Kaikei's Statue of in T#daiji

Christine Guth Kanda

Artibus Asiae, Vol. 43, No. 3. (1981 - 1982), pp. 190-208.

Stable URL: http://links.jstor.org/sici?sici=0004-3648%281981%2F1982%2943%3A3%3C190%3AKSOHIT%3E2.0.CO%3B2-E

Artibus Asiae is currently published by Artibus Asiae Publishers.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/artibus.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Oct 15 10:44:37 2007 CHRISTINE GUTH KANDA

Princeton UniversiQ

KAIKEI'S STATUE OF HACHIMAN IN TODAIJI*

he reconstruction of Tddaiji and Kdfukuji following their destruction by fire in 1180 Tduring the wars between the Minamoto and the Taira opened a new era in Japanese sculptural history. The scale, the symbolic importance of the project, and the new infusion of artistic talent it brought to Japanese enabled to regain some measure of the cultural eminence it had lost during the (794- I I 85). The forty-year period span- ning the last decades of the twelfth and first decades of the thirteenth century which was required to rebuild and refurbish these two vast temple complexes saw the rise of a new school of sculp- ture called Kei after the name of its founder IC6kei (act. late 12th c.). Kdkei and his disciples supplied most of the statues for the halls of Tddaiji and Kdfultuji. Their successful participation in these temple projects established them as the most sought after sculptors of the day. The Kei school's monopoly over the production of sculpture in was to last for several centuries. Kailtei (fl. I 180-1220) was a prominent member of the first generation of ICei school sculp- tors. Through his close ties with Shunjbbd Ch6gen (I I 2 I- I 206), superintendant of reconstruc- tion at T6daiji, Kaikei was given many important commissions at this temple. Among them was the creation of a statue of Hachiman (Plate I a, b), the tutelary deity of Tddaiji, which was completed in 1201. Although this work represents a native Japanese deity () and is the sole statue of this subject by Kaikei, it exemplifies many of the distinguishing traits of his mature style. The labyrinthine circumstances which preceded its creation also bear witness to the interaction of artistic, religious, and political forces at work at T6daiji in the early years of the period (II 8 5- I 3 3 3). TJdaiji and the Hachiman Cult Much controversy surrounds the origins and early history of Hachiman, but his meteoric rise to the national forefront in the eighth century unquestionably resulted in large part from his involvement with the construction of Tddaiji's Great Buddha and his adoption as the temple's tutelary deity, chinjztjin in 749.1 This fifty-two foot gilt bronze image of the Buddha was designed to symbolize the union of church and state under Emperor Shdmu (r. 724-748). Both technically and financially, the project was on a scale never before seen in Japan. Initially

* This article is part of a larger study of Hachiman imagery and its development. The author would like to thank Karen Brock for the many helpful suggestions and comments she made during its preparation. 1 On the origins of Hachiman see Nakano Hayatoshi, Hachiman shinkci- hi no k~nkyfi,2 vols. (Tokyo, 1962). it was to have been erected in Otsu, but in 744, following the failure of the first attempt to cast the bronze colossus, the project was transferred to Tbdaiji in Nara, a temple which was to become a powerful religious institution in the old capital.2 Its cost was so great that in addition to funding from imperial coffers, nationwide solicitation of contributions was required. Gydgi (670-749), a charismatic and rather controversial monk, led the fund-raising drive, traveling throughout Japan on the temple's behalf. An apocryphal account claims that in an attempt to draw support from native deities and the clans serving them, Gydgi even visited Ise to consult the Sun Goddess , ancestress of the imperial family. She is said to have signaled her approval by likening the Buddha Vairocana, also identified with the sun, to herself.3 However, during the Nara and Heian periods, the worshp of Amaterasu declined as the court turned its attention to Hachiman, whose cult, centered at Usa in Oita Prefecture, was deeply permeated with . The Usa Hachiman Shrine was an early benefactor of the Great Buddha construction pro- ject; according to records in the Shbsbin, it sent funds to Tddaiji early in the 740's.4 Usa Shrine, which had close ties with clans of Korean ancestry involved in the mining and forging of metals, may have also provided technical assistance in the casting of the colossal statue.5 Hachiman's support of its construction culminated in his adoption as tutelary deity of Tddaiji in 749.6 The practice of inviting native deities to serve as guardians of Buddhist temples had become widespread in the (710-784) as part of the Buddhist accommodation of the native religious beliefs. However, as cbinjzjin of Tddaiji, the official headquarters of state Buddhism, Hachiman's role was to protect not simply the temple, but the whole nation. Hachiman's prominence at Tbdaiji gave Oga no Tamuramaro and Oga no Morime, the priest and priestess who spoke his will through oracles, almost unlimited political and religious power. This power was soon abused: in 7j j, the priest, the priestess, the minister Fujiwara no Nakamaro, and Gydshin, a monk of Yakushiji, became involved in a political intrigue which resulted in exile for all participants.7 Later, in 767, the monk Ddky6 (d. 772) attempted to manipulate the Hachiman oracles to usurp the throne of Empress Kdken (r. 749-7j 9). This was prevented by the timely intervention of the courtier Wake no Kiyomaro (73 3-799), who during a visit to Usa received an oracle that the imperial line should remain unbroken.8 As a result of the Ddkyb incident, Hachiman assumed new dimensions as the guarantor of imperial legitimacy, and the Usa Shrine was invariably notified when an emperor acceded to the throne.9

Wnthis subject see Serge Elisskeff, "The Bommdkyd and the Great Buddha of Tcdaiji," HaruardJournalof Asiatic Sttrdies vol. I (1936) pp. 84-95. 3 Alicia Matsunaga, The Buddhist Philosoph_yof Assimilation-The Historical Development of the Honj Suijaku Theory, (Rutland, Vermont & Tokyo, 1969) pp. 164-165 4 Nakano, Hachiman, vol. I, p. 164. 5 Ibid., vol. I, pp. I 70-1 71. 6 This event is described in Shokunihongi, Shincho zbho kokushi taikei vol. 2 (Tokyo, 1935), under the year Tempyd shdhd I .XII.z7, p. 206. For a partial translation see Ross Bender, "The Hachiman Cult and the Dbkyd Incident," Monumenta Aripponica 2 (Summer, 1979) pp. I 35-1 36. 7 Shoku nihongi, Tempyd shbhb 6.XI.24, p.22~. 8 For a discussion of this affair and its implications see Bender, op.cit., esp.pp. 138-144. 9 The practice of sending an imperial messenger to notify the Hachiman Shrine at Usa of the accession of an emperor began with Emperor Kdnin (r.770-781), Empress Shdtoku's successor. See Okura Eiichi, "Iwashirnizu Hachiman-gfi sdshi no haikei-kyii seiki zengo no seiji ddkb o chiishin to shite," in Nihon shfikyino rekishi to minxoku, Commemorative Committee for Takeda Chdshii, comp. (Tokyo, 1962), p.91. During the ninth century many temples, first in Nara and later in , followed Tddaiji in adopting Hachiman as their chinjztjin. It was in this setting that images of Hachiman, who following an oracle of 784 was generally known as Hachiman Daibosatsu, were first created.10 According to a tenth century Daianji chronicle, a Hachiman Shrine was established at this Nara temple in 807 by the monk Gydkyd (act. c. 8oc-86o), and an image of the deity was housed within it.11 In 8j9, this same monk founded the Iwashimizu Hachiman Shrine outside Kyoto in response to an oracle that the deity wished to be near the capital in order to insure its pro- tection.~~There a statue of Hachiman was housed in the tenth century.13 The chronicles of both Jingoji, a temple on the outskirts of Kyoto, and of Tdji, in the very heart of the city, record their possession of ninth century images of the deity. The former document, however, refers to a painting of the deity alone, while the latter describes a statue of Hachiman with two attendants. Both the Tdji statues and the Jingoji painting are ascribed in these documents to the hand of Kiikai (774-83>), the founder of the Shingon sect who had served at Tddaiji, Jingoji, and T6ji.14 Although there are many references to the existence of images of Hachiman made during the Heian period, few examples have survived. The earliest is a triad at Tdji, consisting of Hachiman (Plate 2) and two female attendants. It may be dated to the second half of the ninth century on the basis of style and technique. Close in date though far removed in style is a second triad in Yaliushiji (Plate 3).15 Both Hachiman triads were originally housed in shrines in the temple compounds and functioned as , literally, "god-body", the embodiment of the deity's sacred powers. In both cases, Hachiman is portrayed in the guise of a Buddhist monk while his attendants wear secular attire, Originally Hachiman may have held a jewel or rosary in one hand and a monk's staff in the other. Although the iconography of Hachiman's por- trayal seems to be related to that of Jizd Bosatsu, the sole to appear as a Buddhist monk, it is likely that this mode of representation was chosen to signify that Hachiman had adopted Buddhist ways and was following in the footsteps of the Buddha. The image of Hachiman as a monk is a graphic expression of the phenomenon which has come to be known as shimbzltszl shigd, the merging of native Japanese and Buddhist divinities.

10 In 2. V.4, Hachiman had declared himself by oracle to be "Gokoku reigen iriki jintsii daijizai bosatsu," the Great Bodhisattva of State Protecting, Miraculous, Independent Powers. This is recorded in the earliest chronicle of the Usa Hachiman Shrine, compiled during the Jdwa era (834-48). See Usa Hachiman-gti Mirokuji konrytiengi, Dainihon komonjo, iewake IV, Iwashimizu monjo 2 (Tokyo, I~IO),p. 88. " This is based on a tenth century record, the "Daianji Hachiman-gii gochinza-ki narabi Tdchiiin konryii no shidai," quoted in Mori Chizuko, "Hachiman goshinei no seiritsu ni tsuite," Bukkydgeijutsu 128 (January, 1980), pp. 53-15. 12 The founding of the Iwashimizu Shrine is described in Iwashimixu Hachiman-gtigokokuji ryakki, said to have been written by Gydkyd, and Iwashimixu senxa ryaku , a work of 991. See Dainihon Komonjo, iewake 4, Iwashirnizu monjo 5, (Tokyo, 1910) pp. I 10-1 I 5 and Iwashimizu monjo I, pp. 107-1 10 respectively. 13 This statue is the same one referred to in Tddaiji Hachimangenki as having been destroyed in I 140. See text p. 203. I4 Jingojityakki, Dainihon bukkydxensho v0l.23 (Tokyo, 1972) pp. 167-168 and Tihcki, Zoku zoku gunsho ruijii 12, shiikydbu (Tokyo, 19081, PP. 14-56. 15 For detailed discussions see Kuno Takeshi, "Tdji sdzBki no chdzd," in Heian shoki chckokti-shi no kenkyti, (Tokyo, 1974) pp.71-99; Kurata Bunsaku, "Tdji Hachiman sanshin ni tsuite," Yamato 26 (19~8),pp. 1-10; and Saitd Takashi, "Tdji Saiin Hachiman sanshinzd no sakufii: Jdgan chdkoku ni okeru iwayuru rnikkyb-yd no mondai," Bigaku ronkyti 4 (1966), pp.68-92. On the Yakushiji triad see Oka Naomi, Shinxd chdkokti no kenkyti (Tokyo, 1971), pp.145-156; and Yaktishiji, Nara rokudaiji taikan 6 (Tokyo, 1970), pp. 73-75. Plate la Hachivtan: painted wood ; h. 87. I cm. ; by Icaikei; dated I 201 ; Todaiji, Nara. Photograph after Nara Kokitdaiji taikan: Tfidazji vol. 3, plate > 6 Plate ~b Detail of plate la. Plate 2 Hachiman (central figure of a triad); painted wood; h. 109 cm.; 2nd half of 9th c.; T6ji, Kyoto. Photograph after Genshoku nihon no bijcrtsu vol. 5, plate I 7

Plate 3 Hachiman (central figure of a triad); painted wood; h. 38.6 cm.; 2nd half of 9th c.; Yakushiji, Nara. Photograph after Genshoku nibon no bijtltsu vol. 5, plate 69 Plate 4 Gjgga, one of six Hosso patriarchs; painted wood; h. 76.3 cm. by KBkei; dated 1189; Kdfukuji, Nara. Photograph after Kamakura no ch6koku, pl.7

Plate 5 Jki Bosats~c;painted wood; h. 89.5 cm.; attributed to ; late I zth-arly I lth c. ; Rokuharamitsuji, Kyoto. Photograph after Unkei no ch6kokm, plate 31 Plate 6 Amida (detail); painted wood; h. 79.3 ; by Kaikei; dated 1221; Kbdaiin, Mount Kbya. Photograph after Hihi: K@asajz, plate I 80 Plate 7 Hachiman; ink and colors on silk; h. 148, w. 118 cm.; early 14th century; Jingoji, Kyoto. Photograph courtesy of Kyoto National Museum

Plate 8 Hachiman; painted wood; h. 81.8 cm; by Koshun; dated I 328; Museum of Fine Arts, Boston. Photograph courtesy of Museum of Fine Arts Plate 9 Hachiman; painted wood; h. 51.1 cm.; early 14th century; Shbmybji, Kamakura. Photograph after Kamakura no biintszl, plate I 5 5 Plate 10 ha chi ma^; painted wood; h. 5 1.6 cm.; dated 1277-1288; Plate 11 Hachiman (central image of a triad); wood; h. 71.6 cm.; by Ky6kaku; Hatogamine Hachiman Shrine, Nagano Pref. dated I 326; Akaana FIachiman Shrine, Shimane Pref. Photograph after Jiyd bunkaxai vol. 5 ch6koku 5, plate 2 Photograph courtesy of Prof. Shimizu Zenzo, Kyoto University Like the Heian period images described above, the Tddaiji Hachiman is portrayed in the guise of a seated monk with a rosary in the left hand and a staff in the right. Beyond these iconographic conventions, Kaikei's statue has little in common with those in Tdji and Yakushiji. The Tddaiji Hachiman has a realism in form and detail characteristic of much of the sculpture of the .16 Such realism is not grounded in a scientific study of human anatomy and movements, but rather in an interest in precise outward appearances reflecting a new awareness of the realities of the phenomenal world. Hachiman's naturalistic proportions and especially the objective and factual rendering of his clerical costume-with its grey under- garment, ochre robe, and gaudy patchwork surplice-typify the artistic outlook of Kaikei and his contemporaries. The modest closure of this garment just below the collarbone, so different from the bared chest of the Tdji version, is a further indication of the artist's concern with rendering the deity in the likeness of an actual monk. However, Hachiman cannot be mistaken for a mortal. The lotus pedestal on which he is seated and the gold halo framing his head and shoulders are obvious signs of divine status. The restrained treatment of the face, which lacks sharp contours, signs of personality, or age, further distinguishes this statue from historical portraiture. Hachiman's expressionless face may be contrasted, for instance, with that of one of the six imaginary portraits of Hdssd patriarchs carved by Kdkei for Kdfukuji in I 189. Only faint rows of conventionalized wrinkles above and between the eyebrows and to each side of the mouth mar the smooth surface of Hachiman's face; the patriarch on the other hand, (Plate 4) has bulging veins, sagging eyelids, and deeply furrowed cheeks. Such signs of frailty and human mortality are frequently employed in priestly imagery of the Kamakura period to bring out the individuality of the subject, but not in re- presentations of Jizd or Hachiman. Kaikei, like Kdkei and the latter's son Unkei (d. 1223), was firmly committed to the formal integrity of the human figure, but he did not exploit the plastic qualities of his material in the same way. The statue of Hachiman shows far less concern with producing palpable forms which assert themselves in space than with creating a visual unity of surface. Both in its sense of volume and in the treatment of drapery, this image may be contrasted with one of Jizd in Rokuhara- mitsuji (Plate j), attributed to Unkei.17 Its ample frame, over which a complex and varied arrangement of garments creates a vibrant play of light and shadow, has a tactile quality absent in Kaikei's Hachiman. The slender form and shallow, even carving of the statue of Hachiman are traits associated with Kaikei's mature An'arni style, so-called after the religious name given the sculptor by Chdgen. The gradual codification of this style, which would endure for generations among Kaikei's followers, may be clearly documented through the changing appearance of images of Amida carved between I200 and I 22 I .IS The serenity of an Amida carved in I 22 I (Plate 6), with

16 On the Tddaiji Hachiman see Inagaki Su'nao, "Kaikei saku Tddaiji Hachiman shinzd ni tsuite," Kokka 916 (1968), pp.23-30; and Akamatsu Toshihide, "Kaikei saku Tddaiji Hachiman shin26 ni tsuite," in ~okuKamakgra bzrkkj6 no kenkyi (Kyoto, 1966), pp. 41 5-427. 17 On the Rokuharamitsuji Jizd, see Kuno Takeshi, Unkei no ch6kdktl (Tokyo, 1974), pp.165-66. 1s For a list of Kaikei's statues of Amida see my "Kaikei's Statues of MaiijuSri and Four Attendants in Abe no Monjuin," Archives of Asian Art XXXII (1979), p. 22. its carefully modulated treatment of garments, and formalized, even sentimental facial expression exemplifies the fully developed An'arni formula. The Tddaiji Hachiman differs somewhat in construction from other works of the period.19 A technique involving a relatively few components has been used instead of the customary complex interlocking system of multiple blocks of wood (yosegi-pukzlri) favored for most thir- teenth century sculpture. Furthermore, the eyes are painted rather than crystal inserts. Their naturalistic coloring, indeed that of the entire figure, has been exceptionally well-preserved because the image was never exhibited to devotees and was consequently rarely exposed to light. The head and body of the statue are comprised basically of two large units joined front to back leaving a large uninterrupted cavity inside the figure. This hollow area has been covered with a cloth soaked in cinnabar red lacquer. The shoulders, hands, and knees are attached with dowels to the main trunk. The conservative sculpting technique, combined with the painting of the statue's interior with red, a color traditionally thought to have magical protective pro- perties,~~may reflect the sculptor's sensitivity to the non-Buddhist subject represented. The names of the donors and participants in the dedication ceremony completely cover the smooth surface of the cavity within the image. The list begins with the names of members of the imperial family. Among them are the ruling Emperor Tsuchimiltado (r. 1198-IZIO), the retired Emperor Gotoba (r. 1183-1198), and Shfikaku (115-IZO~), Prince-monk of Ninnaji. These are followed by the names of monks of Tddaiji and of the deceased Emperor Goshirakawa (d. II~Z),alive when the project to create the statue was conceived but not at its dedication. Next appears the list of sculptors: Kdshd Amidabutsu Kaikei followed by the names of the some thirty craftsmen who served under him. Many of these may be identified as Kaikei's disciples by the Kai in their names; others, as members of the once powerful En school, now eclipsed by the Kei workshops. A puzzling inclusion among the lesser sculptors is the name Unltei, which cannot refer to the master so named, son of I

'9 See Inagaki, op.cit., p.24. ZO On the protective properties of cinnabar red, see Kageyama Haruki, The Arts of , Christine Guth, tr., Arts of Japan 4 (Tokyo & New York, 1973), p.49. See Kanda, op.cit., esp. pp. I 2-1 3. 22 See Namu Amidabutsu Saxenshi, published in Bijzrtsu kenkyzi go (June, 1934), p.43. 23 Evidence of Mybe's interest in Hachiman may be found in his letters. See Kageyama Haruki, "Mybe Shbnin no tegarni," K6bijutsu 16 (1967), pp.99-102. talented sculptors working in the Nara area, but a member of the Xmidist religious fraternity of which Chdgen himself was the spiritual leader. As a sign of their devotion to Arnida, its members, including Kaikei, all bore names with the characters Arni. The patronage of this powerful group played a decisive role in Kaikei's career.

Special Circumstances Stlrrotlnding the Creatiorz of the T6davi Statl~e Many statues of the Kamakura period have inscriptions revealing the date of their comple- tion, the names of their donors, and of their sculptors. Despite this, it is generally difficult to reconstruct the circumstances surrounding their creation. This is not true of the Tddaiji Hachi- man, which became such a came cdL2bre that the problems arising from the Tddaiji monks' desire to have an image of the deity are mentioned in several chronicles and diaries of the period. The T6dagi Hachiman , compiled in 1237, draws on these contemporary accounts to paint a very detailed picture of the process from conception to completion.24 By understanding the background of Kaikei's statue we may better understand why this portrayal of Hachiman looks the way it does. From the contents of the T6dagi Hachimangenki we learn the following. When the Hachiman Shrine was destroyed by fire in I I 80, no official record of its contents remained, and whether or not an image or other shintai had been housed there was uncertain. What is certain and signif- icant is that in 1188, the Tddaiji monks led by Chdgen wanted an image to be installed as shifztai in the new shrine. Their first step therefore, was to discover a suitable model. Usa Hachi- man Shrine, as the oldest of the sanctuaries dedicated to this deity, was first consulted. Since it did not have an image of Hachiman, it could not help. Next the court and its counsellors were consulted. They advised that although Hachiman might be represented in the form of a monk, uncertainty of the existence of an image before the fire rendered it unwise to give any opinion on the matter. For good measure they added that after the priestly statue at Iwashimizu Hachi- man Shrine (commissioned by Prince Atsuzane in the tenth century) had been destroyed by fire in 1140, the court had ruled that because no image was housed at Usa, permission to carve a new statue was denied. Based on this precedent, they said, the same should hold true at Tddaiji. The court's decision caused some consternation among those involved in reconstructing the shrine. It was followed by a series of inauspicious oracles and dreams. A venerable priest of the Tbdaiji Hachiman Shrine conveniently declared that there was indeed a record indicating that images of Hachiman and two attendants had been enshrined before the fire. But this infor- mation was not thought to be reliable enough to serve as precedent so in I 197, thirty or forty monks went to the court once again. They told that a monk of Tddaiji had had a vision of a red-robed man standing near the Great South Gate of the temple lamenting the lack of a dwelling place. This they identified as the homeless spirit of Hachiman. Moreover, having learned of the existence of a painting believed to be by KGkai's hand, they requested that it be given to Tddaiji to serve as shintai in the new shrine. This painting belonged to Jingoji, but had been given to Retired (r. 1108-1123) and had remained in his collection at Shdkdmydin even after his death.

Z4 T6daiji Hachiman genki, Zoku gunsho ruijii, vol. j part I, shinjibu (Tokyo, 1914), esp.pp 245-248. When word of Tddaiji's request became known, Mongaku Shbnin (d.c. IZO~),who was then attempting to restore Jingoji with the patronage of Minamoto no Yoritomo (I 147-1 199), and a priest of the Iwashimizu Hachiman Shrine each insisted that the painting be donated instead to his institution. In the end, Chbgen's request was refused, and he secretly ordered a new shintai made-the present statue by Kailtei.

The unusual publicity surrounding the creation of Hachiman's shintai was due to the politico- religious implications both of the Tbdaiji reconstruction and of the worship of Hachiman. The rebuilding and refurbishing of this temple was an undertaking which mobilized widely dis- parate segments of the population. It brought together resources from imperial coffers, and from the Minamoto warrior clan, as well as from the public at large. It provided a focus of attention which contributed to the growth of a new sense of national purpose and harmony following the wars between the Taira and the Minamoto. By the opening years of the Kamakura period, Hachiman was among the most powerful kami in the nation. In the Nara period, Hachiman had become a guardian of the imperial family and especially of imperial legitimacy. Identified with Emperor Ojin in the Heian period, he assumed the status of imperial ancestor. His adoption as the tutelary deity of the Minamoto (the Seiwa ) added a new martial dimension to his character in the Kamakura period. These metamorphoses set the stage for Jien's ranking Hachiman on a par with the Imperial Ancestress Amaterasu and the Fujiwara clan's Kasuga deities. Thus in his Gz/k:anshd, an interpretive written in 1219, he states: The Sun Goddess enshrined at Ise Shrine and the Great Illuminating Kami enshrined at Kasuga certainly consulted together and decided (how Imperial rule was to be supported) in the distant past. And the Great Hachiman Bodhisattva and the Great Illuminating Kami of Kasuga consulted together and decided (how Imperial rule is to be supported) in the present. Thus the state was and is to be maintained. It is clear that the decision for the present, made after the state had been buffeted this way and that, has been made for these final reigns and requires that the sovereign have a guardian who has the power of both learning and military might.25 All these factors, however, do not entirely account for the unusual publicity given the crea- tion of Hachiman's new shintai. Although portrayals of native deities had been made as early as the ninth century, the number of shrines housing images rather than the more traditional sword, mirror, or other emblems as shintai was relatively small. Furthermore, knowledge of the existence and appearance of such images was limited because they were generally kept from public view. Because of the potential threat of their powers to spiritually unprepared individuals, even the shrine priest saw them only on rare occasions.26 Consequently, the practice of making iconographic drawings, integral to the Buddhist artistic tradition, was never encouraged in the

25 See Delmer Brown and Ichiro Ishida, tr., The Future and the Past: a Translation of the G~kansbi,an Interpretive History of Japan Written in 1219 (Berkeley, 1979), p. 228. The most comprehensive discussions of Hachiman's metamorphoses are ICIiyaji Naokazu, Hachiman-gi no kenkyc (Tokyo I 9jG), and Ross Bender, The Political Meaning of the Hachiman Cult in Ancient and Early Mediaeval Japan. Ph. D. thesis, Columbia University, 1980. 26 This is apparent from the confusion surrounding the identities of the deities represented at Tdji. See Tihiki, p. 5 5. case of images of kami. It is hardly surprising then that confusion, concern, and drama followed the destruction of the Tddaiji Hachiman Shrine. Whereas the court took the conservative position that the shrine did not need to have an image as shintai, and the clergy at any cost wanted a statue or painting, it is obvious that both sought to anchor their positions on precedent. Making an image of so powerful a deity as Hachiman was no light matter. If the clergy was to overcome the court's opposition, a suitable model was essential. That the Tddaiji had never learned of the existence of the statues at Tdji or even nearby Yakushiji indicates the secrecy surrounding images functioning as shintai. Tales of the militant activities of the warrior monks of various powerful temples during the tumultuous decades preceding the establishment of the are well known from the Tale of Heike and other accounts of the Gempei ~ars.~7From the TCdaqi Hachiman genki it is clear that the tactics used by militant monks to intimidate and press their demands during these wars were still practiced in the I 190's. The visit by thirty or forty monks requesting the painting from Toba's collection, reinforced by the report that the homeless spirit of Hachiman had been seen at Tddaiji's Great South Gate is but a thinly veiled threat against the court. Surely the combined ill-will of the monks and of their guardian deity were factors in the decision involving the fate of the painting of Hachiman. But why was the painting not donated to Tddaiji? Donation of the work to this Nara temple while Jingoji and Iwashimizu were demanding it be handed over to them was obviously politically inadvisable. Refusal to all parties was the only solution which could prevent the outbreak of violence among the rival institutions. The bitter dispute between Tddaiji, Jingoji, and Iwashimizu Shrine over the painting of Hachiman vividly illustrates the competition be- tween temples in Nara and Kyoto which developed during the early Kamakura period. Tddaiji and Kdfukuji were not the only temples facing massive reconstruction projects; many Kyoto temples were also restoring buildings and images which had fallen into disrepair during the turbulence of the twelfth century. Funds for these costly endeavors were in short supply and patronage was sought among all segments of the population. The rivalry among Tddaiji, Jingoji, and Iwashimizu over the prestigious painting of Hachiman by Kiikai affected monks and their laymen patrons alike. How did Tddaiji learn of the existence and whereabouts of the Jingoji painting? The even- tual discovery of the image of Hachiman in the collection of Emperor Toba may point to the assistance of Shiikaku, Prince-monk of Ninnaji, whose name figures prominently in the Tddaiji statue's inscription. Although not mentioned in TCdaiji Hachiman genki, it appears that Shiikaku had become involved in the Hachiman affair in 1195, two years before the Tddaiji monks marched to Kyoto to demand the Jingoji painting for their temple. The diary of Zen- kaku (1174-IZ~),an associate of Shiikaku at Ninnaji, records that in 1195 he was ordered by the Prince-monk to enter a hall on the Ninnaji compound where a painting of Hachiman had been kept since the days of the prince-monk Kakuhd (1091-1153). This Ninnaji version had been secretly copied from the Jingoji painting by Kanliaku, the Ninnaji monk who had removed

27 For instance, see The Tale of Heike, Hiroshi Kitagawa & Bruce Tsuchida, tr., (Tokyo, 1977). vol. I, p.67.

205 the painting from Jingoji and offered it to Toba.28 The Ninnaji copy does not survive, but Zenkaku's diary provides a description of the painting he saw. This god is in the form of an aged monk. He wears a surplice. [He sits on] a red lotus flower. Above is the red globe of the sun. In his left hand he holds a six-ringed monk's staff. In the upper right is a colored cartouche inscribed with his name.29 In general, this description fits a painting of the fourteenth century in Jingoji today (figure 7).30 With the exception of the three-quarter view of the face, and the curious sun intersecting the halo, the painting and Kaikei's statue are very similar. This similarity indicates that Kaikei's statue, though not modeled after the original Jingoji painting, was copied from a work in the same lineage-probably the one in Ninnaji.

Other Statzles of Hachiman The unique history and appearance of the Tddaiji Hachiman becomes more striking when it is compared with other Kamakura period images of the deity. Distinguishing portrayals of Hachiman from those of Jizd, actual monks, or even other kami is often very dficult when a work is undocumented or its prevenance unknown.31 The number of images of the Kamakura period which may be securely identified as Hachiman is indeed small. Among them is a statue in the Boston Museum of Fine Arts (figure 8) inscribed inside the head, "Sdgyd Hachiman [made by] Hdgen Kdshun of Nara Kdfukuji in Kareki 12 [1328]".32 Unfortunately, the pro- venence of the piece can no longer be determined, but taking into consideration Kdshun's activities just before and after the creation of this image, it is probable that it was commissioned for a Hachiman Shrine affiliated with Kdfukuji or Saidaiji.33 The similarities between this work and Kaikei's are often overemphasized at the expense of their considerable differences. Kdshun (act. I 3 I j-I 369), belongs to the Kei school lineage, but his artistic inclinations follow Unkei rather than Kaikei. In fact, he signed one work, "Hdin Kdshun, sixth generation descendant of UnkeiV.34 Gone in Kdshun7s version are the overt pictorial overtones of the 1201 statue. Moreover, Hachiman is freed from the spatial ambiance dictated by the halo and lotus pedestal in the Tddaiji version. The self-contained, somewhat dry quality of the earlier work has given way to a more generous and fluid treatment of garments and

28 Zenkaku sitoki, in Ninnaji monjo, Dainihon shiryd vol.4 part j (Tokyo, 1906), under the heading Kenkya 8.VIII. 17, P.459. 29 Ibid. 30 The geneology of this painting of Hachiman is not clear, but it is possible that it was copied from the version in Toba's collection. For a discussion of this complex issue, see Kageyama Haruki, "Hachiman shin eizd no hatten," in Shintc b&if~ff,(Tokyo, 1973), PP. 219-242. 3l Among the other kami represented in the guise of a monk are Yanagimoto Prlyojin and Jfizenshi Gongen. For reproduc- tions see, Kageyama, The Arts of Shinto, figures 63 and 108. 32 On the Boston Hachiman see Nishikawa Kydtaro, "Kaikei saku Miroku bosatsu to Kdshun saku Sdgyd Hachiman 26," Kokka 916 (1968), pp.9-18; and Tomita Kdjiro, "Statue of Sdgyd Hachiman by Kdshun: Japanese Wooden Sculpture dated I 328", Boston Museum of Fine Arts Bullefin 3j (December 1937), pp. 78-81. 33 On Kdshun's activities, see Izumi Yahiro, "Kyiishfi shdzai no Kdshun, Kdsei zaimei butsuzd ni tsuite," Bukkybgeijufsu 76 (July, 19701, ~SP.~p.97-99. 34 Izumi, op. cit., pp. 94-95. facial features. The diagonal sweep of the garment across the chest and boldly curved folds of the skirt over the lap are not patterned after Kaikei's characteristic style, with its evenly distrib- uted arrangement of garments, but rather after the richer and more complex mode favored by Unkei, as illustrated in the Rokuharamitsuji Jizd (figure 5). Roughly contemporary in date to Kdshun's statue is a work in the Kamakura Shdmydji (figure 9), a temple founded by Hdjd Sanetoki (1224-1276) renowned for its library, the Kane- zawa bunko.35 Although similar to the Tddaiji statue in pose, the deity's wizened face, strange pointed head, and bulbous ears give this image a curiously individualistic appearance. If it is in fact a representation of Hachiman, the conventions used were much influenced by Lohan imagery, which was in vogue during the Kamakura period. This work is close in style to a group of statues of the Buddha and his Ten Disciples dated 1308, also in Shdmydji, and may have been made around the same time.36 Two dated statues of Hachiman in secular guise are preserved in provincial shrines. It is dficult to determine precisely why Hachiman came to be represented in courtly attire. A probable esplanation lies in his identification with Emperor Ojin: his resulting image as im- perial ancestor and god of war were thought to be incompatible with a priestly appearance. One such image (figure 10) is housed in the Hatogamine Hachiman Shrine in Iida, Nagano Prefecture.37 According to an inscription contained within its joined wood-block body, the statue was begun in 1277 and completed only in 1288. It was probably commissioned by the Ogasawara, a branch of the Minamoto who had erected the shrine in I 2 j 7. Close in date, but completely different both in style and iconography from the Kdshun statue is a monumental Hachiman triad in the Akaana Hachiman Shrine in Shimane Prefecture (figure 11). Inscriptions on wooden slips found within the statue provide the date Kareki I (1326); the name Enjd, a Buddhist monk affiliated with the shrine; and that of the sculptor Kydkaku (or Keikaku) of Yamashiro Province. Although Kydkaku's name does not appear on any other extant statues, as the character kaku is commonly found among the names of members of the Kei school working for Iwashimizu Hachiman Shrine (located in the former Yamashiro Province), it is likely that Kydliaku was connected with this institution. This is also supported by the fact that the Akaana Shrine was a subsidiary of the powerful Iwashimizu Hachiman-gC1.38 The Hatogamine and Akaana Hachiman are both depicted seated in courtly attire, yet they present striking differences. The former is attired in the soft garment worn by courtiers during the Nara and early Heian periods, the latter in the stiff, starched one adopted in the last years of the Heian period. While the latter's voluminous robes give him awesome dimensions, the former's more fluid ones, which gently mold his body, give him a more modest and unassuming presence. In Kydkaku's work, an aura of dignity and authority is achieved largely through plastic volume. In this respect, the Akaana Hachiman reflects the conventions used in the secular portraiture of the period. The statue in the Hatogamine Shrine, on the other hand, conveys the

35 On this statue see Suijaku bijtrtru, Nara National Museum, comp. (Nara, 1964), entry 106 and Kamakura no b&tm, Matsu- mot0 Eiichi, comp. (Tokyo, 19>8),entry I 5 5. 36 For reproductions, see Kamakura no bijutru, plates I j 3-1 >4. 37 SeeJEyd bunka~ai5 chbkoku V:Bunkachb Mombushb, comp. (Tokyo, 1974), plate 2 and accompanying text. 38 On the Akaana triad, see Oka, op.cif., pp. 166-171. exalted status of the subject through a more expressive and realistic rendering of the face. This kami's stern, resolute physiognomy, enlivened by glistening crystal eyes, commands the dev- otee's unquestioning respect of Hachman.

The statue of Hachiman at Tddaiji is an important example of Kaikei's mature sculptural style. It also vividly illustrates the contrasting aesthetic outlook of Kaikei and of his contem- poraries Kdkei and Unkei. The inscription inside the statue combined with the account in T6dagi Hachiman genki give an exceptionally clear picture of the successive stages leading to the creation of a major icon. The dramatispersonae in this saga include many major personalities of the period. Because the reconstruction of Tddaiji and creation of Hachiman's shintai were national affairs, some of the problems described in Tadagi Hachiman genki are unique; others, however, are not and thus help shed light on the context of less well documented images of the Kamakura period. Although many statues of Hachiman were carved from the ninth century on, the secrecy surrounding those used as shintai and absence of iconographic drawings makes it impossible to establish any lineages among surviving works. Kaikei's statue is unique among portrayals of Hachiman in that it based on a known pictorial prototype.

Selected List of Chinese Characters in Alphabetical Order Akaana Hachiman-gii $RABg Kijsh6 An'amidabutsu Kaikei FJE~~@J$~@C{@~&?@ Amida PiiS%F% K6shun An'ami Em% Kijzanji Fg &+ Atsuzane 3s Kii'amidabutsu bly6hen Bpfil%FGBff)BaS chinjujin Q48$ Kyijkaku @if Chijgen B'Ei Mongaku Sh6nin Daianji %Z?+ My6e Sh6nin y$JlA Dijky6 hTinnaji Ebin S& Ogasawara /J\%E Enj6 Saidaiji 'Ek3 Fujiwara no Nakamaro %R4$Rjff;E Sazenshii YgS% Gotoba &,%TI shimbutsu shiig6 $$Bg& Gyijgi ;if& shintai $$@ Gy6ky6 K7& Sh6mu 22s Gyijshin fi1B Shijmy6ji @%% Hachiman Dai-bosatsu /\@$A%%% Shiikaku 45% Hatogamine Hachiman-gii L$2%A@B Shunj6b6 Ch6gen &~@%% hbgen %& S6gyij Hachiman @%AB h6in *&t$ Toba ,%a Hijj6 Sanetoki iL;l@%@ T6daiji Iwashimizu Hachiman-giiX&&A@&B T6daiji Hachiman genki %A~A@$&Z! Jingoji @#%% T6ji 35% Jiz6 Bosatsu i&%@$l Tsuchimikado Y@Ei KakuhB 5%& Usa 9% Kanezawa Bunko &.m@ Wake no Kiyomaro %I$%&@ Kankaku %B Yakushiji $@3 Kbfukuji B$Z% yosegi-zukuri %$% K6kei BE Zenkaku )%!% KBken @3