Hired Swords vs. Heavenly Warriors: The Development of Warrior Power in Early Japan


The common treatment of the Heian court found in textbooks and survey histories depicts Japan’s ruling class as a group of leisured and effete aristocrats more concerned with composing elaborate waka (poetry) and mastering esoteric Buddhist practices than the effective governance of the country. Furthermore, efforts during the Taika Reform era to adopt a Chinese-style administration and military are dismissed as complete failures, abandoned only a few decades after their inception. As the court “became isolated to an extraordinary degree from the rest of Japanese society,”[1] and could no longer provide an effective military or police system, “provincial residents were forced to take up arms for themselves…[which] allowed the development of large, private warrior networks.”[2] In their respective works, both Karl Friday and William Farris seek to revise this misperception and argue that “the genesis of Japan’s bushi [warrior class] took place within a secure and still-vital imperial state structure.”[3] In Hired Swords: The Rise of Private Warrior Power in Early Japan, Karl Friday traces the evolution of Japan’s military system, from the foundations laid by the Taika Reforms in 645 to Minamoto Yoritomo’s “epoch-making usurpation of power in the 1180s,”[4] to prove it was court activism that concentrated military control in the hands of the rural elite. Furthermore, Friday believes that the court’s growing reliance on the private martial skills of the gentry was motivated by the desire to maximize the efficiency of its military institutions and reflected the changing nature of Japan’s military needs.[5] William Farris advances Friday’s argument in Heavenly Warriors: The Evolution of Japan’s Military, 500-1300 by arguing that the samurai class of the fourteenth and fifteenth centuries was the “direct descendant of the mounted archers of yore…[and maintains] that an equestrian mounted elite was a critical factor in society, economy, and politics as early as about A.D. 500.”[6] However, while Farris asserts that the imperial reforms were essential to the evolution of Japan’s mounted military elite, he does not support Friday’s belief that the court successfully took control of the military from the hands of provincial elite. In the context of these two works, the evolution of Japan’s military can be divided into three stages: the centralization of military control and the adoption of Chinese-style mass infantry tactics under the ritsuryō codes during the eighth century, the subsequent ‘abandonment’ of infantry in favor of ‘the privately acquired  martial skills of provincial elites and the lower nobility,’[7] and the further organization of private military networks around major provincial warriors during the mid-tenth and eleventh centuries.

Emperor Tenji 天智天皇 (626-671)

At the beginning of the seventh century, Japan was controlled by a loose confederation of great houses. In theory, the royal Yamato house was the most powerful and had the authority to grant regional administrative control to lesser noble houses by empowering regional chieftains as kuni no miyatsuko, or provincial patriarchs. However, “kuni no miyatsuko appointments were permanent and hereditary” and provincial patriarchs had virtually independent control of their territories.[8] After Emperor Tenji seized power in a bloody coup d’état in 645, “[he] and his supporters introduced a series of Chinese-inspired reforms collectively known as the Taika Reforms” that gradually stripped the great regional families of their independence and made them subservient to the interests of the state.[9] Prior to the reforms, armies were “knit together from the private forces of the kuni no miyatsuko,”[10] who controlled both the conscription and training of peasant soldiers. While the function and personnel of the army remained the same – peasants continued to be conscripted as soldiers and regional nobles continued to serve as officers – the Taika Reforms changed the military structure in an important way. Private conscription through the kuni no miyatsuko was replaced by direct conscription by the state and was supervised by the imperial court.[11] However, though these reforms were issued on paper in 645, it was not until the rule of Emperor Tenmu[12] (or Temmu) and his successors that the true centralization of Japan’s military resources began.

During the Jinshin Civil War of 672, Tenmu forced his own nephew to commit suicide and seized control of the imperial throne. Tenmu’s successful rise to power owed much to “the private forces of disaffected members of the provincial nobility.”[13] Without the assistance of these private mounted forces, Tenmu’s victory would have been substantially less decisive, if it was achieved at all. Acutely aware of the “danger inherent in an independent military force,”[14] Tenmu began a series of reforms, collectively called the ritsuryō codes, designed to centralize control of the country’s military. His decree of 684 declared, “In a government, military matters are the essential thing.”[15] Both Farris and Friday identify this declaration as the beginning the court’s attempt to firmly concentrate control of the country’s military resources into imperial hands and the first step towards Japan’s gradual evolution into a military state dominated by the warrior class known as the samurai.

Emperor Tenmu 天武天皇 (631-686)

Emperor Tenmu’s reform of the imperial military system was motivated by two threats: a Chinese invasion from the mainland and the “near-continuous warfare with the emishi in the northeast.” [16] The court’s solution (following a revision of the Chinese model) was to conscript peasants, place them under strong central control, and train them into an effective fighting force. Farris points out four benefits to this system; it was cheap, was explicitly short-term in nature because soldiers were also farmers, was primarily defensive, and “severely restricted the role of locally powerful, hereditary military families.” [17] Under the reforms, “all free male subjects between the ages of twenty and fifty-nine…were liable for induction as soldiers, or heishi.”[18] After conscription, heishi were assigned to a provincial regiment and selected for one of three types of duty; watch duty within their own provinces, service as guards in the capital or on the frontier, or for mobilization for major military campaigns in or out of the country.[19] Despite the predominance of mounted cavalry as the “elite technology of the day,”[20] the state’s reliance on an army composed of primarily infantry troops made sense in the context of the presumed military threats. Almost certainly, a well-organized infantry would have been the most effective defense against a large invasion from China, who also employed mass infantry tactics. Additionally, the court’s military policy regarding hostile emishi[21] forces in northeast Japan was primarily defensive. Infantry troops were ordered to defend the colonists with a series of stockades “into which they could retreat in times of trouble.”[22] Because the emishi were comprised almost entirely of small-units of mounted horsemen, the ritsuryō armies were more than up to the task of defending these fortified settlements.

An important aspect of the military reforms was the organization and command of imperial troops. The ritsuryō codes placed every province under the centralized command of the state, directly administered by a court-appointed provincial governor. Governors were responsible for the training and organization of troops, maintenance of weapons, fortifications and livestock, and had a hand in selecting officers (gunki) in charge of the regiment. In addition to important military functions, the governor’s office, or provincial headquarters, “operated granaries, collected taxes, oversaw the government transportation system and encouraged agriculture.”[23] Each provincial regiment was composed of 500 to 1,000 men and was “overseen by one or more colonels (ki or gunki).”[24] This system was meant to further centralize government control and reduce the influence of prominent families by establishing two layers of administration over the peasantry: the provincial governor (selected from the aristocracy) had supreme control of the province and then delegated control to district magistrates, chosen from locally prominent families.[25] However, rather than subordinating the authority of district magistrates to the provincial governor and the court, this system only reinforced the necessity of local families in the organization and command of troops. Provincial governors were chosen from the higher ranks of the court nobility and, hailing from the capital, could hardly be expected to be familiar with the province they were sent to administer. Therefore, it is unlikely that they were able to “penetrate and control the countryside…without the assistance of the local elite.”[26] Aware of this dependence, provincial governors often appointed the heads of elite local families to the post of district magistrate. Unsurprisingly, these local families were the descendants of the kuni no miyatsuko (provincial patriarchs), who had controlled regional military power prior to the 645 reforms.

Fujiwara no Kamatari

The nature of mounted military skills further concentrated power into the hands of the regional elite. Mounted archers were the most effective military technology of the time and the state tried to maintain the largest cavalry possible. However, they were confronted with two major logistical problems – they could not afford the expense of raising, training, and keeping a large stable of horses and, given the time and expense required to master fighting from horseback, it was ‘impossible to produce first-rate cavalrymen out of short-term, peasant conscripts.’[27] Therefore, the state preferred to draw their cavalry troops from “those who already possessed the necessary skills, and by making the care of the state’s war-horses the responsibility of private parties.”[28] It goes without saying that these positions would fall to those who could devote their time to the acquisition of equestrian military-skills and afford the considerable expense of keeping a horse. Naturally, these positions were delegated almost exclusively to wealthy provincial elites. Additionally, under the gunki, control of the troops fell to those skilled in martial and equestrian practices. In other words, those appointed to lead regiments “were drawn from the same socioeconomic class as the district officials…that is, from those houses that had formerly been provincial patriarchs.” This raises the question of how deeply the organization of the military was changed by the reforms. As Farris states, “Instead of severely restricting the role of local strongmen [kuni no miyatsuko] in the new army…both law and practice handed command over to the local elite.”[29] Not only did the provincial governor owe the successful administration of the province to the cooperation of district magistrates, but the organization and training of conscripted peasant troops was dependent on the enlistment of the military skills of that same class. As Farris explains, “by authorizing these local officers to command the troops…the Court was both making a concession to pre-645 Japanese military organization and also violating an important Chinese principle against the control of the military by local leaders.”[30]

Regardless of the court’s success in creating a centrally controlled army, the changing reality of warfare in the ninth century led to the ‘abandonment’ of peasant conscripts in favor of the privately acquired military skills of provincial elites. Midway through the eighth century, a rebellion in Central Asia ‘shook the T’ang dynasty to its foundations’ and eliminated the threat of a Chinese invasion of the Japanese archipelago.[31] Additionally, the court changed its policy towards the emishi people from one of stationary defense to ‘pacification.’ This pacification policy meant the Japanese imperial state was “now attempting to control not just bits of territory but the (hostile) populations residing on and round them as well.”[32] The hit-and run guerilla tactics of the emishi people were now much more effective against an isolated and unfortified infantry. It soon became clear that “small, highly mobile squads that could be assembled with a minimum of delay and sent out to pursue raiding bandits were far more appropriate” to address the emishi as well as capture criminals and bandits in the provinces.[33] However, the court did not ‘abandon’ the military reforms of the previous century as a failure, but “reached the conclusion that it was more efficient – and probably cheaper as well – to rely on the privately acquired military skills of the provincial elite than to continue to attempt to conscript and train the peasantry.”[34] The expanding military role of local elites opened new avenues for personal advancement to provincial and low-ranking aristocrats, and increased the incentive for men to privately acquire military skills.

Ninth-century Japan was beset with decreasing agricultural production, famine, and constant epidemics.[35] Both the state and private houses were forced to compete over a diminishing amount of wealth. This increased competition had two important effects on military organization. First, aristocrats in the capital cultivated personal ties with the provincial elite to secure a source of independent revenue.[36] This led to the development of vertically-integrated private warrior networks. These vertical relationships were mutually beneficial, because the increasing stratification of Heian society meant “the rights and privileges of each stratum [class], were sealed from below, as well as from above.”[37] This ensured that the interests of court nobles never conflicted with those of the provincial elite. Secondly, the growing instability of the countryside and the increasing number of bandits forced the court to place “new emphasis on the provincial administration as the only base of imperial power in the countryside.”[38] These two factors combined to reinforce the growing separation of the court and private military networks. As vertical relationships between individual aristocrats and provincial governors solidified and the state granted increasing military independence to provincial administrations, court nobles soon found themselves the de facto heads of well-organized, private warrior networks. Provincial governors used their posts to establish military networks in a province but would stay well past the end of their appointment, preferring to exploit their military bases and keep the tax-revenues for themselves.[39] Ironically, to put down these cases of aristocratic banditry and rebellion in the countryside, the imperial court was forced to rely on the same vertically-integrated warrior networks. By offering aristocrats rewards and court appointments, the state was able to employ the military networks of noble families in one province to suppress a rebellion in another. Obviously, the court’s reliance on this system of military organization further reinforced the development of private warrior networks, headed under an aristocratic court family. By the end of the century, the military system was composed to two pillars; the aristocratic warriors of the capital and the fighters based in the provinces.[40] By the mid-tenth century, central aristocratic houses came to be hereditarily identified with martial skills.[41] This military organization along family lines was an important forerunner to the samurai class.

Taira no Masakado

During the previous century, the court had experimented ‘through a process of trial-and-error’ with the organization of an imperial military and police structure that would replace the Chinese-style ritsuryō codes. By the mid-tenth century, this organization was formalized around three important institutions; the ōryōshi, tsuibushi, and tsuitōshi.  Each position was created to deal with a threat to imperial power. The ōryōshi post was created to oversee the recruitment and training of troops. Such positions were usually delegated to district officials and were clearly designed to “exploit the military talents of rural elites.”[42] The court intended to place the growing power of rural elites under the control of the state by “giving the warriors a stake in the survival of the polity.”[43] The creation of the tsuibushi post was similarly motivated. Tsuibushi were central court figures, appointed on a standing basis to hold jurisdiction and law enforcement responsibilities over a single province.[44] As expected, the court figures who received appointments already possessed well-developed military networks in the region. Again, this merely perpetuated the problems of the ninth century. Because the court had no independent military resources, it was forced “to depend on the very class that was most active in outlawry to preserve the peace.”[45] Tsuitōshi was a major title developed in the late tenth century and was designated to extremely powerful aristocratic warriors to put down especially troublesome rebellions and incidents of piracy, usually committed by another aristocratic warrior. Farris provides thorough descriptions of two such incidents, the rebellion of Masakado and the pirate Sumitomo.[46] Unlike the position of tsuibushi, tsuitōshi appointments were originally intended to be temporary and to fulfill a specific objective.[47] Tsuitōshi were granted exclusive control of the entire campaign, which put them at the heads of large and powerful warrior networks. In under a century, this position had also become hereditary, further castrating the control of the imperial state over Japan’s military.

While Karl Friday admits that the creation of these institutions perpetuated the development of private martial organizations, he insists that “they were established in an effort to upgrade the provincial and military system.”[48] Unfortunately, it appears that the creation of these three offices was motivated by precisely the opposite – a scramble to block the steady encroachment of the warrior class by “attempting to buy off bushi leaders with ad hoc military positions.”[49] The court had become further entrenched in its dependence on the vertical military networks of aristocratic warriors in the capital and their counterparts in the countryside, either the provincial governors or the heads of elite military families. Despite his earlier contentions, Karl Friday best depicts the nature of court control in the conclusion of Hired Swords:

“The civil nobility was able to keep warriors at heel by setting them against themselves. Bushi were made to compete with each other for the titles that permitted them to exercise armed force and for the rewards that accompanied military service at court. Whenever powerful warriors, such as Taira Masakado or Taira Tadatsune, dared to rise and challenge the central government…there were always peers of rivals…who were readily persuaded that cooperation with the state was a surer path to success.”[50]

Clearly, this solution merely perpetuated the court’s growing dependence on aristocratic warrior networks and their increased separation from the countryside. Like a house of cards, this instability compounded upon itself until Minamoto Yoritomo seized power the following century.

Minamoto no Tametomo

The success of Karl Friday’s argument rests in the distinction between the centralization of imperial state structure and its power. While Friday argues that the imperial court successfully maintained the centralized nature of the state’s administration until at least the twelfth century, he does not equate this centralized control with power. To the contrary, it is quite clear that the state’s ability to assert its power over the provinces was under significant threat as early as the mid-ninth century, with the organization of private warrior groups and the court’s growing dependence on professional mercenaries. Friday also neglects to connect court edicts and laws, which he thoroughly documents and explains throughout Hired Swords, with specific examples of their successful implementation in the provinces. The mere existence of an imperial military and police system does prove that the state effectively administered these policies in the Japanese hinterlands. Furthermore, Friday does not convincingly demonstrate that the imperial court had the ability to preserve its existence at any time between the sixth and twelfth centuries without the explicit cooperation of regional military networks. In truth, the continuation of imperial control was largely due to the court’s successful manipulation of the zero-sum nature of power for provincial warriors throughout the mid-Heian period. Under these circumstances, court control can only be considered precarious at best. The moment the court could no longer successfully play provincial powers against each other with the promise of rewards and advancement, or when a warlord became too powerful to be suppressed by his neighbors, than the true power of the court (or lack thereof) would be exposed. This is precisely what happened in the 1180s when Minamoto Yoritomo placed the law-making apparatus of the state under his personal authority.

However, Friday’s thorough explanation of the evolution of the imperial military and state structure and his meticulous documentation of sources more than makes up for his less-than-convincing argument. In comparison, William Wayne Farris’ Heavenly Warriors seems poorly organized and beset with tiresome repetitions. Clearly, the task of Heavenly Warriors – to trace the development of the Japanese warrior from the sixth to fourteenth century – is quite ambitious. However, Farris often allows his argument to be lost among frequent inclusions of interesting, but largely irrelevant, secondary sources. Farris also overemphasizes the structure of fourteenth century Japanese warrior in his treatment of earlier periods, which gives the impression that he has organized the historical evidence to support a predetermined conclusion, namely that the basic reality of the Japanese warrior order remained largely unchanged until the 1300s. Despite his repetitious and, at times, overly literal argument, Farris’ thorough descriptions and explanations of major military confrontations ultimately makes Heavenly Warriors an interesting and extremely informative work. While both Hired Swords and Heavenly Warriors have shortcomings, when read together they offer a thorough and extremely accurate depiction of pre-feudal Japan’s state and military institutions and help unify the image of the samurai with the early mounted archer.

Works Cited

Farris, William Wayne. Heavenly Warriors: The Evolution of Japan’s Military, 500-1300. Cambridge: Harvard University Press, 1992.

Friday, Karl F. Hired Swords: The Rise of Private Warrior Power in Early Japan. Stanford: Stanford University Press, 1992.

Varley, Paul. Japanese Culture. Fourth ed. Honolulu: University of Hawai’i Press, 2000.


[1] Paul Varley, Japanese Culture,4th ed. (Honolulu, 2000): 52.

[2] Karl F. Friday, Hired Swords: The Rise of Private Warrior Power in Early Japan (Stanford, 1992): 7.

[3] Friday, 6.

[4] Friday, 176.

[5] Friday, 7.

[6] William Wayne Farris, Heavenly Warriors: The Evolution of Japan’s Military, 500-1300 (Cambridge, 1992): 3.

[7] Friday, 71.

[8] Friday, 10.

[9] Friday, 10.

[10] Friday, 11.

[11] Friday, 11-12.

[12] 天武天皇 “tenmu-tennō” – His name is also commonly written as Temmu. He was the 40th emperor of Japan and ruled the country from 673 to 686.

[13] Friday, 12.

[14] Farris, 46.

[15] Farris, 1.

[16] Friday, 47.

[17] Farris, 48.

[18] Friday, 14.

[19] Friday, 15-17.

[20] Friday, 38.

[21] Emishi (蝦夷) were a group of people who lived in the present day Tōhoku region of northeastern Honshū. Some emishi tribes resisted Japanese imperial rule from the 7th to 10th centuries.

[22] Friday, 48.

[23] Farris, 51.

[24] Faris, 17-18.

[25] Farris, 46-47.

[26] Farris, 52.

[27] Farris, 39.

[28] Farris, 39.

[29] Farris 357.

[30] Farris, 52.

[31] Friday, 44.

[32] Friday, 51.

[33] Friday, 44.

[34] Friday, 49-50.

[35] Farris, 123.

[36] Farris, 125.

[37] Friday, 77.

[38] Farris, 126.

[39] Farris, 126.

[40] Farris, 161.

[41] Friday, 88.

[42] Friday, 143.

[43] Friday, 147.

[44] Friday, 153.

[45] Friday, 156.

[46] Farris, 162-240.

[47] Friday, 160.

[48] Friday, 166.

[49] Friday, 166.

[50] Friday, 176-177.

Tags: , , , , , , , , , , , ,


Comments (2)

  1. amannin


    Very interesting review / comparison of the two titles. More food for thought when it concerns the genesis of the samurai.

    If one were to read “Hired Swords” then “Heavenly Warriors,” what might be a good continuation onto the next subject of Feudal Japan (sharing a similar context or correlation to Friday and or Farris)?

    September 7, 2010
    • constantineintokyo


      Well, “Heavenly Warriors” brings you all the way up to the end of the Kamakura period, which is when the Genji (Minamoto Yoritomo and crew) had established the position of the Shogun as the de facto ruler of Japan and the emperor’s power was largely ceremonial. Even though Farris covers the beginning of the Kamakura period, I would still recommend reading the Heike monogatari, to get a look at the period through a primary source. Then, the Muromachi Period lasted from 1336-1573. The Sengoku Period falls within the Muromachi Period, from roughly 1467 to 1573. The Edo Period (under the Tokugawa bakufu) begins in about 1600. “Legends of the Samurai” by Sato Hiroaki is also a good collection of primary sources.

      After different samurai warlords (daimyo) established control over certain areas, Japan just fell into a period of constant battle (Sengoku). In all honesty, my knowledge of the Muromachi Period is really bad. I tend to jump right from Minamoto Yoritomo to Oda Nobunaga. But, I’ll tell you what I know:

      -“Warrior Rule in Japan” edited by Marius Jansen – It is a collection of essays by noted historians. It really covers the entire period of samurai rule and offers an in-depth look at the development of the bakufu (Shogunate).
      – Books by Stephen Turnbull – He concentrates more on battles, military tactics, and the personalities of specific individuals.

      Books about the late 1500s and the Edo Period –

      – “Hideyoshi” by Mary Elizabeth Berry (about the life of Toyotomi Hideyoshi)
      – “The Taming of the Samurai: Honorific Individualism and the Making of Modern Japan” by Ikegami Eiko (really good book on samurai from the late 1500s onwards)
      – “Shinsengumi: The Shogun’s Last Samurai Corps” by Romulus Hillsborough (about the end of the Edo period)

      However, I’m really a modern Japanese history girl – the 1600s onwards is what I know best. I’d have to ask Nino what books he likes about the Muromachi Period.

      September 7, 2010

Comments are now closed for this post.