Udo Shugendō and the Aisu Kage-ryū

An excerpt from my book Shugendō and Heihō (Mountain, Sword, and Mind Volume I).

Aisu Ikōsai Hisatada (愛洲移香斎久忠) founded Kage-ryū (陰流) after receiving a transmission from a deity that took the form of a spider (蜘蛛; kumo) — in some lines a monkey (猿; saru), hence Saru-tobi (猿飛) becoming Enpi (燕飛) — in Hyūga (日向). The earliest records frame the event as a numinous dream rather than a specific animal: the Honchō Bugei Shōden (本朝武芸小伝) gives only reimu (霊夢; numinous dream), and the animal-form tellings are later elaborations.

The serious scholarship on him (Nakaseko Yoshimichi 中世古祥道) (Nakaseko Yoshimichi 1999) argues that Ikōsai was likely a shugenja (修験者) from the Gokasho (五ヶ所) Aisu family, and that the site of the revelation, Udo, was a yamabushi (山伏) holy ground called “the Kōya of the West” (西の高野; Nishi no Kōya), ranked beside Kumano (熊野).

The tengu/animal-spirit naming in the entire Shinkage stream — Enpi, Tengushō (天狗抄) — descends from a Kage-ryū origin that is itself plausibly descended from medieval Shugendō. The Hikita (疋田; Higo) branch preserves the document form: its chū-gokui (中極意) level is called Tengu-gaki (天狗書), comprising the kata Ranshō, Chōkyoku and Unsai teachings.

The vision itself is easily miscast. Aisu’s revelation was not the crow-billed karasu-tengu of popular imagination; the deity crossed his path as a kage — a shadowing monkey-form — and the iconography that later attached to the school runs through the high-nosed, human-featured register of Sarutahiko (猿田彦) — the same local deity also called Saruhi no Mikoto, for whom the school’s 猿飛 kata is named — the long-nosed monkey-deity and guide-figure of the tenson kōrin whose high nose (高鼻) is one of the standard medieval identifications of the tengu. By the time the set had a stable name it was Tengushō (天狗抄), but the original mythic encounter was with a shadow monkey-form more evocative of an upper-level tengu with human features than of a crow.

The Yagyū-ryū Shinpishō (柳生流新祕抄; Bujutsu sōsho, koma 162–166, pp. 304–313) uses Tengushō (天狗抄) as the name of a technique set:

“The foregoing Tengushō: against one who waits and takes a stance, one initiates, and by outside-and-inside breaks him and wins.”¹

The set includes Kasha (花車), Akemi (明身), Zentai (善待), Tebiki (手引), Murakumo (村雲) and 總南風, followed by Enpi (猿飛). Here 天狗 is purely a label for a curriculum division, with no narrative attached: across this volume the same graph serves as a marker of fabrication, as a literary frame, and as a curriculum heading.

The Kage-ryū revelation site is Udo (鵜戸) in Hyūga — present-day Udo Jingū (鵜戸神宮) in Nichinan, Miyazaki. The Gekiken Sōdan (撃剣叢談) (Mikami Genryū 1790) lineage opens “from Udo Daigongen” (鵜戸大権現). Miyazaki also related that Chūjō Hyōgonosuke (progenitor of Chūjō-ryū) learned sword and spear from the monk Jion (慈音) at the Udo grotto (鵜戶の岩屋) in Hyūga, and there attained mastery.²

Udo’s pre-Edo Shugendō affiliation is well-attested. By tradition it was founded in 782 (延暦元年) under a monk identified as the Tendai cleric Kōkibō Kaikyū (光喜坊快久), who became the first bettō (別当; superintendent-priest) and received the imperial temple-title Udosan Daigongen Abirasan Niō Gokoku-ji (鵜戸山大権現吾平山仁王護国寺). The shrine’s own history records that the sect later shifted to Shingon (真言宗) (Udo Jingū (shrine history) 〔n.d.; shrine self-account〕). Its mountain was at one point celebrated as “the Kōya of the West” and flourished as a great center of Ryōbu Shintō (両部神道). The complex was physically substantial: eighteen sub-cloisters (坊, bō) lined the Hatchō-zaka (八丁坂) approach, and a goma (護摩) hall stood near the cliff-carved Buddha (磨崖仏, magaibutsu) — by tradition a training ground for the Udo-san shugenja. During Aisu Ikōsai’s traditional era (c. 1452–1538), the site was a Shingon-flavored, Ryōbu-Shintō Shugendō complex. This influenced the formation of Aisu Kage-ryū symbolism.³

The “Tendai monk in 782” detail is internally anachronistic: the Japanese Tendai school postdates Saichō’s return from Tang China (c. 805), so Kōkibō Kaikyū and the 782 date belong to the shrine’s own engi (縁起; foundation legend) rather than to independently datable record. The bettō line later ran through Ninna-ji (仁和寺) administration as a Shingi Shingon (新義真言宗) house, which is consistent with the Shingon / Ryōbu-Shintō identity described here.

Tōzan-ha (当山派) was the Shingon-system line, based on Kinpusen (金峰山) with Daigo-ji Sanbō-in (醍醐寺三宝院) as its head temple, whereas the Honzan-ha (本山派) was the Tendai-system line, rooted in the Kumano Sanzan (熊野三山) with Shōgo-in (聖護院) as head temple. This is the standard Tendai-esoteric Honzan-ha versus Shingon-esoteric Tōzan-ha split. But Udo was not a direct Tōzan-ha subsidiary, as Kyushu itself was not Shōgo-in or Daigo-ji territory. The Honzan versus Tōzan dichotomy is a Kinki-centric organizational structure that hardened late: it was at the start of the early modern period that itinerant shugenja settled into village life and were largely organized into Honzan-ha and Tōzan-ha, a process the Tokugawa then enforced through the early-seventeenth-century Shugendō ordinance (修験道法度, 1613) compelling each practitioner to affiliate with one or the other. Ikōsai’s dates predate all of that.

Miyamoto Kesao’s standard typology (Miyamoto Kesao 2024) divides Shugendō into three, not two — Tendai Honzan-ha, Shingon Tōzan-ha and the “various mountains” (諸山; shozan) regional organizations. The regional centers — Dewa Sanzan (出羽三山), Kyushu’s Hikosan (英彦山) and the Chūgoku Goryū (五流; the Kojima Goryū of the Chūgoku region) — were powerful and strongly independent.

Hikosan in particular was a Tendai-system Shugendō organization that held subordinate temples across Kyushu, resisted Honzan-ha’s attempt to absorb its branches, won in court and ultimately came under the Tendai authority of Kan’ei-ji (寛永寺), hence “the Hikosan school” of Shugendō.

In the Muromachi period Hikosan’s records claim a faith-sphere covering the whole of Kyushu. Udo sat inside that southern-Kyushu world — alongside the Kirishima (霧島) complex — not under the Kinki head temples; tellingly, the shrine’s own present-day reciprocal-festival network still pairs it with Kirishima, Kagoshima and Hikosan.

So the precise statement is: the site itself was Shingon / Ryōbu-Shintō in its temple identity, which corresponds to the Tōzan stream in the later binary, but the Shugendō actually operating there in Ikōsai’s time belonged to the independent regional category (the Hikosan-and-Kirishima Kyushu sphere), before the Honzan/Tōzan labels meant anything locally.

We do not find documentation of any formal Honzan or Tōzan affiliation for Udo, and Nakaseko’s argument is only that Ikōsai was a shugenja — it does not assign him a specific line. He could have visited Udo, rather than being ensconced in it.

Pinning him personally to Tōzan-ha would therefore be unsupported on all three grounds (anachronism, region and silence). What the evidence does support is the softer and more useful claim for the Kage-ryū origin: a Shingon-esoteric, Ryōbu-Shintō, revelation-oriented mountain milieu of exactly the register the founding legend assumes.

Two threads in Ikōsai’s wider biography frame the shugenja reading. The first is a competing account of the same mobility: where Nakaseko reads him as a travelling shugenja, an older and more popular tradition — traced by Nakaseko to Yagyū Genchō’s Seiden Shinkage-ryū (正伝新陰流) and broadcast by the budō historian Wataya Kiyoshi (綿谷雪) — makes him a man of the Kumano/Ise suigun (水軍; naval levies) and, by extension, a wakō (倭寇; coastal raider). That reach is what is usually invoked to explain how a Kage-ryū catalogue reached Ming hands: Qī Jìguāng (戚継光) reproduced a fragment of the Kage-ryū no mokuroku (影流之目録) in his Jìxiào Xīnshū (紀效新書) (Qī Jìguāng 〔1560s; rev. c. 1584〕), later carried into the Wǔbèizhì (武備志).

The two readings are not exclusive — a younger son could be both a yamabushi and a man of the sea — but they pull the biography in different directions, and the wakō version is the one that travelled furthest in the popular literature.

The maritime record also gives Ikōsai concrete occasion to have been in Hyūga, independent of any revelation: the Aisu ships joined the Bunmei-era kenminsen (遣明船; Ming-tribute trade ships), and the 1483 voyage that carried Hisatada toward Beijing put in at Hyūga before crossing via Níngbō (寧波). Whatever drew him to the Udo cave, he had reason to pass that coast.⁴

The illustrated Aisu Kage-no-ryū mokuroku (愛洲陰之流目録) held by the Tokyo National Museum, with brush diagrams of stance and cutting lines, is an early primary artifact for the school and for the saru-tobi (猿飛) and enpi (燕飛) naming, independent of later Shinkage emokuroku. Enpi was an Aisu Kage-no-ryū teaching and not invented by Kamiizumi Ise no Kami of Shinkage-ryū (Aisu Kage-ryū (transmission document) 〔Muromachi–Momoyama〕).

Marishiten (摩利支天; Skt. Marīcī) figures in Shinkage-ryū iconography, usually depicted in the preamble to the Yagyū Tengushō emokuroku (絵目録; illustrated transmission scroll). It belongs to the warrior-Mikkyō (密教; esoteric Buddhism) register: a personification of mirage and light who, being unseeable and unharmable, became the deity of invisibility (隠形; ongyō) and victory, carried as a helmet amulet by warriors. The warrant is scriptural — the Fó shuō Mólìzhītiān jīng (佛說摩利支天經) declares that one who upholds Marīcī’s name cannot be seen, known, seized, bound, harmed, or deceived.⁵ The tengu in this case are intermediaries for the deva — or, in the specific case of Aisu Ikōsai, a spider or monkey and its shadow.

The Marishiten link to Ikōsai is concrete: the Hirasawa family preserved a Marishiten shisha-sho (摩利支天使者書; “writ of the messenger of Marishiten”) attributed to his own hand (Hirasawa Michiari 〔Genroku era, c. 1688–1704〕), an esoteric text of travel-protection and divination of exactly the Mikkyō register the Shugendō reading predicts.

That register runs on into the Yagyū Shinkage-ryū descended from this Kage-ryū line, and the Shugendō inheritance is legible already in the school’s earliest writings. The Tsuki no Shō gokui of Munenori and Mitsuyoshi are built on esoteric vocabulary — the nine-syllable kuji (九字) and its ten-syllable jūji extension, written in the palm and cut with the sword-mudra; the seed-syllable (shuji, 種字) read into the shuji shuriken teaching; the five gokui mapped onto the five great elements — while the school’s Kashima tutelary carries the honji of Jūichimen Kannon (十一面観音, the Eleven-Headed Avalokiteśvara) in the Takuan kishōmon. The full documentary treatment of that Yagyū kuji praxis, read against the three-volume Shugendō ritual canon, is taken up in the second part of this volume, in the chapter on the nine syllables.

Matsunoya gaishū on the Monkey in Kage-ryū

Among the earliest attempts to explain, rather than merely repeat, the simian imagery attached to Aisu Ikō (愛洲移香) is an entry in the Matsunoya gaishū (松屋外集), the antiquarian miscellany of Oyamada Tomokiyo (小山田与清, 1783–1847) (Oyamada Tomokiyo, n.d.). The entry stands as item twenty-five of volume fourteen, headed saru no kenjutsu (猿の劔術; “monkey swordsmanship”), and is indexed in the volume’s own table of contents with the gloss Inryū Shinkage-ryū no mokuroku (陰流新陰流の目録). It occupies koma 44–46 of the digitisation, closing on koma 46 with an anecdote from Gyūbamon (牛馬問), juan 4, on Yagyū Tajima-no-kami’s two sparring monkeys.

The entry proceeds in two movements. Oyamada first assembles the received Japanese account of the school’s origin; he then sets beside it a chain of Chinese texts on the White Ape and asks whether the one derives from the other.

Oyamada opens by locating the Kage-ryū (陰流) mokuroku in the Wǔbèi zhì (武備志), juan 86, in the section zhènliàn zhì・jiàoyì (陣練制・教藝; “training regulations, instruction in the arts”), part three, noting the accompanying illustrations of a small monkey at swordsmanship (Mao Yuanyi 1621). He then quotes Matsushita Kenrin (松下見林) at Ishō Nihon den (異称日本伝), middle part, juan 6, leaf 9 recto (Matsushita Kenrin 1693):

“Toward the end of the Ashikaga age there was one Aisu Ikō, Governor of Hyūga. He had polished the frost-blade for many years. He made pilgrimage to Udo Gongen (鵜戸権現) and prayed that his art might be brought to refinement. In a dream the deity manifested in the form of a monkey and revealed the innermost gokui to him. His name became known to a thousand generations, and he named his house Kage-ryū. His disciple Kamiizumi Musashi-no-kami Fujiwara Nobutsuna applied his mind to it, added and subtracted, and styled it Shinkage-ryū (新陰流). There are the techniques enpi (猿飛), enkai (猿回), yamakage (山影), tsukikage (月影), ukifune (浮船), uranami (浦波), rankō (覽行; reading unrecorded), matsukaze (松風), kasha (花車), chōtan (長短), tettei (徹底), isonami (磯波), and so on.”⁶

Oyamada then observes that the Wakan sansai zue (和漢三才図会), juan 20, leaf 1 recto, in the weapons division, gives the same account (Terajima Ryōan 1712). Two features of this passage deserve notice before Oyamada’s own argument is taken up. The Ishō Nihon den names the province as Hyūga, which is internally consistent with the Udo setting but diverges from the Ise and Iyo forms circulated elsewhere; and the technique names follow the usage of Oyamada’s own day, writing 影 in yamakage and tsukikage against the 陰 of the Wǔbèi zhì plate captions. The name 覽行 is not a slip of this manuscript,⁷ for the Gekiken sōdan (Bujutsu sōsho, koma 91, pp. 162–163; kan 1, Shinkage-ryū entry) lists 覽行, with 松風 and 花車, under the sangaku (三學) heading of Kamiizumi’s curriculum. Its reading (rankō) is not recorded in either source.

Oyamada next sets out, at length, the textual history of the White Ape swordsman. He begins from the Wúyuè chūnqiū (呉越春秋), juan 5, “Gōujiàn yīnmóu wàizhuàn” (勾践陰謀外伝), where the Maiden of Yue meets on the road an old man calling himself Yuán Gōng (袁公), who tests her by drawing down a bamboo whose tip breaks and falls. He notes the parallel in the Shìwén lèijù hòují (事文類聚後集), juan 37, and traces the couplet tóu tú Huángshí lǎo, xué jiàn báiyuán wēng (投圖黄石老、學劍白猿翁; “he cast his lot with the old man of Yellow Stone, he learned the sword from the white-ape elder”)⁸ through the Nénggǎizhāi mànlù (能改斎漫録), juan 1, to Dù Mù (杜牧), and adduces both Yǔ Xìn’s (庾信) epitaph for Yǔwén Shèng and Lǐ Bái’s (李白) line báiyuán cán jiànshù, Huángshí jiè bīngfú (白猿慙劍術、黄石借兵符; “the white ape was shamed in the art of the sword, Yellow Stone lent the tally of command”).

The pivot of the argument is the commentary he quotes at Lǐ Bái shījí bǔzhù (李白詩集補註), juan 4, leaf 28 recto, on the line shàonián xué jiànshù, línglì báiyuán gōng (少年學劍術、凌轢白猿公) from “Jiékè shàonián chǎng” (結客少年場). The note is by Yáng Qíxián (楊斉賢), whose jizhu underlies the Fēnlèi bǔzhù Lǐ Tàibái shī (分類補註李太白詩):⁹¹⁰

In Yue there was a maiden skilled in the art of the sword, and the King of Yue summoned her. As the maiden was about to go north to see the king, she met an old man on the road who called himself Yuan Gong. He said, “I hear you are skilled with the sword; I should like to see it once.” The maiden said, “Only test it, sir.” Yuan Gong thereupon sprang up into the bamboo grove; a dry branch broke and fell to the ground.¹¹ The maiden caught the tip; Yuan Gong gripped the butt and thrust at her; she answered the rhythm and entered, three times; then, raising her staff, she struck at him. Yuan Gong flew up into a tree, transformed into a white ape, and departed.

Oyamada quotes the two forms of the episode without remarking on the difference: in the Wúyuè chūnqiū Yuán Gōng draws down a bamboo whose tip breaks, while in the Yáng Qíxián note he springs into the grove and a dry branch breaks. The variance is material to any argument that the Japanese material derives from a specific recension.

Having laid the two bodies of material side by side, Oyamada conjectures that the nomenclature used in Aisu Kage-ryū was inspired by the Chinese white ape legend.¹² Oyamada notes that the Honchō bugei shōden (本朝武芸小伝), juan 6, leaf 1 verso, writes the founder’s name not as 愛洲移香 but as Aisu Koretaka (愛洲惟孝); that the same work, juan 6, leaf 1 recto, gives Kamiizumi as Ise-no-kami and omits his personal name; that the kenjutsu keizu (剣術系図) agrees; and that the school itself appears variously as Aisu Kage-ryū (愛洲陰流), Shinkage-ryū (新陰流), Shinkage-ryū (神陰流), and Shinkage-ryū (心陰流), with the writing of the technique names likewise not uniform (Hinatsu Shigetaka 1716). This establishes, from within the Edo-period record itself, that neither the founder’s name nor the school’s name had stabilised by 1716, and that no single orthography can be treated as the received or original form.

Matsunoya gaishū is not an independent witness to Aisu Ikō. Its Aisu content is quoted from the Ishō Nihon den (1688/1693), with the Wakan sansai zue (1712) adduced only as a source that “gives the same account”; both descend from the Wǔbèi zhì (1621) together with Edo-period lineage literature. Nothing in the entry constitutes testimony about sixteenth-century practice, but it might provide context for why Aisu visualized a monkey apparition in the first place.¹³

The hypothesis is testable in principle. It predicts that the simian terminology of Kage-ryū should prove traceable to the Yuán Gōng and White Ape Duke material rather than arising independently within Japanese transmission. It is weakened to the extent that enpi and cognate names can be shown attested in Japanese transmission earlier than, or independently of, exposure to the Wǔbèi zhì plate and its Chinese literary background — a test that remains available, since the Wǔbèi zhì transmission date is fixed at 1621 and is later than the mokuroku it reproduces. Examining surviving Aisu Kage-ryū mokuroku would be an important step for subsequent analysis.

Textual Mentions

Udo Gongen is mentioned in a lineage document (擊劍叢談; Bujutsu sōsho, koma 91, pp. 162–163):

“I have corrected the graphs from what is recorded in an old Shinkage-ryū menkyo certificate. What that certificate records is a lineage drawn from Udo Daigongen, then Aisu Ikō, Aisu Koshichirō, Kamiizumi Musashi-no-kami, Hikita…”¹⁴

This is a documentary witness in which the deity is the first name in the lineage — not a narrative claim about a revelation but a formal lineage entry. Mikami also notes he emended the writing of Ikō’s name on this certificate’s authority.

The monkey-form revelation is also found in the Yagyū-ryū Shinpishō (柳生流新祕抄; Bujutsu sōsho, koma 159, pp. 298–299):

“Aisu Ikō, Hyūga-no-kami, made pilgrimage to Udo Gongen and prayed for mastery in sword method; and in a dream the deity appeared in the form of a monkey and imparted the divine and marvellous innermost gokui. It was named Kage-ryū. It is recorded in the Wǔbèi zhì, which also gives illustrations of the hand-methods Enpi, Enkai, and so on.”¹⁵

The monkey form mentioned here links the revelation to the technique names, and the citation of the Wǔbèi zhì (武備志) puts a Chinese military compendium into the chain. This can be read beside Hinatsu’s insistence, in the Nen-ryū section that follows (§1.2), that the Udo revelation belongs to Ikō and not to Jion.

Sources and Further Reading

Further reading online

On Aisu Ikōsai and the founding of Kage-ryū, 歴史人 (opens in a new tab) (the Udo retreat and Ming-voyage chronology) and the 陰流 (opens in a new tab) article (the Bubishi-preserved Kage-ryū mokuroku). On Kamiizumi and the founding of Shinkage-ryū, the 新陰流 (opens in a new tab) article preserves the Eimokuroku self-attestation. For the kata catalogues, 肥後・新陰流の形 (opens in a new tab) and 「疋田新陰流」と「柳生新陰流」 (opens in a new tab) give line-by-line Higo / Owari / Hikita comparisons; the 新陰流兵法 関西転心会 (opens in a new tab) and 修猷館 (opens in a new tab) pages give the eight Emokuroku tengu names alongside the eight Tengushō technical names. On the tengu and Shugendō background, Ishizuchi-san on Western Mikkyō (opens in a new tab) and karunamitra.org (opens in a new tab) on Shugendō in practice; 鞍馬天狗 (Japanese Wikipedia) on the Sōjōbō / Maō Daisōjō identification and Kurama-kōkyō’s 1949 split from Tendai; and 鞍馬寺 その8 (opens in a new tab) on Kita Sadakichi’s 1922 article as the earliest print appearance of “Kurama-san Maō Daisōjō.”

The scholarly anchor for Aisu Ikōsai himself is Nakaseko Yoshimichi (中世古祥道); for the Shugendō typology it is Miyamoto Kesao, with Miyake Jun (宮家準) the other principal authority on the Honzan/Tōzan/shozan classification (Miyake Jun 1999) (his Shugendō, Kōdansha (Miyake Jun 2001), is the alternative typological citation). The primary anchors for the founding legend are the Hirasawa-house records (Ōmori Nobumasa 1986) and the Gekiken Sōdan. The competing maritime/wakō reading is surveyed in Nakaseko; its external witness is Qī Jìguāng’s Jìxiào Xīnshū, and the downstream nittō parallel is taken up in work on Ogasawara Genshinsai.

End Notes