A Historical Understanding of the “Ninja”

Irregular Warfare in Late Medieval Japan:

Towards a Historical Understanding of the “Ninja”1

Polina Serebriakova and Danny Orbach

The Japanese ninja are known in the west mainly by the romantic
image of the secret agent, or assassin, endowed with superhuman
powers. Yet there is a lack of serious scholarship on the actual war-
riors who became the subject of the myth. This paper seeks to use
the full extent of primary sources from the period to offer a blueprint
for a historical, rather than mythical understanding of the ninja. They
did not belong to a unified class, but were instead a hodgepodge of
heterogeneous groups, united only by their skills in irregular warfare.
They became “ninja” only retrospectively, as a result of the historical
imagination of later generations.

N injas are known in the west mainly by the romantic image of the secret
agent or assassin, endowed with superhuman powers, dressed in black,
and throwing deadly shuriken (“ninja stars”). This image first emerged in Japan
in the seventeenth century through the Kabuki theater and other forms of folk
art. Subsequently it found its way to the West, where it was nurtured by the
postwar boom of interest in martial arts and the exotic Far East. It was further
popularized, even immortalized, by the movie “You Only Live Twice,” the fifth
film in the James Bond series.

1. The authors are grateful to Nathan H. Ledbetter, Dan Sherer, and Simon Frankel Pratt for
their help and advice, and to Stephen Turnbull for generously sharing a rare historical source.

Polina Serebriakova is a Ph.D. Candidate in Japanese Studies at the University of Cambridge.

Her research focuses on the legitimacy of warrior leaders and political rituals exercised by the
warrior class in late-medieval Japan.
Danny Orbach is a senior lecturer for history and Asian Studies at the Hebrew University of
Jerusalem. He specializes in military coup d'etats, the role of military adventurers in modern
warfare and the history of intelligence.

The Journal of Military History 84 (October 2020): 997-1020.

Copyright © 2020 by The Society for Military History, all rights reserved. No part of this publication may be reproduced, stored, or trans-
mitted in any form or by any means without the prior permission in writing from the Editor, Journal of Military History, George C.
Marshall Library, Virginia Military Institute, P.O. Drawer 1600, Lexington, VA 24450. Authorization to photocopy items for internal
and personal use is granted by the copyright holder for libraries and other users registered with the Copyright Clearance Center (CCC),
121 Rosewood Drive, Danvers, MA 01923 USA (www.copyright.com), provided the appropriate fee is paid to the CCC.

MILITARY HISTORY  997

P. SErEBrIAKOvA & D. OrBACH

The myth is loosely based on real irregular warriors who lived in sixteenth-
century Japan, during the long succession of simultaneous civil wars known as
Sengoku jidai (Warring States Period). In this violent era, conventionally dated
1467 to 1603, the Japanese realm experienced incessant warfare between numerous
warlords, monastic confederations, provincial leagues, and miscellaneous entities
of various sizes.2 Most fighting was done by regular, standing armies, but local
militiamen, mercenaries, brigands, freelance spies and other irregulars also took
part in military operations. Some of them were retrospectively called “ninja.”
The history of these irregular warriors is shrouded in mystery. Few reliable
sources remain, with many documents torched during the civil wars. The “ninja”
became so mysterious as to deter most scholars, a fact that may explain the
extremely low yield of scholarly (in contrast to popular) books on the subject. In
recent years though, more researchers are acknowledging these irregular warriors as
an indispensable part of pre-modern Japanese military history. This is mostly due to
the cooperation between Mie University and the Iga-ryū Ninja Museum, which has
promoted the analysis, interpretation, and digitalization of relevant documents, as
well as to the efforts of scholars like Stephen Turnbull and Yamada Yūji.3
This paper continues this line of research by offering a blueprint for a
historical, rather than mythical understanding of the late medieval Japanese
warriors known today as ninja. After discussing the contemporary and historical
terminology, it examines the irregular warriors who fought in the service of
standing armies, in functions resembling early modern European privateers, or
the commandos of later periods.
The second part of the article is devoted to an interesting episode in which
rural samurai from the Iga region south of Kyoto tried to employ “ninja” tactics in a
guerrilla campaign against an invading army. Their failure, we argue, is typical of the
limitations of irregular warfare against a multi-pronged invading force engaged in
annihilation tactics. The episode is a reflection of a larger state-building process, the

2. There is no single-volume, comprehensive survey of the Sengoku civil wars in English, an

understandable fact, given the complexity of the events, and the vast historical and geographical
range in which they took place; however, there are some excellent studies. See, for example, Mary
Elizabeth Berry’s two seminal monographs: Hideyoshi (Cambridge, Mass.: The Council on East
Asian Studies Harvard University, 1989), and The Culture of Civil War in Kyoto (Berkeley: University
of California Press, 1997), as well as Jeroen P. Lamers, Japonius Tyrannus: the Japanese Warlord Oda
Nobunaga Reconsidered (Leiden: Hotei, 2000); David Spafford, A Sense of Space: The Political Land-
scape in Late Medieval Japan (Cambridge, Mass.: Harvard University Asia Center, 2013); Hitomi
Tonomura, Community and Commerce in Late Medieval Japan: The Corporate Villages of Tokuchin-ho
(Stanford, Calif.: Stanford University Press, 1992), as well as Carol Tseng, War and Faith: Ikkō Ikki
in Late Muromachi Japan (Cambridge, Mass.: Harvard University Asia Center, 2007).
3. Stephen Turnbull, Ninja: Unmasking the Myth (Barnsley, U.K.: Frontline Books, 2017);
Yamada Yūji, Ninja-no Rekishi (Tokyo: Kadokawa, 2016). The most comprehensive list of ninja-
related scholarship can be found in databases created by researchers from Mie University: http://
www.human.mie-u.ac.jp/en/projects/iga/database-jp.html (works in Japanese), http://www.hu-
man.mie-u.ac.jp/en/projects/iga/database-foreign.html (works in other languages).

998  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

transition of the Japanese realm from a chaotic arena of social ambiguity, mobility,
and irregular warfare, to an early modern polity based on a more rigid system of
conventional armies and groups with inherited social status.

The Origins of the Term “Ninja”

The term ninja (忍者) is a twentieth-century pronunciation of a more
ancient compound from the late sixteenth century, pronounced then as shinobi-
mono. The character 忍 (read as nin or shinobi), means “to endure” or to “bear”
in both classical Chinese and Japanese. In late medieval Japanese it acquired an
alternative meaning of “sneaking,” “remaining unseen” or “doing things secretly.”
The character 者, pronounced as sha, ja, or mono, denotes a person or a practitioner
of a certain profession, like in the famous term “geisha” (practitioner of the fine
arts). Ninja or shinobi-mono, different pronunciations of the same word, therefore
denotes “a practitioner of the art of sneaking.”4
The term appeared infrequently in the late sixteenth century, during the last
years of the Sengoku civil wars. In its first known appearance (1581), shinobi-
mono referred to hand-picked, crack troops from Iga (a place central to later ninja
lore) dispatched to a neighboring province to conduct nocturnal reconnaissance
and abduct enemy samurai, merchants, and peasants for ransom. It appears again
in 1589, when it denotes an emissary who held secret peace talks in the Amakusa
Islands, Higo Prefecture.5 Then, in 1603, the first year of the Edo Period,
immediately after the end of the civil wars, “xinobi” reappeared in a Portuguese–
Japanese dictionary to mean a “spy who sneaks into castles to gain intelligence,”
along with two associated verbs: “xinobiuo suru” (to investigate) and “xinobiga
itta” (to enter as spy).6
It popped up with greater frequency during the mid-seventeenth century,
decades after the irregular warriors to whom it refers had ceased to exist. The Buyō
Benryaku, a work complied in 1684, offers this definition:
The Shinobi-mono are men who are able to hide themselves in their
own country as well as in other countries, or to sneak into the
enemy’s castle, strong as it may be, in order to steal its secrets …
recently, the men of Iga and Kōka [were known] as a kind of
people who are able to peep into the enemy’s secrets. From times
of yore, the inhabitants of Iga and Kōka distinguished themselves
in this way, transmitting their knowledge from generation to
generation.7
By using the term shinobi-mono, the author defined the men of Iga and Kōka
as professionals of irregular warfare. Later histories, written in the nineteenth

4. Turnbull, Ninja, 6–7.

5. Turnbull, Ninja, 42, 50.
6. Doi Tadao, ed., Vocabvlario da lingoa de Iapam com adeclaracao em Portugues (Tokyo: Iwa-
nami Shoten, 1960), 606.
7. Quoted in Sasama Yoshihiko, Buke Senjin Sakuhō Shūsei (Tokyo: Kōdansha, 1970), 84–85.

MILITARY HISTORY  999

P. SErEBrIAKOvA & D. OrBACH

and twentieth centuries, mythologized the shinobi-mono, or “ninja,” as clans of

irregular warriors that had existed from time immemorial. Thus the foundations
for the modern ninja myth were laid.
During the Sengoku civil wars though, the term shinobi-mono was rare,
and irregular warriors were usually not described either as a cohesive social or
professional group, or as a well-defined warrior type, in contrast to later Japanese
depictions or comparable discussions of “partisans,” “guerrilla fighters,” or
“irregulars” among Western theoreticians.8 Usually, the term shinobi appears as
an adverb, denoting an activity, and not as a noun referring to a person or to a
particular type of conflict (i.e. guerrilla or low-intensity war). In line with M.L.r.
Smith’s more general observations, guerrilla warfare in late medieval Japan was
perceived as a tactic within war, not a separate category of warfare.9
Contemporary sources accordingly reveal that a number of peasants, priests,
foot soldiers, and samurai were employed by their respective lords in military
assignments that in modern terms might be defined as “commando” or “guerrilla”
operations. Such activities, almost always a small part of larger conventional
conflicts, included the nocturnal infiltration of castles, sabotage, espionage,
reconnaissance, and operations of counter-intelligence. People engaged in such
activities never formed a common identity. They did not belong to organized
schools or to secret clans. The great majority was not even engaged full time
in irregular warfare. Instead, many people who were later described by modern
popular authors as “ninja” were fulfilling routine military, civilian, or religious
tasks most of the time, and were involved only occasionally in warfare. Almost
all of them had ties, either formally or informally, with warlord entities and other
hierarchical structures of political authority.
Some were commoners or Buddhist clergymen. In her study of the Ōyama
pilgrimage system, Barbara Ambros notes that the monks of Ōyama doubled as
mountain guides, spies, and reconnaissance troops for local warlords during the
Sengoku civil wars. Likewise, Sakamoto Tokuichi and Sasamoto Shōji point out
that Oshi (itinerant Shinto priests), wandering poets, and merchants also were used
as spies—a convenient measure because their lifestyle gave them a good reason for
moving between territories of different lords.10 Warlord Takeda Shingen, a major
actor in sixteenth-century Sengoku politics, often dispatched spies, in many cases
itinerant monks and merchants, deep into enemy territory.11 Sasamoto provides

8. Carl Schmitt, The Theory of the Partisan: A Commentary/Remark on the Concept of the Po-
litical. Trans. A.C. Goodson (East Lansing: Michigan State University Press, 2004), 6–14.
9. Turnbull, Ninja, 11–12; M.L.r. Smith, “Guerrillas in the Mist: reassessing Strategy and
Low Intensity Warfare,” Review of International Studies 29 (2003): 37.
10. Barbara Ambros, Emplacing a Pilgrimage: the Ōyama Cult and Regional Religion in Early
Modern Japan (Cambridge, Mass.: Harvard University Press, 2008), 10; Sakamoto Tokuichi,
Takeda Shingen, Sonshi no Senpō (Tokyo: Shin Jinbutsu Ōraisha, 1987), 84–86; Sasamoto Shōji,
Sengoku Daimyō Takeda Shi no Shinano Shihai (Tokyo: Meicho Shuppan, 1990), 191.
11. Sasamoto, Sengoku Daimyō, 191.

1000  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

convincing evidence for Takeda’s habitual use of spies, especially merchants. He

quotes written orders from Takeda to one of his merchant-spies, as well as a letter
sent by Uesugi Kenshin to one of his vassals ordering him to keep an eye on
suspicious behavior of wandering merchants. Takeda’s intelligence operations, as
well as his strategic worldview, were greatly influenced by the classical Chinese
military thinker Sun Tzu, whom he greatly admired.12 A military manual written
in 1653 (fifty years after the end of the civil wars), reports that shinobi spies used a
variety of camouflage techniques and regularly dressed as monkey trainers, street
entertainers, or itinerant priests (yamabushi).13
There were also gangs of brigands known as rappa who served warlords as
long-term contractors, all the while performing certain tasks of irregular warfare.
They belonged to the wider category of “evil bands” (akutō), a hodgepodge of
mercenaries of lower classes, tax evaders, lowly samurai who refused to work for
liege lords, estate overseers who rebelled against their landlords, and other outcasts.
In some cases, such people took to robbery. In others, they were only defined as
akutō by their rivals and treated as such when the occasion arose.14
As in many civil wars around the world, the chaotic reality of the Sengoku era
“privatized politics”; namely, it blurred the distinctions between public and private
conflict, between legitimate warfare and crime. People became used to solving private
quarrels with violent action, often enlisting the support of nearby interested parties such
as warlords, armed monks, urban and rural associations, brigands, or rogue warriors.15
The rappa lived in this legal grey zone, somewhere between the categories of warrior
and criminal, but were almost always tied with warlords, either officially or informally.
Their description in the chronicle of the Hōjō clan is typical in that respect:
Once, chaos reigned… and one always had to keep his bows and
arrows at hand. Back then, there were these villains called rappa.
They were thieves, but at the same time not thieves, admirably
cunning and wicked at the same time … They were granted a stipend
as a part of a warlord’s army … These people skillfully looked for

12. Sasamoto, Sengoku Daimyō, 191; Shan Qun Yang, Sun Zi (Taipei: Zhi Shu Fang Chu
Ban She, 1996), 311, and Ota Gyūichi, The Chronicle of Lord Nobunaga (Shinchō-kō Ki), trans.
J.S.A. Elisonas and J.P. Lamers (Leiden: Brill, 2011), 83–85).
13. Turnbull, Ninja, 102–3.
14. Berry, The Culture of Civil War, 69–70; Dan Sherer, “The Akutō on Ōbe Estate: Lawsuits,
Evidence and Participation in the Late Kamakura Legal System,” in Janet r. Goodwin and Joan
r. Piggott, eds., Land, Power and the Sacred: The Estates System in Medieval Japan (Honolulu:
University of Hawaii Press, 2018), 403–47. Compare this description of Akutō with Joseph
Mackay’s definition of “escape societies,” based on James Scott’s classic study, in Joseph MacKay,
“Pirate Nations: Maritime Pirates as Escape Societies in Late Imperial China,” Social Science
History 37, no.4 (Winter 2013): 551–73; James Scott, The Art of not being Governed: An Anarchist
History of Upland Southeast Asia (New Haven, Conn.: Yale University Press, 2009).
15. Berry, The Culture of Civil War, 7, 13, 70, 98-102, 136–137. For a comparative perspec-
tive and theoretical analysis, see Stathis N. Kalyvas, The Logic of Violence in Civil War (Cam-
bridge: Cambridge University Press, 2006), 79–80, 389.

MILITARY HISTORY  1001

P. SErEBrIAKOvA & D. OrBACH

thieves in our own province, found them and slit their throats.
However, they sneaked into other provinces to engage in mountain
robbery, piracy, night assaults, and abductions, in all of which
they were skilled. They had the talent to devise cunning schemes,
which ordinary people could not even think about […] As for their
wisdom, they were almost as the Gods and the Buddhas.16
Then comes a long story of an attack by a band of 200 rappa, led by an “evil
person” named Kazama or Fūma Kotarō (retrospectively lionized as one of the
legendary “ninja” leaders). According to the chronicle, Kazama led a mixed band.
One of his officers was a “mountain bandit,” another was a “pirate,” and the two
others were “thieves and robbers.” All of them were extremely familiar with the
terrain and had shinobi skills, and could therefore infiltrate the enemy positions to
create havoc in its ranks.17
In letters of the Hōjō clan from the late sixteenth century, there are few or
no descriptions of the members of this gang, but there are some complaints about
their wild behavior. These rappa should have been controlled by the retainers of
the Hōjō clan:
Fūma shall arrive [and reside] until the seventh month and shall be
stationed in six villages; thus, lodging and other facilities shall be
provided for. In case of emergency a part of the income … shall be
distributed. Even the slightest disorder that can be partially traced
to Fūma is unacceptable and shall be reported.18
As many as five retainers are named in the letter as responsible for the safe
stationing of Fūma’s troops in the villages. Such precautionary measures were
common at this time, since warriors of all kinds, including rappa, were known
to oppress peasants and villagers. Fūma’s case is more complicated though, as
revealed in a letter written a year later:
Since Fūma was ordered to leave for his place of origin, I thought
[Fūma’s troops] were not staying in Sunawara. Yet now [they] are
residing [there]! This causes disturbance to the peasants [who] have
filed a complaint, thus [we] order [that] from now on Fūma shall
not be stationed [there anymore].19
Sunawara is situated in Adachi District, Musashi Province, whereas the origin
of Fūma troops is believed to be Ashigarashimo District, Sagami Province.20 The fact

16. Imamura Yoshitaka, ed., Sengoku Shiryō Sōsho, series II (Tokyo: Jinbutsu Ōraisha, 1966),
1:397–98.
17. Imamura, Sengoku Shiryo 1:398.
18. Sugiyama Hiroshi and Shimoyama Haruhisa, eds., Sengoku ibun, Go-Hōjō-shi hen (To-
kyo: Tōkyōdō Shuppan, 1989-1995), “Copy of the red Seal Document,” doc. 1595.
19. Sugiyama and Shimoyana, Sengoku ibun, Go-Hōjō-shi hen, “Copy of red Seal Judgment
Document,” doc. 1677.
20. Ueda Masaaki, et al., eds., Nihon Jinmei Daijiten (Tokyo: Kodansha, 2001), Japan
Knowledge Library, https://japanknowledge.com/lib/display/?lid=5011061746480. Accessed 27
July 2020.

1002  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

that these rappa were not eager to return to their homes suggests that the Fūma group
did not have strong agricultural connections to their homeland. Their employers from
the Hōjō clan, moreover, hardly seem to be in control of their movements.
Fūma’s fame had spread far and wide in eastern Japan; his name can be spotted
in other chronicles, such as those written by Hōjō’s rivals from the Takeda clan.
There, Fūma is described as a bandit similar to a suppa (a synonym of rappa).21 Even
after the Takeda clan was destroyed in 1582, suppa existed as a military formation
and continued to be employed against the Hōjō. The grounds for this statement
come from the following letter by Hōjō Ujikuni, sent either in 1582 or 1588:
right now I inform you: five hundred suppa from Shinano are
coming under orders to seize your stronghold. Be on high alert
day and night, [for] they can come anytime. It is important to have
watchmen at night, at dawn, and at the break of day. It is important
to prepare [and have] nine people in the watch at all times. At the
moment, it is the cold time of year, so shinobi would not approach,
unless it is a moonlit night. And nevertheless, rotate the watchmen
three times during the night, roll down the stones, throw torches
and observe [everything around you].22
The Hōjō consider the suppa as a hostile force, without specifying to whose
camp they belong. The object of their attack is a castle or a fortress, as seen from
the countermeasures suggested by Hōjō Ujikuni to prevent these shinobi from
scaling the walls. The number of warriors (500) suggests this was not a mere
assault of brigands, but a military operation.
The rappa and suppa were thus part-time commandos and part-time robbers
who tended to be engaged in long-term contracts with certain lords such as Takeda
or Hōjō. As a sub-category inside the larger social group of outcasts known as
akutō (evil bands), they resemble somewhat the privateers of early modern Europe,
as well as Chinese and Japanese coastal pirates who interchangeably worked for
the Chinese imperial state, foreign colonial powers, or various rebel groups.23
Shinobi skills were also ascribed to some rank-and-file warriors who served
in regular armies. Guards tasked with preventing night raids on castles were
sometimes called shinobi guards (shinobi-ban).24 In other cases, shinobi were merely
soldiers who volunteered, even once in a lifetime, for dangerous operations behind
enemy lines. In 1600, for example, during one of the small battles preceding the
decisive confrontation at Sekigahara, a petty commander named Eguchi Goenojō
was besieged in his small fort by a superior army. When the situation became
desperate, Goenojō decided to order a daring assault to demoralize, insult, or
perhaps frighten away the besieging army. Though this operation failed to prevent
the fall of his castle, it was immortalized in a military chronicle:

21. Morohoshi Michinao, “Ninja, Onmitsu to Hōgen,” Kokugakuin zasshi 97 (1996): 74

22. Sugiyama and Shimoyana, Sengoku ibun, Go-Hōjō-shi hen, “Letter by Hōjō Ujikuni,”
doc. 2431.
23. MacKay, “Pirate Nations,” 557–59.
24. Turnbull, Ninja, 42.

MILITARY HISTORY  1003

P. SErEBrIAKOvA & D. OrBACH

Now, there was one [warrior] renowned for his shinobi skills … and
on the same night he sneaked into the enemy’s camp, took one flag
from the camp of Naoe Kanetsugu, and another from the barracks
of Kurogane Sanzaemon, and hanged them on a high place at
the main entrance to the castle. Next morning, the men from the
assault force saw this and said: “Such a blunder! Not only did we
fail to occupy such a small castle, but also the flags were taken. That
is indeed a failure.”25
The volunteer who stole the commanders’ flags was not a full-time irregular
warrior. He was a rank-and-file samurai, “renowned” for his special sneaking
abilities. In 1570, two warriors who served the warlord Oda Nobunaga as mountain
guides and sneaked up to enemy fortifications to set them on fire were described
as using shinobi tactics.26 In a similar vein, another historical source lists rewards
bestowed on a middle-ranking warrior who used shinobi skills to seize a castle.27
In both cases, the word “shinobi” does not refer to a person, but to the tactical skills
of irregular warfare.
Kusa (literally: grass) was yet another specific shinobi activity, ascribed to
warriors who specialized in ambush tactics. The term probably referred to their
ability to blend in with the natural environment while fulfilling their tasks. On
Kyushu Island, these warriors were called fushikusa (literally, “lying in the grass,”
which is better translated here as “laying an ambush in the grass”).28 Yatsushiro
nikki, the chronicle of the Sagara clan from Kuma District in Kyushu, produced
the following description in 1557:
In a place called Naima Hatake Moroko in Kamitsu-ura Bay,
fushikusa [warriors] from Amakusa and Sumoto appeared. Both
[enemy] clans’ armies were defeated … In Kugino fushikusa were
discovered when the chestnut grass was cut. Two of them were
killed, two were taken hostage.29
Sengoku lords often incorporated kusa warriors into their armies. Takeda
Shingen, for example, as well as his rivals from the Hōjō and the Satake clans, used
specialized troops to collect tactical information in the field before battles. The Hōjō
deployed samurai on horseback who could ride, collect information, and disappear.
The various lords used kusa as both field intelligence and counter-intelligence troops.

25. Imamura, Sengoku Shiryō, 4:308.

26. Ota, Chronicle of Lord Nobunaga, 157. In the Japanese original, the verb is shinobi-iri
(to sneak and enter). See: Shinchō Kōki, vol.3, 10th Month, 19th Day, last accessed 12.10.2019,
http://www.page.sannet.ne.jp/gutoku2/sintyokouki_4.pdf
27. Sugiyama and Shimoyana, Sengoku ibun, Go-Hōjō-shi hen, “Signed letter of Hōjō Uji-
yasu,” doc. 750.
28. Aragaki Tsuneaki, “Sengoku kassen ni okeru machibuse senjutsu ni tsuite: shinobi to
kusa/kusa-chōgi-no jittai,” in Minegishi Sumio, ed., Nihon Chūsei shi no sai hakken (Tokyo: Yo-
shikawa Kōbunkan, 2003), 245.
29. Kumamoto chūseishi kenkyūkai, ed., Yatsushiro nikki (Kumamoto: Shinchōsha, 1980),
118–19.

1004  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

According to the Hōjō Godaiki, a chronicle of the Hōjō clan compiled in 1615, they
had to find ways to deal with kusa dispatched by the hostile Satake house:
When two armies confronted one other, horsemen, who were
always men of great merit, were sent to survey the area. The warrior
scouts rode to the frontline to study the circumstances of that day.
They crossed the border, climbed to a high spot, examined the
regimental colors of the enemy and swiftly rode back. When a
general went to war and the camps were being deployed in front
of one another, however, foot soldiers were sent to the frontline at
nightfall to hide in the grass. They spied the enemy and returned
[to camp] at daybreak. These [foot soldiers] were known as kusa or
shinobi. At times, these night-time kusa remained there until noon
[of the following day], unbeknownst to our warrior scouts. When
the latter crossed the front line, the kusa sprang up, blocked their
escape route, and killed them.30
In one of his letters, Hōjō Ujinao refers to kusa dispatched not only to intercept
enemy troops, but also to capture rival warriors as hostages, or for intelligence
purposes. This indirectly links them to intelligence gathering, which constitutes,
along with assassinations, the most iconic aspect of ninja nowadays:
Yesterday you came over, we had a meeting and I was very pleased,
especially [with the fact that] you used kusa [warriors] last night
and killed more than ten enemies and captured some alive and sent
them to Okushū (Hōjō Ujiteru). This was particularly good news.
Today I am returning to Odawara, so I will report the details.31
Aragaki Tsuneaki argues that the word kusa may refer not only to warriors,
but also to a corpus of tactical knowledge. The term kusa chōgi frequently appears
in the historical accounts of the Date clan from northeastern Japan. Masamune Ki,
a biography of Lord Date Masamune complied in 1642, has a detailed description
of these tactics:
So, in Mutsu Province battle strategy, there is [something] called kusa
chōgi, or entering the grass, hiding in the grass, plowing the grass, and
searching in the grass. Firstly, kusa chōgi [includes] sending [military
force] from our camp to sneak into other lands… So, depending on
how big or small this force is, people get divided into groups: Kusa
One, Kusa Two, Kusa Three … In Kusa One only infantry is sent up
to two to three chō [220–330 meters] ahead, and [they] infiltrate the
surroundings of the enemy’s castle at night, and that is called entering
the grass. From then on, they hide, waiting for the best opportunity
[to attack], and that is called hiding in the grass.
Then, [if ] one or more messengers emerge from [the castle] of
the enemy’s camp, Kusa One kills them, and that is called plowing

30. Imamura, Sengoku Shiryo 1:395.

31. Sugiyama and Shimoyana, Sengoku ibun, Go-Hōjō-shi hen, “Letter of Hōjō Ujinao,” doc.
2292.

MILITARY HISTORY  1005

P. SErEBrIAKOvA & D. OrBACH

the grass. Indeed, when members of the enemy’s camp learn about
[the ambush], all of them pull out their weapons and set to kill
Kusa One. So kusa warriors run away in different directions, which
gives an advantage [to our troops], [because] Kusa Two and Three
emerge and start competing against each other in defeating [as
many enemies as possible], which brings the ultimate victory.
Also, when [we] learn that [rival] kusa have infiltrated our territory,
a group of two to three warriors from our camp is formed, and this
force is sent away [to patrol the area]; Kusa Two and Three search
for the traces of their [rival] kusa, and then carry out their duty as
soldiers. Kusa One searches for anyone who might have survived
and when their [rival] kusa appears, Kusa Two and Three expect the
enemy to scatter. They shut [the enemy] out and attack [him]... No
matter how good their [rival] kusa are, [their positions] are taken
over from the rear, and they are killed if they attempt to escape.32
Kusa warriors thus were considered a part of the regular army of the Date
clan. They were mainly engaged in guerrilla operations and, at the same time,
executed counter-guerrilla and counterintelligence operations within their own
territory. In other words, kusa were warriors skilled in the tactics of kusa chōgi and
able to detect the movements of their rivals. There was a corpus of knowledge
behind the art of ambush, devised by people employed to do it on a regular basis.
All of the warriors and spies described thus far, samurai and commoners
alike, used their skills in irregular warfare in the service of conventional armies
in the Sengoku civil wars. In the same era, however, there was an interesting case
of rural samurai renowned precisely for their shinobi skills. Instead of serving
in the armies of lords, they used their art to safeguard their autonomy. Their
war against Oda Nobunaga, one of the most powerful warlords of the period,
became a cornerstone of later ninja lore.

Guerrilla Warfare with Shinobi Tactics: The Rise and Fall of Iga and Kōka
The neighboring regions of Iga and Kōka (today in Mie Prefecture) have a
special place in both ninja lore and the actual history of irregular warfare in late
Sengoku Japan. The tactics of the Iga Shū were documented in the esoteric Iga/
Kōga manuscript “Mansenshūkai,” written in 1676, as well as in the chronicle Iga
Kyūkō.33 Several retrospective, post-civil-war accounts from the Edo Period, such
as the Buyō Benryaku (quoted above) and Nochi Kagami, also associate the “ninja”
with these particular regions.34

32. Masamune Ki, vol.4, in Kobayashi Seiji, ed., Date Shiryō Shū, vol. 1 (Tokyo: Jinbutsu
Ōraisha, 1967), 213–14.
33. reproduced in Imamura Yoshio, ed., Nihon Budō Zenshū (Tokyo: Jinbutsu Ōraisha,
1966), 4:425–28, and Iga Kobunken Kankōkai, ed., Iga Kyūkō Iran Ki (Ueno-shi: Iga Kobunken
Kankōkai, 2010), 27–29, respectively.
34. Sugiwara Hiroshi, Sengoku Daimyo (Tokyo: Chūō Kōronsha, 1965), 205–6.

1006  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

According to Pierre Souyri, the tendency toward irregular warfare was born
out of the peculiar mountainous topography of Iga, surrounded as it was by “abrupt
summits, slopes covered with pine trees, narrow, winding paths [and] streams
turning into torrents in deep gorges.” The narrow pathways, called “tiger mouths,”
permitted the passage of only one rider at a time, and were easy to defend with
tactics of irregular warfare: ambushes and night assaults. Kōka was less difficult
to approach, but also partly protected by high mountains.35 Many inhabitants
were mountain woodsmen and hunters, who were used to blending with the
environment and waiting silently for their prey. As with early modern irregular
bands in Europe like the Hessian Jägers, the Iga and Kōka shinobi warfare tactics
were a natural development of their peacetime trade.36
The mountainous terrain, however, is not a sufficient explanation for the
development of irregular warfare skills in Iga and Kōka, or for the fact that such
warriors were organized into autonomous entities. There are many other regions,
such as Higo and Satsuma, with a similar topography throughout Japan, all replete
with hunters. Mountain warfare could also be used in the service of regular armies,
such as with the kusa or rappa in eastern Japan.37
The uniqueness of Iga and Kōga rather lies in a social factor: an unusually
high percentage of rural samurai (jizamurai), low-ranking warriors who held
managerial positions in estates with absentee landlords and lived in close proximity
to peasants, often in small, crude fortifications. Placed between lords and peasants,
rural samurai often had a complicated relationship with both. In certain places, they
formed the backbone of warlord armies. In others, they allied with the peasants
in provincial leagues (kuni ikki), a form of a “horizontal association governed by
written or unwritten codes,” designed to chase away brigands, warlords, and even
agents of the weak central government in Kyoto.38
In some areas, such as Yamashiro Province, also near Kyoto, tensions between
the peasants and the rural samurai broke the league after a few years. In Iga, however,
the geographic isolation of the area fostered cooperation between the upper and
lower echelons of rural society.39 As early as 1477, a priestly source mentions Iga

35. Pierre F. Souyri, “Autonomy and War in the Sixteenth Century Iga region and the
Birth of the Ninja Phenomena,” in John A. Ferejohn and Frances McCall rosenbluth, eds.,
War and State Building in Medieval Japan (Stanford, Calif.: Stanford University Press, 2010),
110–18, 20–22. For a modern historical analysis, see Murakami Tadashi, “Ninja to Onmitsu,” in
Yoshimoto Shinji, ed., Edo Jidai Bushi no Seikatsu, (Tokyo: Yūzankaku, 1966), 237.
36. Souyri, “Autonomy and War,” 111. See also the editors’ introduction to the
“Mansenshūkai,” 405. On the Hessian Jägers, see Christopher Duffy, The Military Experience in
the Age of Reason (New York: routledge, 2016), 272.
37. Turnbull, Ninja, 47.
38. Berry, Culture of Civil War, 39–40; Turnbull, Ninja, 45.
39. Pierre F. Souyri, The World Turned Upside Down: Medieval Japanese Society, trans. Käthe
roth (New York: Columbia University Press, 2001), 189–91, 215; Souyri, “Autonomy and War,”
118–19.

MILITARY HISTORY  1007

P. SErEBrIAKOvA & D. OrBACH

as a province where the locals do not obey the military governors appointed by the
shogun (in fact, warlords who manipulated shogunal appointments).40 In 1560,
a council met in a local temple and formally declared a new league, allied with a
sister organization in neighboring Kōka. The charter of the Iga League declared
its main purpose was to repel invasions by neighboring lords.41 “While everyone
else obeyed their military governors (shugo), the men of Iga and Kōka were free
and did as they pleased,” wrote the authors of “Mansenshūkai” in the post-civil-
war period, clearly lamenting their freedom lost in bygone times.42
Such conditions were ideal for the development of bands with strong horizontal
ties, comprised of peasants and rural samurai, yet led and trained by the latter.
rural samurai leadership was strong enough to ensure cohesion and yet allowed
social mobility. According to Murakami, there was still a class division between
“high” and “low” shinobi, and only the former were employed in truly prestigious
missions. Souyri argues that “although differences in social rank between peasants
and warriors were clear and well established, they were not insurmountable, for
the Iga commune also promoted heroic fighters, regardless of their background.”43
According to the Iga Charter, peasants took full part in military operations, and
distinguished fighters could be promoted to samurai status. Discipline was harsh,
and the league threatened informers and traitors with confiscation of all property,
followed by public decapitation.44
Part of the training in this militarized society was in tactics of irregular
warfare. According to the Iran Ki, a chronicle compiled by a local scholar in the
seventeenth century:
According to the customs of the land, peasants and servants had
to wake up every morning at the hour of the ox [1-3 a.m.], being
allowed to tend their household affairs only at the hour of the
tiger [3-5 a.m.]. After the hour of the tiger, they merrily went to
the Nikeji temple for military training, learning and practicing
especially the art of pity [probably a code name for shinobi arts].
The same chronicle also stipulates, with some exaggeration, that “they have super-
natural powers as spies, so no matter how secure a fortress is, they can enter it secretly,
and in other provinces too the Iga shinobi have been found to be useful.”45
As implied in the Iran Ki, Iga and Kōka rose to prominence not only because
they were able to defend themselves successfully, but because many of their
inhabitants used shinobi tactics to participate in wars outside their borders. Hence,

40. Berry, Culture of Civil War, 35, 39–43.

41. Souyri, World Turned Upside Down, 186, 189–91.
42. Yoshio, Nihon Budō Zenshū, 427.
43. Turnbull, Ninja, 44-45; Murakami, “Ninja to Onmitsu,” 238; Souyri, World Turned Up-
side Down, 119.
44. Souyri, “Autonomy and War,” 191.
45. Iga Kyūkō Iran Ki, 112–13. The quote beginning with “they have” was translated by
Turnbull (Ninja, p.63) from an edition of the manuscript.

1008  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

Hino Kumawaka Maru

training to avenge his
father by swinging across
a stream on a bamboo pole
[by Utagawa Kuniyoshi,
circa 1842]

these fighting techniques, which were originally developed for self-protection,

gradually become a local specialty.46
The earliest reliable account of irregular warfare in Iga and Kōka is given in a
priest’s diary, Tamon-in Nikki. This record demonstrates that warriors from these
two regions fought with shinobi tactics as early as 1541:
This morning, the Iga units entered Kasagi Castle in secret (shinobi
itte) and set fire to a few of the priests’ quarters and so on. They also
burned down outbuildings in various places within the third bailey
and are even said to have seized the first bailey and the second
bailey … According to Ukon, the castle commander … they came
from Kōka in Ōmi and numbered between 70 and 80 men.47
In 1556 and 1560, another diarist documented similar castle raids launched
by Iga men in Takada and Tōichi.48 In 1580, Iga men assisted Kongōbuji Temple,
the head temple of Mt. Koya’s Shingon Buddhist Sect, in its plans to expand its
territory into the lands of a neighboring proprietor:

46. Iga Kyūkō Iran Ki, 112.

47. Eishun, Tamon in Nikki, Tsuji Zennosuke, ed. (Tokyo: Kadokawa Shoten, 1967), 256.
English translation in Turnbull, Ninja, p. 40.
48. Turnbull, Ninja, 41.

MILITARY HISTORY  1009

P. SErEBrIAKOvA & D. OrBACH

Iga units sneaked at night into the castle … [They entered] from
the south [by] getting over the water moat. On multiple occasions
[you] were the first people on the frontline, [you] fought beyond
comparison inside the castle, a few people witnessed that.49
Judging from the context of this letter, the addressee could be a representative
of one of the Iga units, who demonstrated excellent skills in besieging the castle.
Kongōbuji Temple was included in the self-governed monastic confederation of
Mt. Kōya, which remained an independent powerful military force up until 1585.
The document does not include any indication of payment or reward for the Iga
men, which suggests there might have been an alliance between the Iga groups
and Mt. Kōya. In other words, the Iga men likely did not operate as mercenaries,
but as a political faction assisting its allies.
It is also possible to interpret that letter as a statement of authority though.
In the Sengoku civil wars, military letters of acknowledgment often implied that
the recipient was subordinate to the sender. The sender in this case was a temple,
however, and not a warlord with pretentions to the realm, and there is no proof of
vassalage between Mt. Kōya and the Iga groups other than this letter.
Many of the sources on the special operations of Iga men in the sixteenth
century were written a long time after the events, and were therefore subject to
embellishments and the distorting influence of later myths. Most contemporary
sources were destroyed when the warlord Oda Nobunaga invaded Iga in 1581, so it
is difficult to verify the details of most operations.50 Broadly speaking, the available
sources indicate that the Iga men specialized in nocturnal, irregular warfare, and
would penetrate castles secretly to gather information or to set them on fire. That
was a well-known shinobi tactic, also used by soldiers of regular armies.51 It is clear,
moreover, that the Iga men worked independently, without allegiance to lords,
but the sources do not prove that they were mercenaries, as stipulated in later
ninja lore. The sources suggest that all shinobi operations were located in nearby
provinces, where the Iga men had a direct stake. As in 1580, they fought on behalf
of allies in the complicated military environment of the Sengoku civil wars.52
The Kōka men apparently also used shinobi warfare to assist their allies. A
biography of Tokugawa Ieyasu, Mikawa Go Fūdo Ki, based on sources written
around 1640, argues that Ieyasu used the good services of the Kōka shinobi in his

49. “Inukaike Monjo,” in Kōyasan shi hensanjo, ed., Kankō Kaihon Kōyasan Monjo (Kyoto:
Kōyasan monjo kankō kai, 1939), doc. 9137.
50. Turnbull, Ninja, 38-44
51. Ota, Chronicle of Lord Nobunaga, 157.
52. Beyond the five pre-1581 sources Turnbull cites (Turnbull, Ninja, 38–43), there are
several sources on Iga shinobi tactics, but they were written decades after the events. See, for ex-
ample, the following chronicles, written in the seventeenth century: “Asai Sandai Ki,” in Kondō
Heijō, et al., eds., Bukebu. Nendaiki hen. Dai 3 (Kyoto: rinsen Shoten, 1967), 8:297–99; Oze
Hōan, Taikoki, ed. Emoto Hiroshi (Tokyo: Iwanami Shoten, 1996), 54, 90, 368; and the less reli-
able “Mansenshūkai,” Yoshio, Nihon Budō Zenshū, 423–28.

1010  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

assault on Lord Imagawa’s Kaminojō castle in 1552. In an operation reminiscent

of the tactic described in Tamon in Nikki, they entered the castle silently by night,
used fire and other distraction tactics to create havoc, killed most of the defenders,
and captured the commander’s children alive, enabling Ieyasu to exchange them
for hostages from his own family. The contractual relationship between Tokugawa
and the Kōka groups is clear from a passage in the Mikawa Go Fūdo Ki:
The commander Mitsuhara Sanzaemon advised [Iyesu]: “As this
castle is protected by a steep slope, if we rush head on to attack it
our allies will suffer grave loses. Fortunately, among our retainers
there are some people related to the Kōka region in Ōmi Province.
Invite the Kōka men through their kin and let them sneak into
the castle.”53
There were two subsequent assaults on this castle, and Kōga shinobi were employed
in both of them. This demonstrates that the Kōka groups, like their allies from Iga,
were not bound to serve Tokugawa Ieyasu as retainers. They did not even serve him
for a long period, in contrast to the rappa of eastern Japan.
Another example of a lord reaching out to a Kōka group is found amongst the
letters of Nobunaga, later known as the first unifier of Japan. In 1568, Nobunaga
was about to make one of the most crucial and dangerous strategic moves in his
career, marching on the ancient capital of Kyoto. Success would give him control
over the prestigious institutions of the shogun and the emperor. Such a move,
however, would expose Nobunaga’s home territory to incursions by rival warlords.
The march to Kyoto, eventually an astounding success, was made possible only by a
web of alliances that safeguarded Nobunaga’s rear.54 As part of these negotiations,
Nobunaga invited the warriors of Kōka to join him and to pledge allegiance:
Since passage through Ōmi Province is difficult … [our army]
shall set out for your land. In accordance with previous request, you
shall demonstrate your friendship to Nobunaga, and in that time
all of you shall pledge loyalty, and [your service] will be greatly
appreciated. For that purpose [you shall] send [Wada] Koremasa
and [Ōkusa] Kimihiro. Furthermore, three [representatives] shall
submit their answer [to this request].55
Nobunaga addresses the league of Kōka as a collective of low-class warriors,
without any ranks or titles, much as he would address a self-governed village or an
independent monastic confederation. This request is repeated, moreover, to ensure

53. Kuwata Tadachika, et al., eds., Kaisei Mikawa go Fūdo Ki (Tokyo: Akita Shoten, 1976–
77), 1:244, 262.
54. Stephen Turnbull, The Samurai: A Military History (New York: MacMillan, 1977), 131–
34; Berry, Hideyoshi, 37–40; Lee Butler, “The Sixteenth Century Unification,” in Karl Friday, et
al., eds., Japan Emerging (Boulder, Colo.: Westview Press, 2012), 312–13.
55. Okuno Takahiro, ed., Oda Nobunaga monjo no kenkyū, vol. 1 (Tokyo: Yoshikawa
Kōbunkan, 1988), doc. 93. For the context of the Ōmi campaign see also Ota, Chronicle of Lord
Nobunaga, 120–24.

MILITARY HISTORY  1011

P. SErEBrIAKOvA & D. OrBACH

that Kōka samurai are not going to betray Nobunaga in favor of his enemies from
the rokkaku (Sasaki) clan.
Nevertheless, the Kōka men did not remain loyal to Oda Nobunaga. In 1570,
they operated against him in alliance with Iga and the rokkaku clan, and again in
1573, this time allied with Iga and the Asai clan.56 This shows that shinobi from
Kōka and Iga maneuvered between rival warlords, and shifted their allegiance as it
suited them. Always restive, they remained a symbol of chaos in a world that became
gradually more organized. To unify the realm, Nobunaga had to neutralize such
divisive forces. In 1571, he ordered the massacre of the Buddhist clergy of Mount
Hiei, for instance, and subsequently began subduing the leagues of Iga and Kōka.57
At first Nobunaga practiced moderation. As in other civil wars, actors were
often quick to join the power that exercised the most effective control over a given
territory, regardless of their initial sentiments and loyalties.58 In 1574, after the
fall of their allies from the rokkaku clan, Kōka was prudent to surrender without
much of a fight, and thus was ordered only to demolish its castles and dispatch
soldiers to Nobunaga’s armies.59 The men of Kōka subsequently are identified
among the troops stationed in Nobunaga’s stronghold, Azuchi castle. During the
New Year celebrations in 1582, for instance,
the horse guards, Kōka warriors, and others had in the meantime
assembled on the white gravel. After keeping them waiting there for
a while, Nobunaga told them, ‘You must all be freezing. Go up to the
South residence and have a look around the Kōunji Palace.60
Nobunaga thus allowed the Kōka men to observe the most private and luxurious
parts of his castle (the southern palace), which might be regarded as a sign of the
complete submission of the once-independent league of Kōka.
Iga was a tougher nut to crack, as throughout the 1570s it acquired a
reputation as an unrelenting enemy of Nobunaga. The league fought him twice,
in 1570 and 1573, allied with some of his enemies and gave sanctuary to others.
Unlike Kōka, which controlled strategic highways, Iga was remote enough to be
ignored as long as Nobunaga had more dangerous enemies to handle.61 The Iran
Ki notes that in 1579 though, a disaffected rural samurai from Iga deserted into
the camp of Nobunaga’s son, Nobukatsu, in Ise, complaining that the residents
of his region were “extravagantly perverting the heavenly principles” by driving
out their rightful lords. Nobukatsu promptly decided to invade the province,

56. Ota, Chronicle of Lord Nobunaga, 145, 201.

57. A contemporary account of the destruction of Mt. Hiei is provided in the Chronicle
of Lord Nobunaga, 164–66, as well as ryūsaku Tsunoda, et al., eds., Sources of Japanese Tradition
(New York: Columbia University Press, 1964), 305–8.
58. Kalyvas, The Logic of Violence, 132, 145.
59. Turnbull, Ninja, 59–60.
60. Ota, Chronicle of Lord Nobunaga, 421–22.
61. Turnbull, Ninja, 51, 65.

1012  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

probably without coordinating this operation with his father, who was busy on
more important military operations elsewhere.62
Nobukatsu made one of the most foolish mistakes possible when fighting
irregular warriors, however. He ordered one of his vassals to march into the
heart of Iga and rebuild a ruined castle as a base for operations. The Iga men,
well informed by spies, used their favorite tactics and dispersed the invading
force with surprise attacks, driving it into the wooded ravines and out of their
territory with heavy losses.
In response, Nobukatsu staged a grand, three-pronged invasion through well-
known mountain passes. Thanks to his indiscretion and Iga spies (called shinobi-
mono in the Iran Ki), his route was well-known to the defenders. The Iga shinobi
ambushed his forces, scattering the soldiers into rocky trails, deep valleys, and muddy
rice fields. Thousands of soldiers, including a senior general, were surrounded and
pierced with spears. Others killed each other in confusion, or committed suicide.63
Nobunaga was incensed and admonished his son with harsh words:
The other day you were guilty of a fiasco on the Iga border. Take
this as a lesson: The Way of Heaven is terrible indeed, and the sun
and moon have not yet fallen to the earth … you’re immature, so
you’re gullible. That’s what got you into this mess. How mortifying!
How utterly mortifying! ... If your composition is really like this, I
shall not permit the parental tie between us to continue.64
Nobunaga was angry not only because his senior retainers had died, but also
because his son wasted time and effort on a minor skirmish instead of joining him
in more important theaters.65
After Nobunaga defeated his other enemies in quick succession, Iga’s turn
finally came. The local warriors unwisely attracted his attention by attacking one
of his vassals in 1580, in league with the Mt. Kōya monastic confederation.66 The
Iran Ki, written from the point of view of the Iga men, elaborates on this point,
quoting Nobunaga:
The Iga rabble’s extravagance is growing day and night so as to try
our patience. They do not have any concern for high and low and are
spreading lawlessness across the land. Their behavior is a mystery
to me, as they despise the high, do not fear men of rank, behave
rebelliously, and are disobeying the laws of the land. Therefore, I
intend to swiftly hunt and punish them.67

62. Iga Kyūkō Iran Ki, 124; Chronicle of Lord Nobunaga, 329–30.
63. Iga Kyūkō Iran Ki, 126–36; Ota, Chronicle of Lord Nobunaga, 329–30. See also Turnbull,
Ninja, 66.
64. Ota, Chronicle of Lord Nobunaga, 330.
65. Ota, Chronicle of Lord Nobunaga, 330.
66. Turnbull, Ninja, 41.
67. The author of the Iran Ki uses the word “extravagance” (ogori) twice, once in the com-
plaint of the Iga traitor (p. 124), and once in Nobunaga’s speech (p. 152), to show how Nobunaga

MILITARY HISTORY  1013

P. SErEBrIAKOvA & D. OrBACH

It would be tempting, but misleading, to interpret Nobunaga’s sentiment

as a distaste felt by a regular army commander towards partisan and guerrilla
fighters. Nobunaga’s irritation with Iga had little to do with their irregular tactics
and more with their foreign policy and political structure. The Sengoku civil wars
were comprised of multiple private quarrels between strongmen in which “saving
face” and avenging insults were crucial building blocks of political capital. They
were especially important for strong warlords such as Nobunaga, who used their
carefully constructed image to evoke awe among their rivals.
When Nobunaga grew strong enough, he began to transcend the language
of private conflict and revive an older, pre-Sengoku narrative of public authority,
order, and discipline.68 The Iga men, both independent and close to the Kyoto
center of political power, repeatedly insulted him and defied his authority. They thus
became suitable targets for repression according to both the discourse of private
quarrels (saving face) and the rearticulated discourse of public authority (upholding
discipline). It was precisely the political structure of Iga as a self-governed league of
rural samurai, however, that incentivized irregular warfare in the first place.
When Oda Nobukatsu invaded Iga again in 1581, he made use of two shinobi
traitors who served as mountain guides, as well as a group of shinobi from Kōka.
Control of Kōka was crucial, because it was the most convenient route to Iga,
the others being blocked by high mountains. Nobukatsu invaded the region
simultaneously from six directions, including from Kōka, with an army 40,000-
60,000 strong. Each of his contingents was equal in strength to the army defeated
by Iga in 1579.69
The Iga men tried yet again to use shinobi tactics to deter the enemy. In a
much-celebrated attack, Iga shinobi-mono (here the term is used as a noun) raided
Kōka, then in Nobunaga’s territory, to abduct samurai, merchants, and farmers
for ransom. Shinobi-mono are mentioned here as night scouts and assault troops,
distinguished warriors hand-picked from the ranks. They were ordered to employ
great discretion, and not to attack certain families of their former allies, still known
to harbor pro-Iga and anti-Nobunaga sentiments. Presumably, such people could
serve as precious sources of human intelligence.70

was influenced by the informer. Thus the author creates a direct causal link between internal
treason and foreign invasion. Elsewhere in the Iran Ki, both Oda Nobukatsu and his father
Nobunaga use the term kyōto (教徒), translated to English as “thugs” or “rabble,” when referring
to the Iga Shū. See Iga Kyūkō Iran Ki, 130. See also p. 207.
68. On the role of revenge, personal injury and reprisals in the Sengoku civil wars, see
Berry, The Culture of Civil War, 7–8, 32–33, 37, 48. On the narrative of centralized authority, see
Nobunaga speech in 1571, quoted in Tsunoda, Sources of Japanese Tradition, 1:305–6; Butler, “The
Sixteenth Century Unification,” 319.
69. Ota, Chronicle of Lord Nobunaga, 410–13; Iga Kyūkō Iran Ki, 146, 152–53; Souyri, “Au-
tonomy and War,” 118; Turnbull, Ninja, 37; Iga Kyūkō Iran Ki, 152–53.
70. Turnbull, Ninja, 42–43.

1014  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

Oda Nobunaga [Portrait

by Kanō Sōshū, circa
1590]

Such daring attacks, however, could not stop Nobukatsu’s advance. Once the
invading troops were deep inside Iga, the panicked locals did not have enough
manpower to defend all fronts. They could hide in the mountains, as some probably
did, but then they would leave their homes and families to the conqueror’s fire and
sword. Most tried to barricade themselves in wooden strongholds, which were
easily smashed by enemy artillery and fell one after the other. Nobukatsu ordered
the burning of every temple and village he encountered, and his forces massacred
many of the inhabitants. This violence was not indiscriminate, but targeted well-
known samurai households.71
Often, in civil wars, both conventional armies and irregulars have difficulty
identifying friend and foe.72 In Iga, the solution was relatively easy. Nobunaga
and Nobukatsu identified the rural samurai households as the root of the league’s
troublesome political structure and, therefore, slated them for annihilation.
According to Souyri, some shinobi-mono remained in the mountains to engage
in guerrilla warfare, but they were not able to seriously challenge Nobunaga’s
rule. One old edition of the Iran Ki even tells of three Iga shinobi who tried to
assassinate Nobunaga using small cannon, but were only able to kill “seven or
eight” of his companions. According to Turnbull, however, this story seems out of
place, and probably grew in the telling.73

71. Souyri, “Autonomy and War,” 118; Ota, Chronicle of Lord Nobunaga, 412; Mikawa Go
Fūdo Ki, 2:191–92; Iga Kyūkō Iran Ki, 154–55, 59–62, 86–90, 205–6; Turnbull, Ninja, 67–68.
72. Kalyvas, The Logic of Violence, 89–90.
73. Turnbull, Ninja, p.71. This story does not appear in the new version, and probably
the two editions are based on different manuscripts. For the assassination story, see: Momochi
Orinosuke, ed., Kōsei Iran Ki (Ueno, Mie: Tekisui Shoin, 1897), vol 4, p.5.

MILITARY HISTORY  1015

P. SErEBrIAKOvA & D. OrBACH

The lands of Iga were divided between Nobunaga’s generals, who finally
pardoned some of the remaining fighters.74 Gone was the league of Iga, with its
relatively egalitarian regime. Many survivors were offered shelter in neighboring
provinces controlled by Tokugawa Ieyasu. The Mikawa Go Fūdo Ki says that some
remaining Iga warriors (possibly the people who were engaged in guerrilla warfare
against Nobunaga) repaid Ieyasu for his kindness. In 1582, only one year after the
conquest of Iga, Oda Nobunaga was assassinated by a traitorous retainer. Tokugawa
Ieyasu, who had to reach safe territory quickly in order to assemble his troops, was
helped by local shinobi in Iga to cross the area safely on the way to his province.75
By then, the phenomenon of independent irregular warriors was waning.
Warriors skilled in shinobi tactics, some of them from Iga and Kōka, were integrated
into regular armies as special shock troops, and were active during the campaigns of
Nobunaga’s successor, Toyotomi Hideyoshi, in Kyushu and Korea, as well as in the
Battle of Sekigahara (1600) that ended the Sengoku civil wars.76 In 1603, Tokugawa
Ieyasu was nominated as shogun, and the Edo Period began. As a final prize for
long years of cooperation, the shogun appointed 300 shinobi from Iga and Kōka to
guard the castle in his new capital, Edo. The shogunate was still insecure, and the
employment of seasoned shinobi as palace guards could be effective against potential
assassins and spies. Their employment as a hereditary status group continued until
1740, when they were dismissed by Shogun Tokugawa Yoshimune.

Consolidation: The Role of Irregular Warfare in the Sengoku Civil Wars

After evaluating all available evidence, it seems that irregular warfare, in all of
its forms, occupied at best a secondary role in the Sengoku civil wars. These wars,
especially in their later stage, were decided with large-scale conventional battles
such as Nagashino (1575) and Sekigahara (1600), pitting tens of thousands of
troops against one another. No reliable, contemporary evidence is known that
shows shinobi warfare decided an engagement on anything beyond the tactical
level. Privateers and organized brigands such as the rappa and the suppa played
a role in the endless wars between the Hōjō, Takeda, and Satake clans in
eastern Japan, but could not decide them. Kusa had a supporting role, at best, in
conventional engagements. The famous political assassinations ascribed to shinobi
(or “ninja”) are later inventions, typically from the modern era. Intelligence was
surely indispensable in the Sengoku civil wars, as in all wars, but this particular
shinobi activity remains an important subject for further research.
The only groups associated with shinobi activity that played an independent
political role were the leagues of Iga and Kōka. The former’s attempt to defy Oda

74. Turnbull, Ninja, p.71; Murakami, “Ninja to Onmitsu,” 239; Iran Ki, 206–7; Ota, Chron-
icle of Lord Nobunaga, p.413.
75. Mikawa Go Fudo Ki, 2:191–92. This story is corroborated by the Kawasumi Taikō Ki,
one of the Toyotomi Hideyoshi chronicles; see: Kuwata Tadachika, Nakamura Kōya. eds., Taikō
Shiryō Shū (Tokyo: Jinbutsu Ōraisha, 1965), 230–31.
76. Taikoki, 368; Turnbull, Ninja, 74.

1016  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

Nobunaga is an interesting test case for the merits, limits, and wider context of
irregular warfare in a chaotic, but gradually consolidating society.
Comparative studies of guerrilla warfare show that to survive the onslaught
of a bigger, conventional army, irregular warriors should ideally play for time, delay
decision, and wear down the invader until the costs of the campaign outweigh any
potential benefits. To reach such a goal, the guerrilla needs to avoid outright defeat,
often through blending into a sympathetic population, obtaining unreachable
sanctuaries (either in difficult terrain or across a border), or winning support from
external powers. In the end, the invader’s political interest in victory will diminish
enough for the campaign to stop. That holds true especially in modern times, when
nationalism might garner sympathy for the guerrilla among the local population,
and mass media and electoral politics may sap the invader’s will to sacrifice men
and resources. Domestic opposition, from within either the ruling elite or the
general public, will likely rise with the costs of the conflict. 77
In the Sengoku era, a pre-modern complex of simultaneous civil wars,
conditions were quite different. Electoral politics and mass media did not exist, so
insurgents could not generate the political drama that helps wear down invaders
in the modern era. No matter what the men of Iga did, Nobunaga did not have to
contend with domestic political pressure to abort the campaign. Even when his
generals resisted orders to massacre the monks of Mount Hiei, for fear of sacrilege,
he was able to enforce his will. That was certainly true in a place like Iga that
carried no religious significance.78
In other respects, however, there were some similarities between the Iga affair
and modern guerrilla wars. As a succession of civil wars between numerous actors,
the Sengoku conflict incentivized irregular warfare.79 As long as multiple warlords
contested the realm, the Iga and Kōka leagues could shift alliances and use sanctuaries
across borders, thus enjoying the support of anti-Nobunaga warlords such as the
rokkaku. Guerrilla warfare could sap the invader’s political will to proceed, but in a
different way than in the modern world. In the Sengoku civil wars, it was dangerous
to dispatch an army for a prolonged military campaign, as rival warlords could use
the opportunity to invade the army’s home territory.80 Guerrilla tactics that delayed
an invading army thus might convince the invader to stop prematurely, to protect his

77. Smith, “Guerrillas,” 36; Andrew Mack, “Why Big Nations lose Small Wars: The Politics
of Asymmetric Conflict,” World Politics 27, no.2 ( January 1975): 179, 195–96, 200; Yagil Hen-
kin, “How Great Nations Can Win Small Wars,” Azure 66 (Spring 2006): 39–41, 57–60; Ivan
Arreguin-Toft, “How the Weak Win Wars: A Theory of Asymmetric Conflict,” International
Security 26, no.1 (Summer 2001):103–06, 122; Max Boot, Invisible Armies: An Epic History of
Guerrilla Warfare from Ancient Times to the Present (New York: Liveright, 2013), 559–66.
78. For the destruction of Mount Hiei, see Nobunaga’s debate with his generals (1571)
reproduced in Tsunoda, Sources of Japanese Tradition, 1:305–8.
79. Kalyvas, The Logic of Violence, 83.
80. Stephen Turnbull, The Samurai—A Military History (New York: MacMillan, 1977),
131.

MILITARY HISTORY  1017

P. SErEBrIAKOvA & D. OrBACH

rear from attack. Such conditions allowed independent entities such as the leagues
of Iga and Kōka to survive well into the 1570s.
This room for maneuver gradually closed in the mid-sixteenth century,
as Japan underwent the long transition from the Sengoku civil wars into the
more stable Edo Period. As the century came to a close, irregular warfare and
its associated breaching of status boundaries, became an anomaly. To paraphrase
James Scott and Joseph MacKay, state building expanded like a flood, overrunning
independent and semi-independent regions.81 Organizational, economic, and
military developments empowered a small number of stronger lords, who formed
increasingly larger regional coalitions.82
As the number of feuding warlords diminished, and especially following
Nobunaga’s seizure of Kyoto in 1568, Japan was consolidated by Nobunaga and
his two successors, Toyotomi Hideyoshi and Tokugawa Ieyasu. These warlords
strengthened their control over an increasing territorial expanse and, as in other
civil wars, encouraged weaker actors to join them as vassals, which further tightened
their control.83
In this process of consolidation and state building, one of the main tools
at the unifiers’ disposal was the reinforcement of status boundaries. Nobunaga
eliminated independent forces such as the armed monasteries, or the Iga and Kōka
leagues. The Kōka men understood for whom the bells tolled and joined Nobunaga
without much resistance. Being in possession of Kōka, he could now strike Iga
with impunity from six directions. As the number of his enemies declined, he no
longer had to fear a third party attacking his home territory from the rear.
In such conditions, irregular warfare had little chance, especially when most
Iga men, eager to defend their homes and villages, could not practice it to the full
and fortified themselves in strongholds instead of taking to the mountains. Iga
thus fought Nobunaga on his terms, pushed to do so by Nobunaga’s campaign of
extermination. Had the invader merely occupied the main settlements of Iga, the local
shinobi could have hidden in the mountains and worn him down with hit-and-run
tactics. But as Nobunaga’s policy was the utter annihilation of samurai households,
the warriors were forced to fight and, ultimately, perish with their families. To quote
Ivan Arreguin-Toft, “when invading or occupying forces do not exercise restraint in
the use of force, or when their purpose is the destruction of a weak actor’s people,
[guerrilla warfare] can become a prohibitively expensive defensive strategy.”84

81. Scott, The Art of Not Being Governed, 22; MacKay, “Pirate Nations,” 554.
82. Thomas Conlan, “Instruments of Change: Organizational Technology and the Consoli-
dation of regional Power in Japan, 1333–1600,” in War and State Building, 124–51; Butler, “The
Sixteenth Century Unification,” 311–20.
83. Kalyvas, The Logic of Violence, 132, 145.
84. Arreguin-Toft, “How the Weak Win Wars,” 104, 109–10. See also George H. Quester,
“The Guerrilla Problem in retrospect,” Military Affairs 39, no.4 (December 1975), 192; and
Angelo M. Codevilla, Between the Alps and a Hard Place: Switzerland in WWII and the Rewriting
of History (New York: regnery, 2000), 64–68.

1018  THE JOURNAL OF

 A Historical Understanding of the “Ninja”

The consolidation of the country continued after Nobunaga’s assassination

under the leadership of Toyotomi Hideyoshi. This self-made ruler of peasant stock
confiscated all weapons from the peasantry, tied the farmers to the land, and stymied
opportunities for social mobility.85 Tokugawa Ieyasu, Hideyoshi’s successor and
the final unifier of Japan, enforced an elaborate system of hereditary status groups,
designed to keep everyone in his place, in a harmonious hierarchy frozen in time.

Shinobi and Ninja: From Real to Imagined Irregular Warriors

With the establishment of the Tokugawa Shogunate came relative peace,
which rendered redundant not only the Iga and Kōka men, but also irregular troops
like kusa and rappa. Having ceased to exist as self-governed territorial entities, the
independent shinobi warriors from Iga and Kōka devolved into static castle guards
in Edo, the center of the shogunate’s political power.
The demise of the historical shinobi, disparate social groups who worked for
different organizations and held different identities, and whose only common
denominator was the utilization of irregular tactics, gave way to a new, invented
tradition. In the peaceful Edo Period (1603-1868), under the relatively stable
order of the Tokugawa Shogunate and its affiliated domains, the authorities used
extensive networks of internal espionage to safeguard stability, collect intelligence
on the numerous lords, and nip any plot in the bud. Among other terms, such spies
were sometimes called in official documents “shinobi mono,” “Iga men,” or “Kōka
men,” even if they did not come from these particular regions. In the Edo Period,
these terms denoted a group that shared a specific, if loose, professional identity.
They were relatively few and far between, however, and after the first few decades
of the era, other terms for spies appeared more frequently in the sources.86
At the same time, authors, artists, and martial artists reimagined irregular
warriors from the Sengoku era as a unified professional group: the shinobi-mono
(“ninja” in twentieth-century pronunciation). The military art of the shinobi was
even studied in schools, presented as an ancient tradition. Military guidebooks for
shinobi warfare began appearing in increasing numbers. Some of them reinterpreted
local traditions from Iga and Kōka. Others were more fanciful.
Nonetheless, as Stephen Turnbull writes, many mainstays of contemporary
“ninja” lore were still absent. The figures in the guidebooks, for example, threw
no shuriken and were not dressed in black; there were no descriptions of sword
fighting, assassinations, or face-to-face battles. One of the manuals, from 1681,
did recommend wearing dark clothes for concealment, but the famous black
robes and hoods of the ninja probably first appeared in the Kabuki Theatre
and on woodblock prints. Most manuals focused on intelligence collection and
combined classical Chinese aphorisms, insights on espionage, descriptions of
equipment, and frequent references to an unwritten lore. Crucially, some of them
amalgamated different groups of irregular warriors into a single professional

85. Butler, “The Sixteenth Century Unification,” 315–16.

86. Turnbull, Ninja, 87–89, 92–95.

MILITARY HISTORY  1019

P. SErEBrIAKOvA & D. OrBACH

category. In Gunpō Jiyōshū, a manual from 1653, the author refers to Iga and
Kōka men in the chapter titles, and then instructs the readers to pronounce even
the ancient term settō (thieves) as shinobi-mono.87
Stathis N. Kalyvas has argued that “the fact that civil wars are state-building
processes means that their ‘master narratives’ are likely to be contaminated by the
war’s outcome: they will be distorted, and their ambiguities and contradictions
will be erased.”88 So it was with the irregular warfare of post-Sengoku Japan. In
the relatively static world of the Edo Period, it made sense to imagine the shinobi-
mono as a secret order with a frozen identity. While in the earlier chronicles they
are described as related to specific regions (such as Iga and Kōka), chronicles from
the nineteenth century imagine them as floating in space and time, from ancient
China through the Sengoku civil wars, and into the present. The gradual invention
of this tradition formed the basis for the ninja lore of the twentieth century.

87. Turnbull, Ninja, 99–103.

88. Kalyvas, The Logic of Violence, 389–90.

1020  THE JOURNAL OF

Copyright of Journal of Military History is the property of Society for Military History and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.