After being forced from his position as head of the southern government at Canton by Lu Rongting and Tang Jiyao in mid 1918, Sun Yat-sen observed bitterly in his resignation to the southern National Assembly:
Of our country's great tribulations, none is greater than the struggle for supremacy by military men. North or south, it is as if they all belonged to a single den of badgers. Even if they assume "Constitution Protection" designations, yet none will submit themselves to the law or to the will of the people.
These remarks reflect Sun's exasperation with the growing political autonomy of China's military commanders, and in particular with the unwillingness of his own putative military allies to subordinate themselves to his political vision. Although serving specific polemical purposes, Sun's statement is also a notable example of the increasingly popular tendency at that time to blame China's political instability on military men whose unscrupulous struggles for personal power were leading China into one civil war after another. In this view, one needed to look no further for the origins of warlordism than the willful ambitions of the military commanders who were emerging as warlords.
Military commanders, not surprisingly, were less likely to attribute China's political turmoil to their own actions. On the contrary, many of them blamed the nation's military conflicts on civilian politicians. Thus in declaring their cease-fire in November 1917, Wang Ruxian and Fan Guozhang gave this explanation of the origins of the civil war in which they found themselves involved: "Politicians make use of
military men, each holding to their own views, and taking them to extremes . . .. They stir up antagonisms between north and south, and compete in selfish scheming for power." Similarly, Zhang Jingyao, responding to a call for the removal of military men from politics, saw China's disorder as resulting from "the use of military men by politicians struggling for power." In a private letter solidifying a secret alliance with Yunnan's Tang Jiyao in 1919, Wu Peifu likewise asserted that in "recent years, warlords have been used by politicians." Such sentiments should not be dismissed out of hand as political doublespeak hiding the reality exposed by Sun Yat-sen's statement. Rather, these two diametrically opposed viewpoints reveal two different, and equally important, facets of the complicated process that gave rise to warlordism.
Modern China's warlords emerged within a specific and essential historical context, namely, the continuing crisis of political authority that followed the fall of the imperial system. One of Sun Yat-sen's special strengths was his confidence that his revolutionary program was indeed the true expression of the "will of the people." Later Chinese historians have usually conceded Sun's right to speak for the Chinese nation. This should not, however, obscure the very real problem of political authority in the early years of the Republic, in which Sun was only one, and not necessarily the strongest, among various contending voices. Warlordism did not originate simply in the rejection of legitimate political authority by military commanders, but rather in the difficulty of defining which authority was legitimate. One example of this crisis of authority was the continuing constitutional controversy in the central government over the relative powers of the president, the premier, and the National Assembly. Equally important, though, was the conflict between the all-encompassing claims of the postrevolutionary central state and the counterclaims that legitimated self-governing provincial regimes. In the absence of a political consensus able to mediate fundamental conflicts over the structure and distribution of political power, fatal choices were made to resolve the crisis of political authority by force.
The militarization of politics that resulted from the crisis of early Republican political authority created the environment essential to the rise of warlordism. Under these circumstances, the protests by military men that they were "used" by politicians carry some validity. Military men were indeed called into the political arena, not to establish the military's political power, but to resolve the seemingly irreconcilable conflicts of civilian politics, and the decision to use military force came
as much from civilian politicians as from the military itself. Unfortunately, the recourse to military force not only failed to end the crisis of political authority but exacerbated it by encouraging further political fragmentation. These conditions also provided the context for expression of the ambitions of military commanders. At this point the criticism of military men for their role in the creation of warlordism gains greater weight. Contrary to their protests, military commanders were never simply the passive tools of civilian politicians or the mere pawns of political forces beyond their control. The militarization of politics that drew military commanders into the political arena also empowered them by making the military force they controlled the main determinant of political power. By taking advantage of the competing political authorities that vied for their support, military commanders were able to increase their own political autonomy. In the process, many commanders became political authorities unto themselves. In other words, they became warlords. The political issues that had originally provided the context for early Republican military conflicts were therefore increasingly subordinated to the self-interested struggles for power of warlord "badgers." The triumph of warlordism thus constituted not only the domination of politics by the military in general but the personal political domination of individual military commanders.
The rise of warlordism destroyed the hopes for a national revival that had accompanied the founding of the Chinese Republic. Successive wars intended to create unity led only to greater political fragmentation and further civil war. Resources needed to strengthen the nation's economic foundations were committed instead to supporting growing armies. National defense was neglected as military commanders devoted their forces to domestic struggles for political power. While the Republic formally continued to exist, the practice of military rule made a mockery of republican institutions. Military intervention in politics ended up exacerbating, not resolving, China's political problems. It was no accident that the term warlord (junfa ), with its pejorative sense, came into popular parlance in the period after the outbreak of the North-South War. In the political discourse of the day, the identification of military commanders as warlords was not merely descriptive but reflected opposition to their political dominance. The emergence of warlordism as a political fact thus gave rise to anti-warlordism as a political issue. At the same time, the existence of warlordism, and the struggle against it, redefined the subsequent course of Chinese political development.
The reshaping of Chinese politics can be clearly seen in the con-
sequences of the failure of provincial self-government movements to resolve the problem of warlordism. First, by the mid 1920s, the ideas of provincial autonomy and federalism were generally discredited as solutions to China's political problems. Indeed, it is only recently that federalism has gained new supporters within China's democracy movement. Arthur Waldron has shown a resonance between the antiwarlord goals of earlier self-government movements and the current revival of federalist sentiment. Increased local self-government is again offered as an alternative to a political system that, in the aftermath of the Tiananmen massacre, is portrayed as held in place by military force. The context for this revived interest in federalism is, of course, considerably different. Whereas early Republican federalism sought to stabilize a condition of political disintegration, current proposals are reacting against the overwhelming power of a strong, centralized state. Ironically, the establishment of this centralized state had its roots in the disenchantment with provincialism and federalism in the 1920s. The reorganization of the Guomindang, or Nationalist Party, by Sun Yat-sen and the founding of the Communist Party both took place in the context of this disenchantment. Their subsequent programs reflected a shared commitment to centralized, unitary solutions to China's political problems.
Provincial autonomy was not the only political casualty of the triumph of warlordism. Andrew Nathan has shown how factional struggles for the control of the Beijing government, culminating in Cao Kun's 1923 use of bribery to obtain the presidency, undermined faith in the efficacy of constitutionalism to provide political stability. The inability of provincial constitutions produced by self-government movements to restrain the actions of provincial warlords contributed, on a local level, to the same result. The militarization of politics under the warlords may have similarly tainted perceptions of pluralistic politics. Under the influence of modernization theory, Lucian Pye has suggested that the open competition for power represented in warlord struggles was a tendency toward a more "modern" (i.e., pluralistic) form of politics. Although this view of the progressive features of warlordism is certainly too sanguine, Pye is correct in observing that, in the end, "repugnance toward the styles and objectives of the warlords became repugnance toward a competitive power pattern."
Warlordism thus had a broad transformational impact on Chinese politics. Insofar as warlordism penetrated all political activity, it problematized all the assumptions that informed politics in the early
Republic. By the same token, though, the conditions of warlordism stimulated a search for new solutions to China's political problems. Indeed, the linkage between warlordism and political fragmentation made the elimination of warlordism the sine qua non of the search for Chinese national strength and unity.
Even as anti-warlordism emerged as an overriding political issue, the existence of warlordism placed military force irrevocably at the center of Chinese politics. The militarization of politics in essence made military force a political necessity. Thus, the civilian proponents of provincial self-government movements were ultimately forced to find military allies. Mao Zedong drew the same conclusion in a more general context when he noted: "War can only be abolished through war—in order to get rid of the gun, we must first grasp it in our hand." Mao's observation also recognized another lesson that could be drawn from the inability of civilian provincialism to bind its military allies to its own ends. The political effectiveness of military power depended whose hand actually grasped it. Thus Mao's demand that every party member understand the "truth" of the statement "Political power grows out of the barrel of a gun" was immediately followed by assertion of another, equally important "principle": "The Party commands the gun, and the gun will never be allowed to command the Party." The key to political power was not just military force per se, but the control of military force. This concern was no less pressing for those seeking to lay claim to state power in the Republican period than it had been for the emperors of preceding dynasties.
The solution to the problem of military control, as ultimately implemented by both the Nationalist and Communist parties, was both political and organizational. The fundamental political solution, of course, was the party itself. In the 1920s, the Nationalist and Communist parties established strong ideological programs and organized themselves around new forms of political mobilization that went beyond the elite politics of the late Qing and early Republican eras. These measures provided them with expanded political constituencies and a new legitimacy that promised at last to resolve the early Republic's crisis of political authority. The existence of these parties therefore provided a political context for the incorporation of the military into a new integrative framework.
The organizational solution to the problem of military control was the formation of "party armies" based on the model of the Soviet Army. Ideological indoctrination of soldiers and officers in these armies provided a focus for loyalty beyond the personal chain of com-
mand that was the basis of warlord organization. The potential for warlord-like political autonomy was also weakened by the establishment of a system of political commissars. Recent research has suggested that rather than ensuring strict "civilian" control of the military, the commissar system in both Nationalist and Communist party armies functioned more to incorporate the military's political participation into the party's power structure. Paradoxically, then, the ultimate effect was not to depoliticize the army, as had been attempted in the early Republican provincial regimes, but to harness the politicization of the army in a manner that would strengthen party leadership. Under these circumstances, the party army not only provided a means to reverse the fragmentation of military power but also finally made possible the application of military power in the service of national unification.
In the end, the development of party armies sealed the fate of warlordism. No warlord could compete with the party army's capacity to accommodate increasing numbers of troops or its organizational cohesion. The final struggle to reconstitute the Chinese state thus turned on the conflict between the two parties that built these armies. It would be overly simplistic to reduce the final victory of the Communist Party over the Nationalist Party to military factors. At the same time, military force and political power were irreversibly linked, and military factors did play a crucial role in the outcome of this struggle. For example, the Communist Party's inattention to military power in its early years led to its devastating defeat after its 1927 break with the Nationalist Party. This disaster showed that mass movements alone were no match for a hostile modern army and led directly to the development of the People's Liberation Army. Likewise, the absorption of largely unreconstructed warlord armies into the National Revolutionary Army and relaxation of political controls over the army in general allowed the survival inside the Nationalist party army of what James Sheridan has described as "residual warlordism." This condition contributed to the eventual inability of the Nationalists' numerically superior army to inflict a decisive defeat on the Communists. Nor did the political importance of the military cease after the Communist victory in 1949. The People's Liberation Army's role in quelling the disorder of the Cultural Revolution and in suppressing demonstrations in 1989 shows the continuing linkage of military power to politics in Chinese society. This linkage thus remains an important legacy of warlordism to modern Chinese politics.
The emergence of warlordism was historically contingent on a
specific combination of social and political circumstances in the early twentieth century. The continued political importance of the military in China does not mean that warlordism will necessarily rise again. Nonetheless, the politicized nature of the Chinese army does create continuing potential for military interventions. Whether the military will again intervene as an autonomous political force depends on the opportunity presented by unfolding political events in China. The most pressing question is the ability of the Chinese Communist Party to weather the current crisis of Communism, which has already shaken or brought down most other Communist governments. It is not difficult to envision the collapse of the Communist Party creating a crisis of authority similar to that which plagued the early Republic, while at the same time cutting the People's Liberation Army adrift from its political foundations. In such an eventuality, the reappearance of something similar to warlordism would always be a possibility. Nonetheless, the uneven development of warlordism in the early twentieth century, with its suggestion of lost political alternatives and paths not taken, serves as a caution against uncritical belief in the predictive powers of history. History may mirror possibilities, but the contingencies of the moment might turn the political influence of the Chinese army in unexpected directions.