The term for the Japanese warrior, bushi («martial servitor»), which became common from the late Heian period on, was preceded in earlier centuries by the words tsuwamono and musha.
Bushi, like the other two words, specified the professional warrior as distinguished from peasant conscripts, court military officials, and palace guards. Tsuwamono is a native Japanese noun of unknown morphology, and musha (or musa) a Sino-Japanese binom meaning «martial one.» The professional warriors were private fighters who at first foughtentirely in their own interests and for their own ends. They later entered the personal retainerships of court nobles or, more commonly, military lords, the greater of whom were called toryo, an other Sino-Japanese word, the constituents of ‘which denoted «rooftree and rafters,» that is, the chief members of a roof and, by extension, also «chief person.» The term was commonly used to mean the chieftain of a major clan or family. As private fighters, thetsuwamono were fundamentally different in character from govern ment troops, who were conscripted by authority of the state and fought solely for it.
During the seven hundred years from the twelfth century to the nineteenth, Japanese society was dominated by warriors, who first appeared as a potent independent political force in the country at the time of the revolt of Taira no Masakado in the mid-tenth century.
Masakado and many of his supporters and opponents were professional fighting men, mounted archers descended in their skills per haps from the prehistoric hunters found in the archaeological record of the eastern provinces. Mounted warriors are known in Japan from before the seventh century, but they first became professional and a locus of autonomous political power during the tenth century, when their arms were employed not on behalf of the government for its purposes but in pursuit of private interests.
As the organizational skills of such warriors grew, they formed retainer-based militarybands of increasing size and power, structures that eventually en abled them to contend for control of the country. Their triumph overa weakened imperial court in the twelfth century led to the establishment of a new, warrior-ruled system of government in the Ka makura period (1185-1333).
The popular modern image of the Japanese warrior has beenshaped largely by colorful literary and pictorial representations of fif teenth- and sixteenth-century battles in which swordplay figured prominently, and also by the knightly figure of the samurai as found in the status system of the Tokugawa bakufu. The Heian warrior was neither. He was primarily a mounted archer who wielded the daggerand sword when his supply of arrows was gone and fighting had be come hand-to-hand. At this time the bow rather than the sword was the weapon of choice. The warrior’s status was not, like that of the Edo-period samurai, clearly established near the top of a centralized military society, nor did he see himself, his superiors, or his society in the ideal categories employed by later Confucian pedagogues.
De rived from a verb meaning to attend on a superior, samurai was usedin that sense (in its earlier pronunciation of saburai) in an early eighth-century chronicle to designate civilian attendants assigned tothe old or infirm or in the menial service of the sovereign, and it con tinued to be used in that nonmilitary sense in the Heian period for low-ranking servants of the nobility. As a title within court society, samurai was also applied to military attendants.
The main forces of the statutory (ritsuryo) imperial court in theeighth century were peasant conscripts, mostly a foot soldiery orga nized into units of three hundred or fewer to a thousand or moremen. Military service was a heavy economic burden for the con scripts, who were required to furnish their own weapons, clothing, and other equipment, specified by law to include a bow, bowstrings,arrows, a quiver, two swords, leggings, boots, a reed hat, a whet stone, and containers for water, food, and salt. Each ten-man unit was also required to supply a tent, tools, cooking equipment, and moxa.
The conscription of a single man, it was said, could bring about the collapse of his entire family.
Modeled on the armies of T’ang China, the statutory conscript system provided the imperial court with a well-articulated military force, lightly armed and trained. But the operation of the system was plagued with trouble from the beginning; individual units sometimesdegenerated into little more than work gangs in the employ of offi cials. In the absence of major internal enemies after the middle of the eighth century and the fading of the perceived threat of Chinese and Korean aggression, the system was generally abandoned in 780 and 792, retained then only in what were considered strategic «frontier» areas, namely, the northeast (the provinces of Dewa and Mutsu) and the Dazaifu provinces of Kyushu.
It was partially replaced by small groups of «stalwart youths» (kondei), elite mounted fighting men who first appeared in the eighth-century east, where their martial skillswere likely learned in some part from their redoubtable Emishi enemies, especially in the matter of fighting from horseback. (For an ac count of the expeditions against the tribal Emishi in the northeast around 800, see Chapter 1.)
Little is known about kondei duties and functions. In 733, three hundred conscripts from Mutsu and Dewa skilled in mountedarchery were named kondei. It is significant that kondei were first ap pointed in the military districts responsible for the subjugation of theEmishi, and that mounted archers were recognized as being more ef fective than foot soldiers in combat with those tribal people. The kondei of 792 and after were members of powerful local families igunji, «district officials») and cultivators sufficiently prosperous to equip, mount, and support a warrior, recruited from what was likely a substantial pool of hunters in the area.
Provisions and two grooms (Jbatei) aged seventeen to twenty were supplied for each kondei, one of the grooms actually functioning perhaps as a batman.
Although the kondei were no doubt militarily superior to the peasant conscripts of the old system, their small numbers nationwide in dicate that they could not have fully replaced the conscripts. There were fewer than four thousand scattered in units of twenty to two hundred throughout Japan, and they could have functioned, one may assume, only as a constabulary in the protection of government facilities and operations.
The central government had in effect ceded responsibility for the preservation of the public peace outside the vicinity of the capital to the landed class.
The conscript system was not disbanded in Kyushu until 928, re placed there in part by units of «select youths» (senshi), who were similar to the kondei in origin and presumed functions but were somewhat better attended.
Although there were fewer horses andarchers in Kyushu than in the east, the creation of senshi units stim ulated a similar, if belated, development of a warrior class there, too.
The disestablishment of the conscript armies and their replacement by small units suitable only for the security of government of fices and facilities signaled the abandonment of state responsibility for the peace of the country. Reflecting that shrinkage of governmentintention, the sources for the period record no actions involving ei ther kondei or senshi. The name kondei office (kondeidokord) survived within provincial-government structures until the end of the Heian period, but the offices came to have merely titular existences.
The development of a warrior class in the tenth century and after cannot be directly linked to either the kondei or senshi, but the men of those units were clearly products of socioeconomic circumstances similar to those that led to the emergence of the private professional warrior. They were mounted fighters and shared common origins with the direct ancestor of the warrior class, the tsuwamono. Both the kondei and the tsuwamono came largely from a prosperous landed class, especially in the east, which was accustomed to hunting fromhorseback and had a long tradition of bearing arms. Where the kondei were insufficient in number to suppress banditry, provincial landlords became more active in military preparedness. They relied in creasingly on arms also to settle their disputes and to extend their holdings when they found the opportunity.
Under the statutory laws the provincial governor was conceived of as a civilian authority and his staff was not permitted to bear arms.
However, in the ninth century, because of lawlessness in the coun tryside, governors of certain provinces were granted permission to engage armed men to protect themselves and their headquarter compounds and to enforce their orders. By the beginning of the tenth century and perhaps earlier, when serious disorders occurred in a province, the court issued additional police or military titles tothe governor or his deputies to empower them to raise and lead war riors in order to apprehend miscreants and restore order.
These special commissions were intended to be temporary to meet emergen cies, but they often became long-term and sometimes hereditary. In the eleventh century, provincial headquarters continued to developas military centers, regularly garrisoned, and the governor was au thorized to engage the services as needed of professional warriors tsuwamono) of the locality.
The word tsuwamono in its earliest known uses referred broadly to the means of warfare – weapons, provisions, and skills – and also to the persons who made use of those to wage war – warriors. By the ninth century, the term was restricted to the human element of the definition, the fighter, and in the tenth century tsuwamono came to designate more specifically the professional fighter. It was in that sense thatTaira no Masakado’s «fame as a tsuwamono» was recorded in Shomonki.
Tsuwamono acquired retainers (rotd, «lads» or, more literally, «young companions»), but initially not in the relationship ofa lord to his vassals. The term roto referred rather to the fellow tsuwa mono whose interests were in common with those of their leader and who fought under him. Since they were not strictly subordinate to their leader, they are perhaps better described in their earliest form as companions. Rudimentary organizations of tsuwamono and row developed by the middle of the tenth century, thereafter growing in size and strength, but roto tended to maintain their relationship with their leader only as long as it served their own interests, a tendencythat remained to some extent even after the subsequent develop ment of personal bonds of loyalty and subordination had led to the formation of the large-scale warrior retainerships that dominated the political history of the twelfth century.
In acquiring retainers, tsuwamono may have been following the ex ample of provincial governors. At any rate, the first occurrence of the word roto is the case of a former governor of Tosa, Ki no Tsurayuki,who mentions roto among his retinue as he sets off on his return voyage from Tosa to the capital in 934. Roto are not found with tsuwa mono until a few years later, at the time of Masakado’s revolt.
The practice of acquiring retainers seems to have been widespread among governors. The value of retainers to the governor came to bemore than simply personal security; they were employed for admin istrative work and were also available for police or military functions. They appear to have been key instruments in the extraction of wealth from a province. In one case, it is reported, the dispersal of a governor’s retainers after his death resulted in the impoverishment of his family. But a governor was a civil official, and when his termof office expired, his retainers might remain to serve the next gover nor or seek employment with the governor of another province. The tsuwamono, on the other hand, was a professional, hereditary warrior vested with interests in land, and his row might continue in service with him and his family indefinitely.
The tsuwamono warrior was the product of sweeping changes in landholding that in the eighth century and after transformed the public, imperial lands and people of the statutory regime into astratified society of wealthy, powerful private landholders and a sub servient peasantry.
The new landholders were in some cases local in genesis, their lands and labor force growing from an original familyand its state allotment of land, and in other cases stemming from external authority, descending from officials and military officers ap pointed in a province by the central government. As the holdings of the new lords expanded and control of an increasingly independentpeasant workforce grew more difficult, and as commendation converted land into shoen estates with complex and often conflictive ex ternal relationships, military might became a useful and necessary adjunct of landholding. The nucleus of the military forces created on the great landholdings was the tsuwamono, who enforced the will ofthe landholder, defended the holding against external threats, offi cial and private, and kept the peace.
The earlier shoen were owned by higher nobles and by major temples and shrines that, as absentee landlords, sent a manager to over see cultivation and remit revenues. However, by the tenth century, most new shoen were established by families, local or newly arrived from the capital, who commended land to a high noble or religiousinstitution or member of the imperial family (ryoke or honjo) with in fluence enough to protect the shoen from interference by provincialauthorities and from provincial exactions. In return the owner manager paid his protector a percentage of the rents collected from farmers. Both the new shoen and the older type founded by an owner in the capital required local military force to keep order on the shoen,to collect rents and taxes from the cultivators, to stave off the gover nor’s men, to resist encroachment by nearby shden, and to expand into kokugaryo (lands of the provincial government) or other lands as opportunity arose. Tsuwamono were often the owners or managersof shoen. In the tenth and eleventh centuries they were also found in creasingly among the administrators in the offices of the provincial headquarters and managers of the province’s lands.
As military might became a necessary element in successful land holding, two noble warrior clans established themselves as tsuwamono both in the vicinity of the capital and in the provinces, especially and most fatefully in the east. Their success was partly attributable to theirfamily lineage, the multiple lines of both clans originating from im perial offspring removed from the imperial family in the cost-cutting of Emperor Saga’s time or after and given the noble clan names of Minamoto orTaira.
Many of those demoted princes remained in the capital pursuing con ventional court careers, they and their descendants sometimes rising to high office and rank, while others sought their fortunes in the provinces, turning there to the profession of arms. But the most prominent of these warriors maintained their connections at court and, from time to time, held guard or police offices in the capital. The Minamoto name was granted to one or more princes by eleven emperors from Saga (r. 809-23) to Sanjo (r. 1011-16).
Themultiple lines created in that manner are distinguished in nomen clature according to the name of the ancestral emperor: Saga Genji,Murakami Genji, and so on, «Genji» being the Sino-Japanese read ing of the characters with which «Minamoto clan» is written. The bakufu of the Kamakura period (1185-1333) was founded by the lineof the Minamoto usually known as the Seiwa Genji, so called be cause they were descended from a prince named Tsunemoto (d. 961), who was supposed to be a grandson of Emperor Seiwa. But adocument discovered in the nineteenth century indicates that Tsu nemoto was actually a grandson ofYozei (and thus a great-grandson of Seiwa). Not all students of the subject accept the testimony of the document, but if it is correct, we may suspect that the name of the criminally inclined and finally deposed Yozei was avoided by family historians in favor of a more auspicious genealogical origin for the great line.
Tsunemoto’s son, Mitsunaka (912-97), served the Northem (regental) House of the Fujiwara as its military arm in overaw ing its rivals. For three more generations until the end of the eleventh century the Seiwa Genji held court positions in the guards,appointments to lead punitive expeditions, and rotations as provincial governors, and gained a reputation as the most formidable war riors in the land.
«Minamoto» was a common noun meaning «source of water, origin.» In the clan name, it was used in the sense of «source of court officials.» An alternative interpretation assumes that the choice of the name by Saga, its first grantor, was influenced in the court of that sinophile emperor by Chinese precedent, giving it the sense «of the blood.» The precedent occurred in China in thereign of an emperor of the Northern Wei (386-534), who, in recog nition of the common Hsien-pei origins of a Mongolian prince with himself and the officers of his court, bestowed as a family name on the prince the character with which Minamoto is written – Yuan (in a modern Chinese pronunciation).
The princely recipients of the clan name of Taira were typically ofa remoter degree of imperial descent than their Minamoto coun terparts: they were never sons of emperors (as Minamoto usually were), but grandsons or genealogically even further removed from the throne. There were several lines of the clan, springing like the Minamoto from the progeny of different emperors (Kammu, Koko, Nimmyo, and Montoku), but it was the first created lines, the four Kammu Heishi lines (Heishi is the Sino-Japanese reading of «Taira clan,» descending from Emperor Kammu, that produced the Taira who figured most prominently in the history of the period. See Taira genealogy Figure 9.7). First bestowed in the ninth century, the Taira name, in its orthography and Sino-Japanese reading, echoed that of Kammu’s new capital at Heian, the characters for which were sometimes also read «Taira no Kyo» or «Taira no Miyako.»
Less closely related to emperors than most of the Minamoto and more thoroughly excluded from the higher offices of the court, some Taira early left the capital for the provinces, where their prospects were more encouraging. There were two chief branches of theKammu Heishi, one a courtier line of mostly minor and middlinglevel officials; the other, founded by Kammu’s great-grandson Taka mochi (fl. 889), became a warrior line that included many of the best-known military men of the final century of the Heian period, and it is almost always Takamochi’s line that is meant when the term «Kammu Heishi» is used. Takamochi was appointed vice-governor of Kazusa around the end of the ninth century, married locally in the province of his appointment, and expanded his landholdings inKazusa and Shimosa. His descendants included in the third genera tion Masakado, whose campaigns, from 935 to 940, may be said to have marked the emergence of professional-warrior power in Japan.