The revolt in the east began inconspicuously in 935 as an armed squabble among members of a Taira family who had settled in the Kanto area perhaps forty-five years earlier. They had grown intogreat landowners there, exercising control over a broad area that in cluded the provinces of Kazusa, Hitachi, and Shimosa (the northern part of Chiba Prefecture and the southern half of Ibaraki). As the conflict evolved in fits and starts, one of the Taira leaders, Masakado,emerged from his Shimosa base as the chief military power and prin cipal arbiter of disputes throughout the southern Kanto area.
The disputes in which he involved himself centered on the resistance of landowners to provincial exactions, Masakado taking the part of the aggrieved landholders. The fighting increasingly assumed the natureof rebellion against provincial authorities, who were mostly Masa kado’s kinsmen. But perhaps because the eastern provinces had along history of banditry, unrest, and minor revolts, the central government at first paid scant heed to Masakado and his fighting, intervening only when a suit was brought against him by one of his vic tims, and then simply to assess a mild punishment that was almost immediately canceled in a general amnesty.
The military forces of Masakado and his opponents comprised the leader’s close followers (Jusha), a permanent bodyguard or army of mounted warriors (400 in the case of Masakado), banrui (allies), and provincial troops recruited from among the cultivators. Banruiwere minor warriors, usually mounted, who were independent col laborators and quick to retire from the battlefield when the fighting went against them. Both Masakado and his enemies used the tactic of burning down the houses of their opponents’ banrui and peasantfoot soldiers and destroying their stored grain to weaken their re solve, from which we can understand that they were men of meager resources.
Although the organization of the armies of Masakado and his enemies were generally the same, they differed in one decisively impor tant respect: the anti-Masakado forces were an alliance of warrior leaders, at times as many as five or sixTaira and also a Minamoto anda Fujiwara, each with his own group of close followers. These al liances were known as to (associations), largely egalitarian leagues of warriors that joined together to further their common interests. (The term to is used in modern times to refer to political parties.)
Shortly before Masakado’s time, for instance, pack-horse haulers had formed»hire-horse associations» (shuba no to) and were terrorizing and loot ing agricultural villages along the roads of the Tokaido and Tosando, their depredations becoming so intolerable that the court establishedbarrier stations (sekisho) at key eastern passes to exercise some con trol over their movements. Nonhierarchical in membership at first, such leagues in the twelfth century developed into small, loosely knit warrior bands like the famous Seven Musashi Associations (Musashi shichiio) in the Kanto or the Matsura Association of Kyushu.
Professional warriors of Masakado’s time were archers on horse back. They did not by choice engage in hand-to-hand combat withswords. Nor did the mounted men carry lances, as they did in Eu rope. Masakado’s forces also included peasant foot soldiers armed with spears and shields, employed in mass tactics, in the tradition of the earlier imperial conscript armies.
The fighting between Masakado and his opponents, as it evolved through its several stages, was destructive and murderous, but the participants continued throughout the first years to appeal to the court at Kyoto for justice. Imperial authority, although conspicuous by the absence of its physical presence in the east, was still recognizedeven by Masakado himself. Then, at the end of 939, Masakado be came involved in a dispute over taxes between a member of the local gentry in Hitachi and the governor of the province. Before the affair ended, he had burned the provincial headquarters, seized the official signet of the province, and made off with the key to the provincial storehouse. Even the inward-looking court at Heian could scarcely ignore so direct a challenge to its authority or income.
At the urging of a certain Prince Okiyo, a former provincial offi cial in Musashi who had thrown in his lot with Masakado in the course of another dispute with local officials in 939, Masakado set out then to seize control of all the eight Kanto provinces, acting on Okiyo’s famous observation that the punishment for rebellion in many provinces was no worse than for that in one.
Claiming that he was obeying an oracle from the Great Bodhisattva Hachiman andjustifying his action on the ground that he was a descendant of Em peror Karamu, Masakado proclaimed himself the «New Emperor» (shinno), in contradistinction to the old tenno at Heian. Masakado
appointed new officials to the provinces and he also began to fash ion a rustic version of the central statutory government centered in Shimosa.
Masakado sent a message to the capital addressed to Fujiwara noTadahira, under whom in his youth he had enrolled himself as a fol lower. He sought the regent’s understanding of his actions, and also suggested that his ambitions did not extend beyond the Kanto, thus proposing in effect a division of the country between the regentalFujiwara and his own warrior family. By that time, however, the authorities at court were thoroughly alarmed, convinced that Masa kado’s forces would soon be descending on Kyoto.
They adopted a three-pronged policy aimed at ending the rebellion: (1) prayer; (2) the appointment of Pursuit and Apprehension Agents (tsuibushi) in the eastern provinces and, in the spring of 940, three or fourmonths after Masakado’s attack on the Hitachi provincial headquar ters, the dispatch of a court commander, Fujiwara no Tadafumi, tothe east; and (3) promises of reward to provincial leaders who suc ceeded in subduing Masakado. Significantly, it was the third prong of the policy – reliance on provincial warrior leaders – that actually brought Masakado down, although the court subsequently insisted on attributing its success to the divine efficacy of the first.
While the specially deputed court commander was still en route to the east at the beginning of 940, Masakado was surprised at his base in Shimosa with a depleted force and was killed by the joint forces of his cousin, Taira no Sadamori, and the Shimotsuke Suppression and Control Agent (oryoshi) Fujiwara no Hidesato.
The failure of the revolt seems to have been chiefly a result of organizational weakness. Masakado’s lack of a retainership system meant that his coalition force of antigovernment landholders tended to fragment after initial successes had achieved the aims of the landholders, and the dispersal of his large force of cultivators to their agricultural pursuits left him exposed at a critical moment to the overwhelming alliance of hostile forces. (The Shomonki says that Masakado’s force, which had once numbered as many as six thousand, had been reduced to a thousand men.)
The imperial court’s military response to the revolt may not have been as ineffective as it seems. The punitive expedition dispatchedfrom the capital against Masakado arrived at the scene after his de feat, but its expected advent in the fighting presumably entered into the calculations of the contending sides. Hidesato participated in thefighting as an imperially appointed officer, holding one of the key ti tles recently created by the court as it reconstituted its military.
Following the abandonment of the conscript-army system in the late eighth century and afterward, the statutory regime continued as before to rely for military strengdi on the richer and more powerful elements of rural society, asserting authority and control through aloose, evolving structure of ad hoc and permanent titles that depu tized individual warriors to exercise force in defense of the regime or for the maintenance of public order. The social and provincial origins of the regime’s military strength remained much the same as earlier. But the nature of the military’s relationship to the court shifted from the employed service of the Nara period, when men were recruited atthe government’s pleasure and recompensed for their labor, to deputized agency in the ninth and tenth centuries, when the power of pri vate warriors was confirmed and legitimatized by conferral of courttitles that may have been useful in the building of retainership sys tems.
Such deputized warriors might fight the battles of the court and enforce its laws, but first and foremost they fought in their own interests and took the offensive when they had most to gain.
The titles awarded to warriors and the functions of the title hold ers as the functions evolved during the period gave the government’smilitary arm more than simply a paper existence. The resulting mil itary and police system enabled the regime to survive rebellion and rampant banditry for close to four centuries, and the new system was in that sense a successful adaptation of statutory institutions to a changing society. But lacking significant numbers of men and weapons, and having no major military leaders of its own, the court at Heian was able to do little more than survive. It did not have the strength, certainly, to impose its unilateral will on the society.
The dispatch by the court of a force under the command of a se nior noble to put down Masaskado’s revolt continued the militarytraditions of the statutory regime, but the ineffectiveness of that ef fort signaled the approaching end of the traditions. The commander of the army subsequently led a more successful imperial army against the pirate-rebel Sumitomo at the Dazaifu, but after that there were no more such armies.
Instead, there came to be four chief, continuously filled titles given to warriors entrusted with military and police duties outsidethe capital. One, «General of the Pacification and Defense Headquarters» (chinjufu shogun), was a title of the Nara period that ac quired a markedly different meaning as it came to be used duringthe Heian period. Another, «Offenses Investigation Agent» (kebi ishi), was originally created for police purposes in the capital city in the ninth century, but in the tenth it expanded nationwide, suchagents then being established in each province by the local govern ment. Two new titles formed the backbone of imperial military framework: «Suppression and Control Agent» (prydshi) and «Pursuit and Apprehension Agent» (tsuibushi).
The Pacification and Defense Headquarters had been established in Mutsu Province in the first half of the eighth century to provide security for Japanese colonizers against Emishi in the region. Itslarge army of conscript soldiers from other provinces played a cru cial role in the subjugation of the Emishi, continuing operations against rebellious and bandit elements all through the ninth century.
It was staffed by four grades of military officers, headed by a general {shogun). Beginning in the tenth century, the office of General of the Headquarters tended to become hereditary, remaining in a single clan or family line for as many as five successive generations. Thetitle fluctuated between lineages but remained always in three lead ing warrior families directly descended from high nobility at the capital: sublineages of the Uona line of the Fujiwara, the Kammu Heishi, and the Seiwa Genji.
When the Emishi submitted to the Japanese presence in the northeast during the ninth century, the Pacification and Defense Headquarters {chinjufu) lost its military raison d’etre, becoming thereafter mainly a civil administrative organ, little or no different perhaps from a normal provincial-government office. The holder of the title of General typically held concurrent appointment to thegovernorship of Mutsu, which in the eleventh century was associated with great wealth. Despite the loss of the office’s primary mili tary function, there was no other military title that brought so muchprestige to an eastern warrior, but the title seems to have been primarily a recognition of existing power and a reward for military ex ploits, not a grant of power or the commission of military duties.
In response to the upsurge of banditry in the countryside, the po lice and judicial powers of the provincial governments came to be separated and vested in special officers called, like their counterparts in the capital. Offenses Investigation Agents.
Such officers existed at latest by the middle of ninth century, and by the tenth they are found in nearly one-third of all provinces, from Musashi in the east to the Dazaifu in Kyushu. Multiple kebiishi appointments might be made for a single province, and two, presumably especially lawlessprovinces, are known to have installed kebiishi also at the subprovincial, or district, level. Initially selected from rich and powerful families in a province, by the tenth century rankless peasants were hold ing the title, and their appointments, which had originally been made by the central government, were in the hands of the provincial headquarters. The appointees then tended to be functionaries of the headquarters.
Fujiwara no Hidesato held the title of Suppression and ControlAgent when he and Taira no Sadamori defeated and killed Masa kado in 940, and Minamoto noTsunemoto (d. 961), the ancestor of the Seiwa Genji who was a member of the court’s army dispatched against Masakado, held the title of Pursuit and Apprehension Agent. The two titles had emerged in the ninth and tenth centuries as ad hoc court appointments of private warriors to mobilize andlead their own and sometimes other local military forces in the sup pression of revolts and banditry.
Originally different in function, the two titles came to be substantively the same in the tenth century, when they lost their ad hoc character and were continuously filled, each appointee having jurisdiction in a single province (Hidesato was, for instance, the oryoshi of Shimotsuke) and in principle actingunder the direction of the provincial-government headquarters. Ap pointed by order of the central government, the agents were usually chosen from among powerful local families, frequently from those whose members were already part of the provincial headquarters staff. The posts often became hereditary, and as officials actually resident in a province, appointees were often able in the late Heian period to take over the military and administrative authority of the provincial government.
During the half year before Masakado set the fateful fire at the Hi tachi provincial headquarters, the court had been rocked by news of a major revolt by resettled Emishi in Dewa and a renewed outbreak of piracy in the west, in the Inland Sea (Seto Naikai) area, where piracy had been a continuing problem since the last half of the ninth century. The piracy was soon contained, but then just two weeks after word of the burning of the Hitachi headquarters reached Kyoto, the court learned from the vice-governor of Bizen that a powerful pirateleader named Fujiwara no Sumitomo, a great-grandson of Moto tsune’s biological father, Nagara, and therefore a distant cousin of the current regent, was about to take to the sea again.
A few days latercame the terrifying news that Sumitomo had caught up with the vicegovernor in Settsu as the latter was returning from Bizen to the capital, cut off his ears and slit his nose, killed his son, and taken pris oner his wife. Buffeted by alarming reports from east and west, thecourt seems to have been on the verge of panic, fearing that Sumi tomo and Masakado were acting in concert and that their joint forces would soon be assaulting the virtually defenseless capital itself.
However, Sumitomo was not a tsuwamono of the Masakado type. His rebellion followed its own course in its origins and resolution. Sumitomo had formerly held a minor provincial post in Iyo (present Ehime Prefecture) on the island of Shikoku, where he built a largepiratical following that was active from as early as 936. Sumitomo or ganized bands of pirates operating in the area, most of whom were likely poor fishermen and petty seamen, and he began to plunder the sea lanes and ports of the Inland Sea, the transportation routes forthe movement of government revenues and products from the west ern provinces to the capital.
After the renewal of his activities in 939, he attacked governmen tal headquarters in the provinces of Shikoku and along the Honshu coast, burned ships, and spread havoc throughout the coastal areas. Sumitomo’s extensive control of the sea grew into a direct challenge to governmental authority, but it was not until after the defeat of Masakado in early 940 that the central government was able to turnits full attention to the pirate chief. The authorities in Kyoto had ear lier attempted to check his depredations by appeasement (he wasgranted Fifth Rank), but now a government force made up of pro vincial leaders finally succeeded in attacking and, with the aid of a turncoat, reducing Sumitomo’s stronghold in Iyo. The pirate leaderfled then to Kyushu, where he occupied and sacked the lightly de fended Dazaifu headquarters, continuing his predatory ways until at Hakata he suffered another disastrous defeat in 941 at the hands of court-dispatched land and sea forces supported by deserters from his own ranks and he met his death.
The revolt of Masakado and the piracy of Sumitomo proved to be not as serious threats as the court imagined; but some key elementsin the origins, prosecution, and suppression of the revolts were em blematic harbingers of the great historical changes that came duringthe following two and a half centuries. The suppression of Masa kado’s revolt did not result in an assertion of court authority in the east. Rather, the revolt and its suppression confirmed the private wealth and military strength of the local leaders. Masakado and his principal opponents, Taira no Sadamori, Fujiwara no Hidesato, and Minamoto no Tsunemoto (the vice-governor of Musashi who brought the initial charge of rebellion against Masakado at court), were the ancestors of most of the important warrior leaders in the east during the remainder of the Heian period, and it was from the east that the decisive military power flowed in the fateful twelfth century.
All four men were alike in being descendants of the highest strata of the court nobility (the imperial line and the Fujiwara clan). Their ancestors had received appointments in the ninth and tenth centuries to provincial or military posts in the east and had settled and prospered there after the terms of their offices had expired. In the typical case of the founder of a major warrior family in the east, an ancestor of noble extraction first came to the area with thecourt-conferred title of General of the chinjufu in Mutsu, and his de scendants often continued to hold that title after him in a limited hereditary fashion. The powers of the office of the chinjufu General in the tenth century probably depended mostly on an appointee’s own private military resources, but it was a title of great prestige, conferring on its holder what amounted to the titular leadership ofeastern warriors, and its official civil-military functions were still important. Although the Emishi had been decisively defeated by Japan ese forces in the ninth century, large numbers in the northern andcentral parts of Dewa and Mutsu continued to live in a semi independent tribal society beyond the purview of regular provincialand district officials. It was the responsibility of the chinjufu to man age and control them.
Bearing the old and distinguished title of chinjufu General, or the somewhat less grand title of governor or vice-governor of a province,and tied to the court not only by blood but also by occasional em ployment at the capital in court or noble-family military functions, these men and their descendants quickly became major loci of armed power in the eastern provinces, providing military protectionfor local gentry against their enemies and contending among them selves for domination in an area. Sometimes called «military nobles»(gunji kizoku) by historians, they were also, in a sense, trouble waiting to happen, ready to join in any fray that promised to be of ad vantage to their own interests, as Masakado did in the spring of 939 when he rushed to the relief of a district official in Musashi who had come under attack by a provincial governor bent on the collection of revenues and other wealth. Masakado himself was a professional warrior, a tsuwamono in the contemporary lexicon, a predecessor of the late-Heian bushi. «He gathered many fierce tsuwamono as his companions and engaged in battles as his occupation,» in the wordsof a story about him in the twelfth-century Konjaku monogatari shu.
He differed from the earlier warrior of the statutory state who was in fundamental ways an employee of the state who fought for its purposes. The professional warrior of the tenth century and afterwas a private fighter who did battle as his own interests dictated, le gitimizing his actions when he could as government business.
The failure of the two tenth-century revolts is attributable chiefly in both cases to the fragmentation of the rebel forces, alliances chiefly, in Masakado’s case, of a warrior leader and rural gentry, and, in Sumitomo’s, of a renegade official and local piratical forces. After Masakado’s successes against the provincial government forces ofHitachi and other Kanto provinces, the local gentry, their chief purpose achieved, dispersed, leaving Masakado with fewer than a thou sand men in his final, losing battle with Sadamori and Hidesato.
Thepattern was similar in the case of Sumitomo, who lost two of his pirate chieftains and twenty-five hundred men to the bribes of the governor of Iyo in 936 and who was later routed when his subcomman der surrendered to government forces and led them to the rebel’s base in Iyo Province. There were no adequately cohesive socialbonds to hold the rebel forces together once their individual inter ests were served. Such bonds emerged only with the strong vassalage system that in the late eleventh and twelfth centuries transformed rural society.
Although the government’s forces were perhaps inherently no stronger than the rebels’, the court was able, nevertheless, to field punitive forces sufficiently powerful to suppress the revolts, thanks to its own surviving military and financial resources and to its successin allying itself with private warriors very like the rebels themselves provincial «military nobility» with strong familial and formal ties to the court. For such figures, their services to the government in the suppression of rebels must have seemed simply the continuation in a different framework of the private battles that marked the expansion of warrior strength in the provinces.
The revolts of Masakado and Sumitomo, although not revolu tionary in effect, were true watersheds – in the opinion of some, the beginning of the «middle age» (chusei). At least they were the firstmajor efforts by provincial military leaders to assert their autono mous strength. As the old ties of the military nobility with Kyoto weakened and the strength and unity of the provincial military grew, the court’s position became ever more isolated and precarious,maintained perhaps as much by ideological, religious, and senti mental values as by any important material strength.
The Masakado rebellion marked the advent of the private profes sional warrior in Japanese political history, and with his appearance seems also to have emerged the first lineaments of a recognizablewarrior ideal. The loose code of conduct that in the Meiji era became generally known as bushido («way of the warrior») is not di rectly traceable to the tsuwamono, but several prized attributes of the tsuwamono may be thought of as having constituted an initial essay at a warrior ethic. Knowledge of the early ideal is slender, since the men left little personal record of themselves, and other sources areproblematic, existing as unique, single texts of uncertain authentic ity, as in the case of Shomonki, or surviving as tales probably oral in origin and collected over a century after Masakado’s time (Konjaku monogatari-shu). The picture found in such sources is sketchy, with none of the detail of knightly perfection displayed for our admirationin later sources from the thirteenth century and after, but some im portant elements are already in place.
Defining characteristics of the tsuwamono’s conception of himselfare early seen in a letter written at the beginning of 940 by the pro totypical tsuwamono Taira no Masakado. Written just at the time of his assumption of a royal title at the head of his newly established state in the east, and addressed in explanation of his actions to the imperial regent at Heian, the letter as recorded in Shomonki justifiesthe new state on three grounds: (1) the precedents of the past, history having recorded earlier instances of usurpation of central au thority by arms; (2) Masakado’s descent from an emperor, a lineage that could entitle him to exercise sovereignty over the country; and (3) his military power, with which he had been endowed by Heaven, not by any governmental authority.
Masakado’s argument followeda rationale suited to an age when the government had largely abdi cated its defense and peacekeeping responsibilities, and the security of the country was by default mostly in the hands of tsuwamono, or so they claimed.
The bushi of later ages was depicted as placing supreme value on loyalty to his lord, but the tsuwamono as represented in Masakado’s letter found his central values in historical practice, familial descent,and military might. Those values underlay, or were expressed in, sev eral characteristics of the ideal tsuwamono. Familial descent perhapsmore than anything else determined the warrior’s status, for the pro fession of arms, like all other professions at the time, was hereditaryand closed at least in its upper reaches to those outside the tsuwamono families, the «houses of martial valor» (buyu no ie) or «bow and-arrow houses» (yumiya no ie), as they were sometimes called.
There was in consequence a strong emphasis on family honor and a tendency to attribute one’s accomplishments to the family line. On being praised and rewarded by his master the crown prince for aseemingly impossible arrow shot that killed a fox on the roof of Fu jiwara no Kaneie’s house one night, Minamoto no Yorimitsu (948-1021), a court warrior and founder of a major sublineage of the Seiwa Genji, is said to have responded, «It was not I alone who shot that arrow. I was aided by my family’s guardian deity in order not to bring shame on my ancestors.»
Prayer to the guardian deity of a family and to the family ancestors was essential. That centrality of the family line led to the warrior’s ritual proclamation before battle of the names of his more notable ancestors, a practice that became customary by the twelfth century, at least in literary accounts. That recitation of ancestral names reflected an ancient belief in «wordspirit» (kotodama), according to which words were thought to be en dowed with spirit and the intoning of a name gave the speaker the protection and aid of the spirit of the name’s referent.
Although military might may have been in practice the most sig nificant trait of the tsuwamono, victory in battle was not necessarily the highest goal. What seems to have been prized most in a warriorwas his readiness to lay down his life in battle, to lead his men himself on the battlefield, and his ferocious tenacity in the face of overwhelming odds. His utter disregard for his own life in pursuit of mil itary honor could require him to be indifferent even to the welfare of his own children.
A well-known Konjaku tale tells the story of aneleventh-century hero named Fujiwara no Chikataka, who was reprimanded for his weakness when he reported to the renowned war rior and governor of Chikataka’s province, Minamoto no Yorinobu, that his son, a child of only five or six, had been taken hostage and was being held under threat of death by a desperate gang of bandits trapped at his house. Yorinobu laughed and said: I understand your tears, but they do no good. A warrior must be resolved to grapple with anyone, be he devil or god. Blubbering away like that is stu- pid. Let them kill him – he’s just one child. A tsuwamono has to be prepared for that. He’s not worth much if he lets concern for wife and children give him pause. Fearlessness depends on thinking neither of oneself nor of one’s wife or children.
On the other hand, other Konjaku and Skomonki stories exemplify that the true tsuwamono spared and protected women.