Gempei Wars that were confined to the island of Honshu destroyed farmers’ lands and families. When the Taira house was finally defeated, the Minamoto established the bakufu, a military government, in which real power was in the hands of the Minamoto and their samurais, leaving the status of the emperor and his court only to be nominal. From this time, the feudal age in Japan began.

The establishment of Japan’s first warrior government, the Kamakura bakufu, represented both a culmination and a beginning. Since the tenth century, an increasingly professionalized class of mounted fighting men had served in local areas as estate administrators and policemen and as officials attached to the organs of provincial governance. By the twelfth century, warriors had come to exercise a dominant share of the total volume of local government, but even after two hundred years they remained politically immature. The most exalted warriors were still only middle-level figures in hierarchies dominated by courtiers and religious institutions in and near the capital. The bakufu’s founding in the 1180s thus represented an initial breakthrough to power on the part of elite fighting men, but the fledgling regime was scarcely in a position to assume unitary control over the entire country. What evolved was a system of government approximating a dyarchy. During the Kamakura period, Japan had two capitals and two interconnected loci of authority. The potential of warrior power was clear enough to those who cared to envision it, but the legacy of the past prevented more than a slow progress into the future.

Despite its aversion to fighting, the bakufu was created by war, the Gempei (Genji versus Heishi, or Minamoto versus Taira) conflict of 1180-5. This was a much more complex upheaveal than its name implies. Far from being a dispute between two great warrior clans, as it is so often depicted, the Gempei conflict was a national civil war involving substantial intraclan fighting and also pitting local against central interests.1 Indeed, the character of the violence was responsible for the type of regime that was created. Likewise, the backdrop to the conflict was a product of society’s tensions and is therefore integral to the history of the Kamakura bakufu.

To understand the limitations of both the warrior victory and the resulting government, we need to trace the rise of the warrior class in the Heian period as well as the ascendancy of the Taira in the years just before the Gempei War. The original blueprint for imperial government in Japan did not envision a military aristocracy as the mainstay of administration over the countryside. Yet as the courtiers in the capital became more confident of their superiority, they began to loosen their grip over the provinces, exchanging governance over a public realm for proprietorship over its component pieces. The country was divided into public and private estates (the provincial lands known as kokugaryo, and the estates known as shorn), under the authority of governors and estate holders, respectively, who themselves made up the courtier and religious elite. The owners of land at the topmost proprietary level were thus exclusively nobles and clerics. The purpose of this privatization of land was to secure a flow of revenue that exceeded what was provided by the holding of bureaucratic office. In turn, this permitted an increasingly extravagant life-style in the capital. The division of the country was predicated in this way on the desire of shoen owners to be absentee landlords. Yet it was equally dependent on those owners’ ability to draft into service a class of willing and obedient administrators.

This loosening of control from above also loosened the cement that bound the provinces to the capital. A degree of local instability ensued, which caused the lower ranks to look to one another for mutual support and protection. Leadership fell to persons of distinction whose principal source of prestige was an ancestry traceable to the capital. Thus, unlike the invaders who promoted the feudalization of Europe, local leaders in Japan were men with long pedigrees. They also retained their central connections, which meant that the developing class of provincial administrators were less members of local war bands than members of groups that were forming to secure the peace. This did not preclude outbreaks of lawlessness. But courtiers could always brand such outbursts as rebellion and enroll others as their provincial agents. In this way, at any rate, local and central remained essentially joined for the duration of the Heian period.

The warriors who were becoming the true captains of local society were called zaichokanjin, or resident officials attached to provincial government headquarters (kokuga). Although the governorships themselves continued to rotate among courtiers in Kyoto, positions within the kokuga became hereditary. Later, during the early stages of the Gempei War, the developing cleavage of interests here was exploited by the founder of the Kamakura bakufu, Minamoto Yoritomo. However, during the two centuries preceding 1180, patrons in the capital were able to channel the energies of provincial subordinates towards mutually beneficial ends. On the one hand, the locals were given extensive powers in the areas of tax collecting and policing. But on the other hand, these same locals were obliged to work through their superiors to secure new appointments or confirmations of old ones or to secure justice in the frequent legal battles between kin and nonkin rivals. Neither the local chieftain nor the clan head (if this was a different person) was empowered to provide these services on his own authority; he too was dependent on the support of a central patron. The result was that ownership and administration, authority and power, became separable, with little risk to the capital-resident proprietor. So ingrained was the psychology of a hierarchy in which the center dominated the periphery that in the absence of some regionally based patronage source such as the bakufu, courtiers in the capital, no matter how effete, could remain the superiors of warriors, no matter how powerful the latter were.

But Kyoto protected its interests in other ways, too. One of the most ingenious was to promote a handful of men as career governors. These persons might then be moved from province to province, much as modern ambassadors are moved today. The origins of this practice have not been adequately studied, but by late in the eleventh century the use of such representatives, now called zuryo, had become interwoven with the competition between the Fujiwara and retired emperor patronage blocs in the capital. By this time, governorships had become, in a sense, commodities circulating among the elite. The proprietary province (chigyokoku) system, as it was called, was designed to allow patronage groups to function on both sides of the local land ledger (shoen and kokugaryo), with the governor as the principal instrument of manipulation. What is important to us is the identity and character of the journeyman governors who now came to be employed by the ex-emperors and Fujiwara. They were from the Taira and Minamoto, particular scions of which were recognized as career troubleshooters for provinces possessed by their patrons. Thus, to cite one example, Taira Masamori received successive appointments to at least nine provinces, as did his son Tadamori after him. And the latter’s son, the illustrious Kiyomori, was governor of three provinces before beginning his historic ascent in the capital.

The leaders of the Taira and Minamoto need to be appreciated in this light. They were not, as they are usually depicted, regional chieftains chafing under courtier dominance. Rather, they were bridging figures – military nobles in the truest sense – between the great central aristocrats, who were their patrons, and the great provincial warriors, who were their followers. The leaders’ dual character, born out of service to two constituencies, is essential to an understanding of the slow progress of warrior development in its initial phase. It is also basic to the incompleteness of the warrior revolution that was later spearheaded by the bakufu.