily gained the upper hand. Despite being outnumbered, the Kumamoto garrison did not capitulate, and contrary to Saigō’s expectations, it repulsed repeated assaults. Meanwhile, the government not only mobilized the standing army and called up reserves, but it also enlisted thousands of shizoku volunteers as “police” auxiliaries. Fresh units arrived daily, and the well-equipped Imperial army counterattacked. After several days of fierce fighting, on March 20, the government forces captured the key pass at Taharazaka. Both sides suffered heavy losses. The Satsuma army executed an orderly retreat and set up a new line of defense. During the next two weeks, the Imperial army assaulted this line while additional units advanced on Kumamoto from the south. Threatened with encirclement, Saigō abandoned the siege and retreated. Although not yet defeated, the rebellion had clearly failed. The government controlled all of northern and central Kyushu, which severely curtailed the recruitment of additional forces. Most able-bodied shizoku from Satsuma had already enlisted, and Saigō was forced to conscript peasants, who had little incentive to fight on behalf of the shizoku, and convicted criminals, who were not likely to be dedicated soldiers. Nevertheless, Saigō and what remained of his army continued fighting through the summer. By September, only Saigō and a few hundred troops were still in the field. On September 23, 1877, confronted by a vastly larger government force in the hills north of Kagōshima, Saigō refused a personal plea from Yamagata to surrender. The next day the army attacked; the rebel force was annihilated; and Saigō committed suicide on the battlefield rather than allow himself to be captured.
Unlike previous shizoku uprisings, which were small and poorly organized, the Satsuma Rebellion severely tested the government’s capacity to wage war. To defeat the large and well-trained rebel forces, the government had to mobilize the entire standing army and reserves and enlist an additional 7,000 shizoku as “police” auxiliaries. Of the 65,000 soldiers sent to the front, 6,000 were killed in action, and 10,000 were wounded. The financial cost of prosecuting the war was staggering. Direct expenditures totaled ¥42 million, a sum equal to 80 percent of the annual budget.
However costly the war had been to the government in men and coin, the oligarchy had reason to view the outcome with considerable satisfaction. The annihilation of the Satsuma army – eighteen thousand rebel troops were killed or wounded, and Saigō and his lieutenants died in battle or by suicide – eliminated the only shizoku force capable of threatening the central government. Moreover, the performance of the Imperial army vindicated the government’s decision to adopt universal military conscription. Approximately two-thirds of the Kumamoto garrison were conscripts; although outnumbered and short of supplies, they withstood a fifty-day siege, thereby preventing Saigō’s army from moving into northern Kyushu. To say that the victory proved commoner conscripts to be superior to samurai is nevertheless an oversimplification, for the majority of government soldiers were also shizoku. The Imperial Guards and “police” auxiliaries, which were exclusively shizoku, bore the brunt of the fiercest fighting. Nevertheless, the overall efficiency with which the government prosecuted the war amply demonstrated the advantages of military modernization and particularly of centralized command and national recruitment. The Satsuma Rebellion marked the final attempt by disaffected shizoku to overthrow the Meiji government. What does the failure of the greatest revolt reveal about the limits of shizoku rebellion?
After the first three weeks, Saigō’s army was outnumbered and outgunned. The core of the Satsuma army consisted of six infantry regiments of 2,000 men each, in addition to artillery and the rearguard. After entering Kumamoto, they were immediately joined by two bands of local shizoku, the politically conservative Gakkōtō and the Kyōdōtai (an association of nominally progressive shizoku affiliated with the Popular Rights movement) and an additional 5,000 volunteers from nearby provinces. Thus, at peak strength the rebel forces numbered no more than 22,000.
In contrast, the government initially fielded an army of 33,000 and sent an additional 30,000 before the end of the war. The rebel army was short of guns and munitions throughout most of the war. The looting of the Kagoshima arsenal secured an initial stock of guns and powder, but when the army marched out of Kagoshima in February, each soldier carried only one hundred bullets, enough for two to three days of combat. Attempts were made to purchase arms abroad, but even if the negotiations had succeeded, the Imperial navy’s control of the sea would probably have prevented their delivery. The government occupied Kagoshima in April, cutting off the overland supply of munitions that were manufactured locally.
Tactical miscalculations, and especially the decision to lay siege to Kumamoto Castle, undoubtedly hastened defeat. Because of the strategic advantages enjoyed by the government forces, it was unrealistic to expect victory in a prolonged war. The best hope lay in a rapid advance to link up with sympathizers and create the impression of success before the government had time to mobilize. If Saigō had proceeded directly to Fukuoka and thereby carried the rebellion into northern Kyushu, additional groups might have declared for the rebellion. As the Shimpūren, Akitsuki, and Hagi uprisings of the previous autumn had demonstrated, shizoku disaffection was intense, and the decision of the “liberal” Kumamoto Kyōdōtai to join Saigō when it appeared that the rebellion might succeed suggests that the timing was indeed a critical factor. How long would Itagaki and the Risshisha have held off if Saigō’s army had crossed the Inland Sea? One does not have to be a cynic to recognize that opportunism was a powerful determinant of political alignments in the early Meiji period. But by failing to stay on the offensive, Saigō forfeited his one chance of victory – a general uprising against the Tokyo government.
However, military defeat was inevitable given the very narrow political base of the rebellion. Although Satsuma had been made into a bastion of counterrevolution, the high level of mobilization achieved there between 1874 and 1877 depended on a combination of factors unique to the domain: size and composition of the warrior population, virtual autonomy from central authority, sanction and support of prefectural officials, and Saigō’s prestige as a “founding father” of the Meiji Restoration. Because these conditions could not be duplicated, Satsuma stood alone. In fact, its strength was also its weakness, for the leadership’s parochial political loyalties stood in the way of horizontal alliances. From Saigō’s return to Kagoshima in the fall of 1873 until the attack on the Kumamoto garrison in February 1877, mobilization had been carried out entirely within the prefecture; no effort was made to encourage, aid, or link up with like-minded shizoku bands outside Satsuma. As we have seen, Saigō steadfastly held aloof, even in the autumn of 1876 when the disaffected shizoku in nearby Kumamoto, Akitsuki, and Hagi went on the offensive. When Saigō finally moved against the government and authorized sending emissaries in search of allies, the optimal moment for a general uprising had passed. Potential allies had already committed themselves to local uprisings, which produced little more than suicidal insurrections. Lacking advance notice of Saigō’s rising, in some cases sympathizers were not able to act quickly enough.
In short, Saigō failed to capitalize on his greatest asset. Arno Mayer observed that counterrevolutions are similar to revolutions in that both “feed on socio-economic dislocations, discontents and cleavages.” But despite widespread disaffection, frustration, and despair among the shizoku nationwide, the antigovernment movement in Satsuma remained stubbornly parochial. This doomed it to defeat. Because mobilization was restricted to Satsuma and did not tap shizoku discontent nationwide, Saigō’s rising produced a formidable, but geographically limited, military threat. As Goto Yasushi has observed, despite geographical proximity, similar grievances, and a common enemy, the shizoku rebels were incapable of acting in concert; dispersed and isolated, each was defeated.
The Satsuma leaders also failed to mobilize politically disaffected commoners. Class itself need not determine the scope of counterrevolutionary mobilization. Historically, recruitment to counterrevolution has depended on the degree to which “segments of different classes experienced or were apprehensive about declassment, defunctionalization, or alienation.” It is not simply displaced elites and formerly dominant classes who make up the cadre of counterrevolution, but strata of all classes, “whose fears and anxieties [are] heightened by crisis conditions.”54 However, the Satsuma leaders showed no interest in propagandizing and agitating the rural poor, who, no less than the impoverished shizoku, constituted a “crisis stratum” whose grievances and fears might have been turned against the government. As we have seen, small landholders and tenants gained little benefit from Meiji land policies; indeed, in some areas they had violently opposed its social, educational, and religious policies. How great the potential was for inciting rural uprising and what effect extensive social disorder might have had on the outcome of the Satsuma Rebellion cannot be known. But when the Satsuma army first entered Higo and laid siege to Kumamoto, the poor farmers in the Aso district carried out widespread attacks against the homes and property of local landlords and moneylenders. Although not encouraged in any way by the rebel army, they apparently assumed that the Satsuma army represented the poor and disadvantaged and associated the Imperial army with the rich. Against a background of resentment of the new land tax and allegations of malfeasance by village headmen, rumors circulated that Saigō had abolished the land tax and canceled outstanding debts. But the Satsuma army showed no interest in the peasants, except as suppliers of food, labor, and draft animals. In fact, they treated the local population so roughly that before long opinion swung against them.
Counterrevolution was defeated because none of the rebellions drew on more than a fraction of the potential recruits to the anti-govemment cause. Because the leaders did not develop models of collective action suited to mass mobilization, they failed to tap the vast reservoir of discontent among the declassed samurai nationwide. Generally, the revolts replicated the patterns of mobilization that characterized the tōbaku (overthrow the bakufu) movement: voluntary associations of like-minded men, on the one hand, and the domain as the territorial and emotional unit of collective action, on the other. However, whereas limited mobilization of shishi (men of spirit) and militant action by Chōshū succeeded in toppling the Tokugawa bakufu in the crisis conditions of the mid-1860s, a decade later the fully centralized Meiji state managed to suppress any single group that rose up against it.