The honzon, or primary icon,[1] of the Japanese Buddhist temple Kanshinji is a statue of the bodhisattva Nyoirin Kannon (Figs. 1, 2, 5, 6, 12).[2] The ninth-century figure takes central position on the altar of the temple's main hall (Kondō) and is the locus of daily rituals performed at the temple. It is, however, rarely seen. For all but two days a year, the three-foot-high statue is secreted behind the double doors of a lacquered zushi, or shrine (the Nyoirin Kannon is behind the curtained double doors at center, Fig. 3). Annually, on April 17 and 18,[3] devotees flock by the hundreds to Kanshinji, located deep in the mountains south of Osaka (Fig. 4), to pray to the temporarily revealed secret image (Fig. 5). From a respectful distance they beseech the Nyoirin Kannon for blessings through the power of the “wish-granting gem” held in the innermost right hand of its six arms. Judging by the historical record and applying comparative stylistic analysis, art historians generally date the Nyoirin Kannon statue to about 840. It is thus the earliest surviving Japanese representation of this particular bodhisattva,[4] one of many Esoteric forms[5] of the Kannon (Sanskrit: Avalokiteśvara).

The Kanshinji image is well known to scholars as well as devotees because of its age, excellent preservation, and modern status as an emblem of Esoteric expression. In 1897 and again in 1951 the Japanese government designated the Nyoirin Kannon as a Kokuhō, or National Treasure, securing its status as part of the cultural patrimony and an important work within the canon of Japanese art history.[6] Despite its early canonical status specific details about the construction and preservation of the statue were not known until 1955. In the cold December of that year, Kanshinji's beloved secret image was assaulted by a zealot, after which the temple allowed it to be examined. The criminal files of the Osaka Prefectural Police Office detail the incident:

    A religious fanatic, mentally unstable, wandered about the
    country and stayed in various temples, studying books on
    religion and art. He became obsessed with the beautiful
    bodhisattva of Kanshinji temple. The man purportedly wished
    to obtain for himself the power of the Nyoirin Kannon,
    symbolized by the wish-granting jewel held in the icon's
    fight hand, which he imagined to be contained within the
    physical body of the statue. One night he dreamed that a
    red, moonlike sphere flew out from within the statue. The
    sacred icon was thus rendered powerless, and the man
    determined to destroy the statue and the hall that housed
    it. He went to Kanshinji and hid inside the main hall until
    it had closed to the public. Alone, he located the Nyoirin
    Kannon within its secreting shrine and tried to carry the
    image away. Finding it too heavy, in frustration he broke
    off two of the statue's hands.[7] The man then took the
    broken pieces outside and burned them ceremonially in rice
    fields near the temple. Days later, reading that the police
    were searching for the vandal, he turned himself in and
    confessed his crime.[8]

Following the incident, a team of art specialists was summoned to the temple to scrutinize and repair the damaged icon. The first definitive technical and material study of the secret statue was published one year later in 1956 by Nishiawa Shinji, a historian of sculptures who worked with the team.[9] Twenty-two years later, in 1978, he published another article on all the extant statues of Kanshinji, highlighting the Nyoirin Kannon image and discussing a range of relevant historical documents with which he attempted to construct a chronology of the extant works. By this time other scholars had also investigated the statue.[10] From a material viewpoint, limiting exposure over many centuries has superbly preserved the icon and its meticulous surface decoration (Figs. 6, 7). The artistic execution and overall preservation of the Kanshinji Nyoirin Kannon have earned it the praise of art historians, and it is frequently cited in the literature as a consummate example of period style and technique.

The main objective of this essay is to identify and examine how the Kanshinji Nyoirin Kannon attained its privileged position in both ancient and modern art histories, sectarian histories, cultural histories, and temple histories. Evaluation is a concept that comprises a range of activities and behaviors. The reception and status of works of art are changed by their own evaluational history.[11] Creative works and religious icons alike are singled out and become authoritative, that is, canonical, within a culture because they fulfill certain criteria of expression or perform certain functions. Art historians believe that the Kanshinji Nyoirin Kannon satisfies established criteria for artistic achievement among Buddhist statues, and the Japanese rank it among the greatest cultural assets of the country, a status that is readily accepted by non-Japanese scholars. While acknowledging the Nyoirin Kannon's historical and artistic significance, I wish to critically examine idealized treatments of the icon in order to open it up to new analyses and to reveal the need for close examination of other works made for the temple, both lost and extant. This essay will both participate in and examine a range of assessments concerning Kanshinji's religious icons.

The Nyoirin Kannon's current pride of place in both local tradition and academic discourse derives from its new status as a hidden image a century or more after its creation, legends and beliefs surrounding its origins, twentieth-century assessments of the statue's aesthetic and technical merit, and its modern promotion as a model of Esoteric expression in art.[12] Although such a history of meanings is part of the life of an icon, we should not assume a transcultural or transcontextual role for the Nyoirin Kannon statue. Canonization has reconstituted the Nyoirin Kannon's earlier function and setting, the historical circumstances of its production, and the larger history of the monastery—all with a bias toward the authority of the modern honzon.

This essay analyzes the ninth-century context at Kanshinji in light of the historical record, with awareness of the partiality that colors scholarly and popular literature. Second, it assesses the accretion of new meanings over time for the earliest icons at the temple, the addition of new icons and contexts and their functions, and the processes through which the Nyoirin Kannon has been isolated for both religious worship and academic praise. A third objective is to correct flawed characterizations of Esoteric Buddhism and Esoteric icons that derive from faulty deductions and assumptions about the Kanshinji Kannon.

Most troubling of modern interpretations is the unexamined assumption that the Nyoirin Kannon statue was always the horizon of Kanshinji and that it is the only one of the temple's early icons created under imperial sponsorship. In fact, a temple record dating to 888 shows that the Nyoirin Kannon image was originally but one icon among a group of statues and paintings made for the altar of the monastery's lecture hall. Two Buddha statues from this group still exist, and, like the Nyoirin Kannon, they may be dated using historical records and stylistic criteria to the mid-ninth century (Figs. 8–11).[13] One of the two extant Buddhas will be presented here as the most plausible honzon during the early history of Kanshinji. Although the first reliable record to describe the Nyoirin Kannon as a horizon is a pilgrimage journal of 1378 by the priest Kenki, the Kanshinji sankei shodō junreiki (A pilgrim's account of a visit to the various halls of Kanshinji),[14] Japanese scholars nonetheless assert that the Nyoirin Kannon has always been the honzon of the monastery and a secret icon, as it is today. They further maintain that the Nyoirin Kannon was made at least a decade before the two extant Buddhas and under the auspices of imperial sponsorship.[15] They base their conclusions on their estimation of the Nyoirin Kannon's superior artistic quality and fine preservation, fix its date according to stylistic criteria, and interpret the historical record with a bias toward an earlier date for it.

My own proposal is that the extant gih-wood-with-lacquer image of Butsugen butsumo Buddha (Figs. 8, 9, 17) is an equally likely honzon for the temple during the ninth century (and probably later), and that the Nyoirin Kannon's current designations developed over time. Widening the discussion should be relevant even if proof establishing the currently accepted hierarchy (Nyoirin Kannon as the sole main image), chronology (Nyoirin made first), and patronage (imperial sponsorship of the Nyoirin Kannon) emerges in future studies. Today the two extant ninth-century Buddhas are kept in the temple's museum (treasure hall). I do not intend to argue for reinserting the extant Butsugen butsumo Buddha statue into the modern liturgical context of today's worship hall, however, because the current religious status of icons cannot be contested on the grounds of “original history.”

Contemporary writers on art and literature may question the validity of a canon, its sources, and its supportive structure. Martin Jay proposes a plurality of sometimes competing visual regimes in modern France, none of which he sees as inherently superior.[16] Yet there are many who find the authority of the canon justifiable. Harold Bloom, for example, writes of appraising canonical authors: “ ‘aesthetic value’ is sometimes regarded as a suggestion of Immanuel Kant's rather than an actuality, but that has not been my experience during a lifetime of reading.” To the question of what makes a work canonical he notes, “the answer, more often than not, has turned out to be strangeness, a mode of originality that either cannot be assimilated, or that so assimilates us that we cease to see it as strange.”[17] A semiotic slant denaturalizes the canonical status of a work by acknowledging the contingent relationship between the perceived value of the work when it was created and the values of the interpreting culture.[18]

Art historical opinion regarding the Nyoirin Kannon statue has been presented through two primary types of description. The first stresses formal, stylistic, and material qualities, such as the statue's aesthetic value, expressive tenor, comparative relationship to previous types of statuary, method of construction, and ornamentation. This assessment is then related to characterizations of Esoteric Buddhism. The second type of description is based on historical records. The majority of the literature is by Japanese scholars, and their opinions have been adopted unquestioningly by Western researchers.

In the earliest analysis of the Nyoirin Kannon statue in English (1072) Sherwood Moran states:

    The subject of this study, the Nyoirin Kannon, is the
    principal image of the Hondō [main hall]; in fact, it
    is the only art object at Kanshinji that is of outstanding
    importance. In addition to this, it should be said that as
    representative of the best of Early Heian sculpture it
    stands among the highest of the National Treasures of the
    country. For its particular kind of sculpture, it may be
    said to be a superb combination of certain characteristics
    of the preceding period, along with the most satisfactory
    and ultimate expression of Shingon esoteric art. [19]

Moran's description parallels contemporary opinion among Japanese scholars following the zealot's attack: he praised the statue's artistic excellence, cultural significance, and unsurpassed expression of a Shingon Esoteric aesthetic. His essay does not elaborate on the features of Esoteric art; agreement about an “Esoteric aesthetic” was already widely held by Japanese scholars and the Japanese general public alike.

The Nyoirin Kannon statue has become the exemplar for a vague notion of Esoteric expression that is grounded in aesthetic response but not in doctrine, practice, or the perceived efficacy and power of the icon.[20] Descriptions of an image not only address appearances but also reveal attitudes toward the icon. Nishimura Kōchō, Esoteric priest and art conservator, writes in a book on the Kanshinji Nyoirin Kannon:

    Artistically, the work is of immense value. The figure is
    gloriously modeled, the flesh translucent …. It might
    not be appropriate to describe a Buddhist icon as sensual,
    and certainly the Nyoirin Kannon transcends carnality. But
    it exhibits a feminine quality that lingers on the eye, and
    simultaneously transcends this sensuality to convey the
    mysterious depths of the secrets of Shingon Esotericism.[21]

Many publications use photographs that artificially enhance through lighting and composition the sensual and mysterious traits described by Nishimura (Figs. 1, 2, 12); most of these are accompanied by texts penned by Shingon scholar-priests. Nishimura's text is an extreme example among many similar approaches to the icon.

Echoing general Japanese scholarly opinion, an American researcher described the image as follows:

    The sacred image of Nyoirin Kannon… exhibits a
    majestic, corpulent beauty; the round face and half closed
    eyes express great composure, the full lips do not smile but
    give an impression of aloofness. The garment is elaborately
    detailed and richly colored, while the skin is or a muted
    golden tone. The mysterious nature of this figure is
    enhanced by its six arms, each bearing great symbolic
    significance ….

    …Its rounded fleshy proportions, short neck and
    crescent-shaped eyebrows richly illustrate Esoteric
    sculptural characteristics of the Early Heian period.
    Through its entrancing, otherworldly beauty, the Kanshinji
    figure seems to embody the essence of the mysterious
    doctrine of Esoteric Buddhism.[22]

These excerpts from the copious literature on the statue are similar to many analyses that promote the Nyoirin Kannon as the model of an Esoteric aesthetic characterized by mystery, profundity, and sensuality. This definition reflects the opinions of nineteenth-century apologists for the Esoteric Buddhist tradition who in a process of “reverse orientalism” have implicitly or actively sought to neutralize the religious power and ritual framework of an icon—as suggested by its appearance, setting, or function—to more acceptable notions of artistic achievement, authenticity, historical value (such as noble patronage), or the individual goals of meditation.[23]

The Japanese exegesis raises many red flags. It is theoretically if not historically linked to what has been called the “Protestantization” of Buddhism: the notion that the thaumaturgic, sacerdotal, and image-based ritual functions of Buddhism are at odds with “true” soteriological and philosophical Buddhism. In this scheme, Buddhist icons are understood primarily as pedagogical tools.[24] They are said to inspire or visually convey the experience of the Absolute, but they are not identified as magical or powerful entities in themselves. The effect of such scholarship has been pervasive and now supports an essentialized image type for the Esoteric school of Buddhism. Paired with other canonizing strategies, the paradigmatic Esoteric style represented by the Nyoirin Kannon has determined the slant of extensive literature on Esoteric art.

Because Shingon ritual items and painted or sculpted icons are often withheld from public view, high-ranking Shingon priest-scholars became the dominant spokesmen for all aspects of the Esoteric tradition. They include authors of some of the most influential works on Esoteric art, including Sawa Ryūken, Ishida Hisatoyo, and Yamamoto Chiky&omacr, along with eminent Shingon Buddhist scholars such as Toganoo Shōun and Matsunaga Yūkei. As prime interpreters of the Esoteric tradition these men defined earlier religious elements according to Shingon doctrine and often became apologists for the Esoteric schools under the influence of prevailing Western attitudes toward Buddhism. Many Western researchers or general readers are unaware that these authors are clerics.

Barbara Herrnstein Smith suggests in Contingencies of Value that we canonize by suppressing the temporality of a work and shifting its definition to more comprehensible or acceptable ground. She states,

    For one thing, when the value of a work is seen as
    unquestionable, those of its features that would, in a
    noncanonical work, be found alienating… will be
    glozed over or backgrounded. In particular, features that
    conflict intolerably with the interests and ideologies of
    subsequent subjects (and, in the West, with those generally
    benign “humanistic” values for which canonical
    works are commonly celebrated) … will be repressed or
    rationalized, and there will be a tendency among humanistic
    scholars and academic critics to “save the text”
    by transferring the locus of its interest to more formal or
    structural features and/or by allegorizing its potentially
    alienating ideology to some more general
    (“universal”) level where it becomes more
    tolerable and also more readily interpretable in terms of
    contemporary ideologies.[25]

Sawa Ryūken, a Shingon priest and leading scholar of Esoteric religion and art, “rationalizes” the non-Japanese appearance of Esoteric art, making it (to quote Smith) “more tolerable and also more readily interpretable in terms of contemporary ideologies”—in this case a projection of unique Japanese attitudes and aesthetics. Sawa writes:

    Esoteric statues … tend to display volume and
    solemnity in the physique and awe-inspiring emotions in the
    face. Since these characteristics are not to be seen in
    Chinese statues of similar divinities, I attribute them to
    the reflection of a distinctively Japanese attitude toward
    religion. Although these traits are generally referred to as
    “esoteric,” in fact they are more a direct
    expression of purely Japanese feelings projected on the
    borrowed vehicle of Buddhist sculpture. This is the way the
    Japanese of the age felt that gods should look.[26]

In Sawa's analysis, the human spiritual experience is responsible for the “look” of the statue by projection, the “expression of purely Japanese feelings.” The power of the icon is shifted so that its aesthetic and spiritual impact conveys “a distinctively Japanese attitude toward religion,” but even this is not a direct expression of the work. It is a vehicle “borrowed” for the expression of a Japanese attitude, which suggests a further layering of neutralized bodies, including icon, art, and nation. Sawa further describes the ninth-century style associated with Esotericism as “fleshy, heavy,” and the Nyoirin Kannon as “soft in a weirdly bewitching fashion … part of the pure stream of [the Shingon master] Kūkai's Esoteric aesthetic.”[27] The aura that has customarily surrounded Esoteric icons in both academic and religious descriptions has hidden from view both the historical processes of canonization and earlier historical contexts. Within all the false claims made by Shingon priest-scholars, Esotericism is held firmly to its singularized status as an abstruse belief system, even in its attempted assimilation at the representational level.

The monastery of Kanshinji was established about 837 as a temple of the Shingon Esoteric Buddhist tradition. The English term Esoteric Buddhism refers to a wide range of teachings and historical schools within Mahāyāna Buddhism, including Indian Vajrayāna or Mantrayāna, Tibetan Tantrism, and Japanese Mikkyō (the Japanese equivalent of the Chinese mijiao), and there is considerable confusion and misuse among the terms. Mikkyō means, literally, “secret teachings”—thus the English translation “esoteric.” Esoteric Buddhism usually designates the practice of powerful and efficacious meditation arid ritual, including incantations, eidetic constructions of images (visualizations), fire rituals, sequenced hand or body movements, and invocations of powerful forces. Access to these practices is open to those who have received certain initiations and secret transmissions from a qualified master. Vajrayāna (vajra, or thunderbolt, vehicle) is an emic or insider's term with historical bias: it was developed by Indian practitioners to refer to a third and higher yāna superior to that of Theravāda (that is, Hinayāna Buddhism) or Mahāyāna. It is probably most appropriate to use it when referring to the traditions of India, Tibet, and Mongolia, and not to the Esoteric traditions of China, Korea, and Japan. Strictly speaking, Tantric Buddhism should be used to refer to traditions that advocate practices found in a particular class of Buddhist literature, tantra (it must be noted that tantra exist in the Hindu, Jain, and Buddhist traditions, but here I refer only to Buddhist tantra). Buddhist tantra, deeply related to Vedic-Brahramic traditions and drawing heavily from yogic systems in all the Indian traditions, achieved distinction around the sixth century C.E.; well-known examples include the Hevajra Tantra and Kālacakra Tantra. Tantra is also a practice. Tantra set out ritual practices, religious proscriptions, yogic techniques, and philosophical doctrine. David Gordon White proposes the following working definition of Tantra:

    Tantra is that Asian body of beliefs and practices which
    working from the principle that the universe we experience
    is nothing other than the concrete manifestation of the
    divine energy of the godhead that creates and maintains that
    universe, seeks to ritually appropriate and channel that
    energy, within the human microcosm, in creative and
    emancipatory ways.[28]

The Tantric traditions (typically used to refer to Tibetan practice) and Esoteric Buddhist traditions (typically used to refer to east Asian Tantric Buddhism) are rich, complex, and multiform and have long defied scholarly definition. Alhough Japanese Shingon is not the same as Tibetan Tantrism or Indian Vajrayāna, they display a number of common elements. Many of the scriptural and ritual texts in the Shingon canon are shared by other Tantric traditions. The Buddhist master Kūkai (774–835) brought the Shingon teachings from China to Japan in 806. Best known by his posthumous title, Kōbō Daishi, Kūkai is among the most venerated of persons—Buddhist or other—in Japanese history. He was ordained a member of the clergy in 804, after several years of practicing Buddhism as a privately ordained (thus unofficial and banned by the state) mendicant known as an ubasoku, in order to join a government-sponsored mission to China. Kūkai returned in 806 after less than two years of study, most of which was under the Esoteric master Huiguo.[29]

The term shigon, a transliteration of the Chinese zhenyan, literally means “true word” and is the translation of the Sanskrit mantra—a voiced string of syllables used to effect change or gather power in ritual practice.[30] Mantras feature prominently in Shingon ritual practice. Shingon is one of two Esoteric Buddhist schools in Japan, the other being Tendai (founded by Kūkai's contemporary Saichō, 767–822 C.E.). Kūkai's presentation of the Chinese master's teachings constituted a systematic body of Esoteric texts, practices, and religious objects and effected a substantially new inflection of earlier (unsystematized) esoteric ideas, practices, and images. Over time, proponents of the Shingon sect defined earlier forms and expressions of esotericism in contrast to their own tenets, labeling them zōmitsu (or miscellaneous esotericism).[31]

Kūkai, famous in his day, became a major Japanese cultural icon both in medieval and contemporary Japan. He succeeded in integrating his Esoteric teachings with existing rituals and beliefs in Japan, while he established support for Shingon at the monastic and imperial government levels. Many of the state rituals of medieval China were Esoteric, and Kūkai effectively introduced the Esoteric doctrine to Japan by citing its efficacy in China. The clergy and nobility ultimately accepted Kūkai's transmission as valid and efficacious,[32] a reception aided by the fact that Shingon was new and Chinese (Kūkai's initial friendship with Emperor Saga was driven by the emperor's fascination with the former's considerable knowledge of Chinese culture, especially poetry, calligraphy, and writing brushes). The material, ritual goods of Esotericism were certainly part of its appeal to the Japanese court and clergy. Kūkai brought with him from China hundreds of Esoteric sūtras (scripture), commentaries (sastras), and—in greatest number—ritual manuals (viddhis), along with painted mandala[33] and patriarch paintings, relics, Esoteric statues, ritual implements, iconographic sketches, and other items integral to ritual practice. Most of these texts, icons, and ritual goods were new to Japan, and they had tremendous cultural impact. The transmission of Esotericism highlights the unacknowledged fact that icons and imagery are not didactic tools or ritual aids for Esotericism but constitute the teachings themselves.

Shingon relies on the transmission of teachings from a qualified master to an adherent during many years of training, including textual study, oral transmission, and ritual instruction. The ritual language of mantras, prescribed hand movements (mūdras) of ritual practice, and meditative mental states—including visualizations—ritually replicate the “three mysteries” and three-part “body” of the Absolute Buddha, Dainichi (Sanskrit: Mahaavairocana), whose power and grace a Shingon adherent may unite with during ritual practice.[34] In this way Shingon doctrine, like the Vedic view of man as a microcosm of the cosmos, posits man as participant in the Buddha body (Dharmakāya), a state achieved through the ritual process. Whereas Tantric practice and thought typically expressed the duality of existence in sexual terms, Kūkai's writings expressed the same concept through discourse on and rituals of the Diamond and Womb World mandalas.

Statues and paintings (especially mandalas) and ritual implements feature prominently in the ritual systems of Shingon. But ritual texts are not always explicit about their use. Popular understanding erroneously represents the mandala as the visual focus of a rite or iconographic map of deities evoked in a rite. Statues and paintings function as an agent in Esoteric ritual, directly and indirectly. Because the Esoteric teachings allow that without images and icons, it is impossible for the practitioner to realize the ultimate Buddhist truth within himselt, imagery is part of what might be called an Esoteric logic of universal similarity. Imagery thus has a new function in Esoteric Buddhism. In exoteric (that is, non-Esoteric) contexts, images of the deities stood in for the absent Buddha. A wide range of sutras advocate image making and worship, but the distinction between image and deity is usually made clear. Kūkai blurred that distinction. He wrote:

     The dharma is fundamentally unable to be conveyed in words,
     yet without words it cannot be manifested. The ultimate
     reality[35] is beyond form[36] but in taking form it is
     comprehended.… The great variety of postures and
     mudrās depicted [in drawings and paintings] come from
     the great compassion [of the Buddha]. With a single glance
     [at them] one becomes a Buddha. The secrets of the sutras
     and commentaries are depicted in a general way in diagrams
     and illustrations, and the essentials of the Esoteric
     teachings are actually set forth therein. If [the diagrams
     and illustrations are] discarded, both those who receive
     and those who transmit the dharma will experience
     difficulty, for [the diagrams and illustrations] are none
     other than the source of the oceanlike assembly.[37]

The ritual practice of Shingon Esotericism is said by its adherents to reveal the profound and hidden teachings of the Absolute Buddha, teachings that otherwise remain hidden. Like Vairocana Buddha in the Flower Garland (Avatamsaka Sūtra), Kūkai describes the body of Mahatvairocana as inconceivable and formless, and as the originator of all forms. Mahavairocana is understood as the Dharmakaya, the timeless and formless Buddha body. Kūkai notes in the passage above that the absolute truth is beyond form, but at the same time images and forms are the source of the cosmos of deities. Elsewhere, he quotes the words of his Chinese teacher, Huiguo:

    Each of the three mysteries interfuses equally with the
    others to pervade all the corners of the world.
    Practitioners must therefore understand that all the
    objects of their sight are the all-permeating body [of the
    Dharmakā ya]. All the sounds they hear are…the
    voices of the [Dharmakāya]…. The practitioners'
    mind that understands this principle underlying all the
    sights and sounds of the world is the reality that is the
    divinities of the mandala. The reality is the divinities;
    the divinities, the practioners' own minds.[38]

Artistic representation and ritual practice, which I understand as distinct traditions with many common functions, expressions, and goals in the Shingon tradition, are conflated in the art history literature within the singular “ritual context.” That is, art, understood as an expedient in the ritual process, is thus effectively absent when ritual is present (context, seen as more complex than object, subsumes the object, which is understood as fixed in meaning). Conversely, “fine art” is a primary vehicle sutras the apologists' rendering of ritual, because its aesthetic or didactic construction—frequently supporting canonization—diverts attention from the apotropaic and thaumaturgic structures of ritual and art that constitute (in whole or part) rituals. Although the ritual function of icons varies considerably in the Esoteric Shingon tradition and is often not documented, we would do well to remember the critical function of icon and image in Esoteric praxis and in the fundamental exposition of ultimate reality as presented by Kūkai.

The 883 Official Register and Inventory for Kanshinji (Kanshinji kanroku engi shizaichō) is the earliest extant record of the temple and its holdings (Fig. 13).[39] According to reliable documentation found within the Register, a founding date of about 837 can be proposed for Kanshinji; thus, the inventory was compiled more than forty years later. The first section of the document describes the size and setting of the monastery: fifteen chō (one chō equals 2.45 acres) in Kawachi Province, Nishikori region, in the midst of the southern mountains. The four boundaries are “to the east, Ino waterfall; the west, Oninfukudani valley; the south [no description]; and the north, the mountain(s) of the Dragon spring temple.”[40]

The next portion of the Register provides a summary of the temple's early history, derived from several primary sources. From it we can deduce that the priest Shinshō (797–873) took up residence on the current site of Kanshinji from the year 827 for ten years.[41] Shinshō was a distinguished Shingon priest who trained primarily under Jitsue (786–847), Kūkai's senior disciple and designated heir.[42] The site's status apparently changed from private to official when in 837 Jitsue assigned the temple-hermitage a local administrator; the site was named Kanshinji at this time. The Register further notes that in the sixth month of 869 (Jōgan 11), forty-two years after he first took up residence at Kanshinji, Shinshō requested and was granted government recognition and support for the temple with its designation as a jōgakuji, or government-subsidized temple.[45] As if to signal assent, a government bequest of several estates (shōen) was made to the temple four days before the monk's petition was granted.[44] Shinshō explained in his petition that there were many pious and legitimate monks residing at the temple, whose ceaseless recitation of the sutras for the good of the nation, in accordance with the wishes of his master, Jitsue, could only be assured after his own death with the government's favor. Later legends connect Kūkai with the founding of the temple; although the earliest records do not mention him in this capacity, he may have known the site for reasons discussed below. Over time Kūkai came to be linked both with the building of the monastery and the creation of the Nyoirin Kannon statue (along with other statues concealed in the main hall today), all of which certainly occurred after his death. Whether historically substantiated or apocryphal, Kūkai's association with Kanshinji and its modern honzon figures prominently in the historical construction of meaning for each.

Kanshinji is situated along the route between the ninth-century capital of Heian-kytō (modern-day Kyōto)—and the previous capitals of Nara and Asuka—and Mt. Kōya, where in 816 Kūkai founded the important Shingon monastery of Kongōbuji. Even before Kūkai's time, countless travelers passed through the Kawachi region around Kanshinji and the current site of the monastery on their way northeast to Asuka or Nara from Naniwa (modern-day Osaka) on the Inland Sea.[45] We know from Kūkai's writings that he journeyed several times to Mt. Kōya with his disciples from 816 until his death in 835.[46] Jitsue and other priests also lived on Mt. Kōya at different periods prior to their master's death. They necessarily traveled through the Kawachi district to reach Mt. Kōya, which is one likely explanation for Kanshinji's founding by Jitsue and Shinshō. Unusual features of the temple plan and the iconography of the contents of the image halls (to be dealt with below) also suggest that the site had particular associations or meanings to Kūkai's disciples, and perhaps to the master himself. Regional influences may also have pertained.[47] Late popular accounts of the temple's founding or its miraculous Nyoirin Kannon icon (some of which will be noted in the next section) may be understood as ways of recognizing the region's ancient magical traditions and a popular claiming of the site once government sponsorship became more tenuous.[48]

The 883 Register lists the worship halls and other buildings standing at the temple by that year; it also provides the contents of some halls. Such inventories are invaluable documents for the study of temples and religious practices. The main precinct contained a three-by-three-bay chancel hall (that is, a five-bay square) named the Nyōhōdō and a five-by-five-bay chancel (that is, a seven-bay square) lecture hall (Kudō); the Register gives the contents for these two primary worship halls. Next are listed a six-by-seven-bay structure known as the fire-ritual hall (Gomadō), a fifteen-meter banner (tō) that served symbolically as a pagoda, a bell tower, and bell and gong.[40] The next important structure (bō) of six-by-seven bays served as the sutra repository, for which the textual contents are listed. The Register then names a storehouse (hōzō) and its contents of ritual goods; three buildings (of five-, seven-, and nine-bay chancels) comprising the sōbō, or monks' precinct; and finally a taishu-in, or public precinct, with a five-bay-wide refectory, an altar for kami (native gods), a large kitchen, a large cooking room, a milling structure, a rice storage room, a horse stable, a cow stable, a bathhouse, pots, kettles, and cauldrons.[50] This section of the inventory ends with a list of its landholdings and donations to the temple. With the exception of the highly unusual Nyōhōdō building in the main precinct, the plan is simpler than that for other Esoteric mountain monasteries of the time, as it lacks, for example, a pagoda.[51] (The disposition of buildings, including the main hall, in tile mountainous location today is shown in Fig. 4. The contents of the halls will be considered below.)

It is possible that not all tire structures and items inventoried in 883 existed in 869, when Kanshinji was granted state support as a jōgakuji. A completion date of 840 may be inferred, however, for the two image halls (Nyōhōdō and lecture hall) and other primary buildings (probably the dormitories, refectory, bell tower, and sutra repository: possibly the fire-ritual hall and banner). The source for this date is found in a petition for a bell to be cast for Kanshinji in the year 840.[52] Such a request was made only when the important halls of a monastery were completed or well under way and there was a need for a timepiece for daily priestly activities.[53]

Exoteric monastic compounds of the eighth century featured a main hall and lecture hall as image halls for liturgy and gathering, along with a pagoda (or a pair of pagodas) and, as above, dormitories, dining halls, repositories, and so on. Except for those that had not been planned as Esoteric temples from the start, early Shingon monastic plans did not include a main hall although another hall may have served a similar function (such as the Nyōhōdō).[54] The lecture hall was typically retained in the Esoteric layout for sutra recitation and instruction, and with a traditional nmin hall lacking, it became the major image hall. This traditional “lecture” hall (that is, a hall for sutra recitation and doctrinal lectures, as well as rites) was also a primary image hall, typically featuring a raised platform of statues. This type of hall was vital because Esoteric monasteries retained many of the same functions, liturgies, and rituals as did exoteric sites, such as sutra recitation, the lighting of lamps, offerings, or the gathering of monks before an altar of statues. Arguably, the most important hall in an early Esoteric temple was the Kanjōdō (also called the Shingondō), used for initiation rites known as kanjō (Sanskrit: abhiseka). These important rituals of a priest's training included the conferral of the precepts and the identification of each one's personal deity on the mandala.[55] Unlike established exoteric ceremonies held in statue balls, kanjō might involve only the supplicant and master. Whereas an exoteric image hall fieatured a raised altar of statues, the earliest Japanese initiation halls, derived from Tang Ghinese Esoteric examples experienced by Kūkai, featured altars laid out with ritual implements, mandala paintings, and a fire-ritual (goma) altar.[56] Kanshinji had no Kanjōdō in 883, the year the first inventory was made, nor in later centuries. It is likely that the Nyōhōdō, listed first for the main precinct, functioned as a ritual space similar to an initiation hall.

In terms of its name and its contents, however, the Nyōhōdō at Kanshinji is without precedent among Heian-period Shingon temples. Literally, “hall in accordance with the law,” the structure had a three-bay chancel (mōya) with an aisle (hisashi) on all four sides and a cypress-bark roof. An inventory of its contents in the Register names five paintings or sets of paintings (fourteen paintings total, all in the hanging scroll format) and one small gilt-bronze Shaka (Sukyamuni, the Historical Buddha, henceforth Shakyamuni) statue, named last in the list. Neither the hall nor any of these works survives today.[57] The first painting listed is a large Womb World mandala (Taizōkai mandara) measuring over five feet wide (eight fuku), one of two mandala images central to the Shingon Esoteric representational tradition established in Japan by Kūkai (Fig. 14 gives an example of the type).[58] Listed second, as a pair, are two paintings of the Nyoirin Kanjizai bosatsu (that is, Nyoirin Kannon), the same deity represented by the statue under discussion.[50] A note indicates that one Nyoirin image is executed in gold paint, the other in polychrome pigments. Listed third are a set of paintings depicting the Godaikokuzō, or Five Great Kokuzō, or Storehouse (Sanskrit: ōkōśagarbha), bodhisattvas; next, a single Fugen-enmei bosatsu (Sanskrit: Samantabhadra bodhisattva) image; and fifth, a set of “the five great divinities,” Godaison (that is, Myōō bosatsu, Lords of Radiance bodhisattvas; Sanskrit: Vidyōrōja).[60]

The Nyōhōdō is unusual in that it features paintings rather than statues. This strongly suggests, as above, that it was an Esoteric ritual practice hall. The paintings in the Nyōhōdō would not have been displayed all at once but taken from storage and hung alone or in sets for particular ceremonies. The overall iconographic meaning of the Nyōhōdō is unclear because the paintings listed do not derive from a single ritual text or sutra. Tire most important image would have been the Womb World mandala. It is one of a pair, the “Two Worlds mandala” (Ryōkai mandara) within the Shingon tradition. Although used in distinct rituals that, together with the goma rite, comprise a primary Esoteric initiation sequence,[61] the pair of mandala symbolizes and embodies the perfect spiritual and structural union of the vast assembly of Esoteric deities. The mandala of the Two Worlds, depending on one's viewpoint, are a symbolic visual synthesis and a conceptual matrix of an Esoteric Shingon ritual practice system thought to embody dimensions of the experience of a Buddha. Their structures and potential are always present, subtly or more overtly, in all Esotericism. The Womb World mandala “represents the ritual construction of the realm of enlightened beings delineated in the Mahāvairocana Sutra.”[62] Surprisingly, there is no mention of a Diamond World mandala anywhere in the Register or in subsequent records. If the Womb World mandala alone was displayed in the Kanshinji Nyōhōdō, it is an isolated example ot' unknown usage or iconography. This unusual feature is accentuated by the fact that each deity or deity group of the recorded paintings in the 883 Nyōhōdō hall is among the assembled deities of the Womb World mandala. The prominence of the Nyoirin Kannon in two painted versions is also unusual, if not unique. It is possible that paintings of the Nyoirin Kannon were used in ninth-century/Esoteric rituals. Ritual texts of the Sanbōin lineage feature visualizations of the deity.[65] Finally, the order in which the paintings are listed in the inventory might indicate a hierarchy of importance in an as yet unidentified ritual.

Image halls in exoteric Japanese monasteries, typically a main hall or lecture hall, usually designate a honzon, or primary icon. Image halls in early Esoteric monasteries, typically lecture halls or a pagoda rather than a main hall, featured a Dainichi Buddha or the Five Wisdom Buddhas (Gochi nyorai). An initiation hall or other ritual hall would not usually have had a horizon but, rather, sets of paintings brought out for ritual use, especially the pair of Diamond and World Womb mandala. The bronze Shakyamuni statue listed last in the Nyōhōdō inventory (last, presumably, because it is the only statue among paintings and represents a distinct object category/ritual category) likely had a permanent position in the hall for worship. Although the only statue and a common exoteric honzon, the Shakyamuni icon almost certainly did not function as a horizon in the Esoteric Nyōhōdō. Womb World iconography and possible initiation function are suggested by the contents of the Ny ōhōdō. If we open up our assessment of the hall to consider historical developments and not just specific iconographic clues, then the unusual hall and its contents may be understood to express the complementary functions of exoteric and Esoteric ritual and doctrine. The supreme exoteric Buddha, Shakyamuni—represented by a sculpture, the traditional exoteric locus of worship—is juxtaposed with divinities of a mandala, the paradigmatic Esoteric emblem, featuring (for as yet unknown reasons) individual deities and the assembled gods of the Womb World mandala.

It is impossible to determine the ritual function of the paintings listed in the inventory for the Nyōhōdō, together or individually, without specific ritual records. The existence of two Nyoirin Kannon paintings in a new type of hall with seemingly unorthodox usage of a Womb World mandala without a Diamond World mandala suggests special iconographic symbolism or ritual use for the hall. Some scholars believe this indicates special veneration of the Nyoirin Kannon within the hall by the 880s. In focusing on individual deities or direct correspondences between texts and icon groups, scholars have overlooked meaningful indications of function. A comprehensive view of iconographic correlations is helpful.

The hall is replete with Womb World mandala imagery.[64] The Womb World mandala, or mahākarunāgarbhodbhavamandala, is literally “the mandala generated from the womb of great compassion.” The Mahāvairocana sutra, among others, associates the womb with the receptive feminine principle; apart from the obvious physiological link between woman and womb, the Shingon tradition stresses the womb as a metaphor for the female. The female aspect of the mandala may be salient to this hall or the temple, as the discussion of patronage below will clarify. The central motif of the Womb World mandala is an eight-petal lotus flower (Fig. 14). Fugen-enmei bosatsu, the deity of the Nyōhōdō painting list, occupies one of these petals (at the southeast position) and symbolizes universal wisdom as the embodiment of pure and innate bodhicitta, the aspiration for enlightenment within each of us. Fugen is also a “feminine” deity, typically paired with the boyish bodhisattva Manjusri (Japanese: Monju); later secular images parodied courtesans and prostitutes as Fugen seated on its identifying elephant vehicle. In the Mahāvairocana sutras, which claims to have been transmitted from Mahāvairocana Buddha to the bodhisattva, Fugen manifests itself as Vajrasattva (Japanese: Kongōsatta), leader of the bodhisattvas. Finally, the form of Fugen named by the Register, Fugen-enmei, refers to its power to grant longevity (enrmei). Numerous ritual manuals imported by Kūkai describe the Esoteric worship of Fugen-enmei for long life.[65] A statue of Fugen-enmei was also part of the earliest recorded altar of statues in the lecture hall of Kongōbuji on Mt. Kōya, the mountain training center established by Kūkai. The contemporary Tendai priest Saichō performed the Fugen-e dempō (rite of Fugen) and the Nyoirin dempō, among others, for protection of the state; these rites were also well known to Kūkai.[66]

The five-deity painting set listed in the Register's Nyōhōdō section, the Godaison, depicts awesome Esoteric bodhisattva transmigrations (for a sculpted example, see Fig. 18). The five are the focus of rites for the protection of the sovereign and country, such as the Esoteric rite of the Benevolent Kings (Ninnōhō), first performed by Kūkai in 810. Paintings of the five are central to the Latter Seven-Day rite (Goshichinichi mishuhō), established by Kūkai in 834 as an Esoteric ritual complement to the exoteric Misai-e and first performed at the Shingonin (Shingon chapel) within the imperial palace.[67] They also figure in the 839 lecture hall altar designed by Kūkai for Tōji temple in the Heian capital (Figs. 16. 18, 24). Most noteworthy for the Womb World emphasis of the Nyōhōdō, four of these five important Godaison bodhisattvas are situated in the Womb World mandala (Fig. 14) in the so-called Court of the Mantra Holders, in a horizontal row just below the central eight-petaled lotus.[68] Thus, in addition to being part of the dominant Womb World iconography, this set of paintings may suggest rituals for national protection or relics worship.

The Five Great Storehouse bodhisattvas, listed next, also part of the Womb World mandala, are the five transformations of Ākāśagarbha, “One Whose Storehouse of Wisdom Is as Vast as Empty Space,” a deity of wisdom and good fortune. More than any other known impetus, a ritual text on this deity prompted Kūkai to leave his Confucian studies at the Nara State College and begin Buddhist training.[69] The activities of the Kokuzō are thought to bestow treasures, and the deity's symbolic (samaya) form in the Shingon tradition is the jewel, nyoiju. A rite associated with one of the Storehouse bodhisattvas is the Morning Star ritual (Gumonjihō), wherein the practitioner achieves an enlightened state through the energy of the magical wish-fulfilling jewel. In the Gunumjihō the Storehouse bodhisattva is understood as a planet or “morning star.” Typically, this physically demanding rite is conducted in isolation in the mountains over many days, often with a view of the sky through open windows.[70] Worship of the stars was of special significance at Kanshinji, a site deep in the mountains along the route to Mt. Kōya.

Imagery and symbolism related to relics worship, as manifested by the nyoiju wish-fulfilling jewel (held in the right hand of the Nyoirin Kannon) in the Esoteric tradition, was very prominent in the early esoteric temples and teachings. The nyoiju held special meaning for the imperial court and was a primary motif, along with the relics it symbolizes, in the Latter Seven-Day rite held within the imperial palace. The extended relationship of the wish-fulfilling gem to the Womb World mandala and Nyoirin Kannon—the first two types of paintings listed in the holdings of the Nyōhōdō—can be understood to convey strongly feminine symbolism and as a reference to the great bodhi mind contained in the eight-petaled lotus at the center of the Womb World mandala. (The particular import of the Womb World and nyoiju gem symbolism at Kanshinji will be discussed further in subsequent sections.) Scholars have overlooked relationships and meanings posited by the lost ninth-century Nyōhōdō inventory in favor of extant works. The Womb World mandala and its deities are central to the site, as is the power of the nyoiju. Also of plausible relevance are imperial rule, national protection, and relics worship. Moreover, some of the deities are relevant to texts imported by Kūkai or rites he introduced, as would be expected in temples founded during the early decades of Esoteric dissemination.

Next, let us turn to the lecture hall. The 883 Register describes the lecture hall as a five-bay chancel with an aisle on each side (that is, a structure seven-bays square) with a cypress-bark roof (Fig. 13).[71] Its contents of eight statues and three paintings are noted as follows:

    one gilt (konjiki) Butsugen butsumo nyorai statue [Sanskrit:
    Buddhalocanā] Buddha];

    one gilt (konjiki) Miroku nyorai statue [Maitreya; Buddha of
    the Future];

    one sandalwood color (danjiki) Yakushi nyorai statue
    [Bhaisajyaguru; Buddha of Healing]; [notation remarks:]
    lacks a mandoria, dedicated by the novice Yakuman;

    one polychromed (saishiki) Nyoirin bosatsu statue; [notation
    states:] approximately three shaku[72] in height, wooden
    statue;

    one painting of the Birushana [Vairocana]; [width] 3
    fuku;[73]

    one painting of the Shō Kannon [Ā
    valokiteśvara bodhisattva];

    one Tang [Chinese] portrait statue of a priest;

    one painting of Kongōdoshi [Kani-krodha or
    Subūhu; a lesser deity];

    one polychromed (saishiki) statue of Bishamontennō
    [Vaiśavana krodha; Guardian of the North]; [notation
    remarks:] dedicated by Kiyotaki Yoshio;

    a pair of ten statues [deva; lesser deities, probably
    guardian figures]

Whereas the bronze Shakyamuni statue and all the paintings originally housed in the Nyōhōdō are presumed lost, four statues listed by the Register for the original lecture hall are extant today: the first two Buddhas, Butsugen butsumo nyorai (Figs. 8, 9, 17) and Miroku nyorai (Figs. 10, 11),[74] the Nyoirin Kannon bodhisattva, listed fourth (Figs. 1, 2, 12), and the Chinese priest statue (Fig. 15). The Register does not indicate when the various statues and paintings were completed, nor which icon, if any, is the honzon, nor the arrangement of images on the altar. It notes the patrons for two works, the Yakushi Buddha and Bishamontennō guardian, both of which were important icon types in the early Heian period and at Tōji, the early Esoteric temple founded by Kūkai. It does not indicate that the Nyoirin Kannon statue was a secret image, nor that it was housed in a shrine, as it is today. The Register explicitly notes the statue's height and composition, calling attention to its material representation. Although this notation may support the hypothesis proposed by, some scholars that the statue was a hidden image, the size of the Vairocana painting in the lecture hall is noted, as is the composition of the Yakushi Buddha, distinguishing these two works as well. A benefactor for the Nyoirin Kannon is not mentioned, but it is noted for two other statues on the altar, which is suggestive in light of scholarly opinion today that links the Nyoirin Kannon statue—and no others—to imperial patronage.

The lecture hall statues would have been arranged on a raised altar, as in earlier non-Esoteric image halls or the contemporaneous Esoteric lecture hall at Tōji (Figs. 16, 18, 94). The paintings listed in the inventory would likely have been displayed on specific ritual occasions. Although the hall of statues is in keeping with the exoteric image hall, new Esoteric deities (statues of Butsugen butsumo Buddha and Nyoirin Kannon and the painting of Vairocana Buddha) are placed with exoteric favorites such as Yakushi and Miroku. As is the case for the Nyōhōdō, there is no single textual source for the iconography of the deities on the lecture hall altar. And again, it may not be fruitful—or even appropriate—to seek an iconographic source. Rather, it may prove most informative to be guided by knowledge of Kūkai's liturgical interests and general principles of Esoteric doctrine. The icon and its history are mutually constituting. The temple is best appreciated in light of practice, patronage, and the relationship of Kanshinji and its founders to events of the early Heian period.

Historical inquiry of this type may help answer the following questions: What is the meaning of the original lecture hall assembly? When and why was the Nyoirin Kannon statue placed in the zushi as a secret image? When did it become the honzon of the hall and the temple? What is the history of the two extant Buddha images? When were they removed from the lecture hall, and why? Similarly, what has become of the other statues and paintings listed in the Register, and can the changes that took place on the lecture hall altar be traced or documented?

This essay began with a discussion of the setting as it is today (Figs. 3–5): the Nyoirin Kannon is secreted by itself as the honzon of the main hall, a structure that did not exist in the ninth-century temple complex, behind the double doors of a black-lacquered shrine raised high on a three-by-one-bay altar with an elaborate canopied ritual altar below and before it at center. It is not until the fourteenth century that a zushi shrine for the Nyoirin Kannon is mentioned in temple documents. A photograph of the main hall from about 1950 (Fig. 3) shows two zushi to either side of that of the Nyoirin Kannon. The doors at right conceal a fourteenth-century wooden Fudō myōō (Sanskrit: Acalanātha) revered by Emperor Godaigo (r. 1818–39); behind the left doors is a fourteenth-century wooden Aizen myōō statue revered by Emperor Gomurakami (r. 1339–68). According to tradition these two images were made by Kūkai, and they gained importance during the Nanbokuchō period (1336–92) due to their close associations with the two emperors.[75] Both were added to the altar some time after 1439, the date of the current main hall. All three statues, according to temple tradition, were made by Kūkai's hands. Arranged across the raised altar in front of the three closed zushi are four guardian statues (Shitennō) dating to the tenth century.[76] Today a pair of Diamond and Womb World mandala paintings are permanently displayed on a panel between two pillars, one to each side, one bay south of the altar, with a ritual altar (dan) before each mandala (the Diamond World mandala is visible in Fig. 3). Neither the two extant ninth-century Buddha statues nor that of the Chinese priest (Fig. 15) are worshiped in the main hall; they are now displayed in the temple's museum (treasure hall).

Today's designated horizon, scholars agree, corresponds to the polychromed wooden Nyoirin Kannon statue described in the 883 Register, judging by correspondences in size, style, technical features, and preservation. Scholarly opinion differs on the iconographic sources and liturgical function of the statues inventoried in the lecture hall. Stylistic appraisals prompt many scholars to date both extant Buddhas to the second half of the ninth century, ten to twenty-five years later than the Nyoirin Kannon.[77] Debate arises in part from strategies that give preference to the honzon and in part from the inconclusive meaning of the order in which the icons are listed in the lecture hall section of the inventory. Such lists ordinarily rank statues and paintings in a canonical hierarchy, beginning with icons of the Buddha class, followed by hodhisattva, deva, and other lesser deities. It is not certain in the case of the Kanshinji Register whether the order indicates the horizon of the respective hall in first position or, alternatively, a hierarchy of ritual or other ranking, because there are no comparable inventories for the early ninth century.[78] In eighth-century inventories, for example, those for Saidaiji and Tōdaiji, the order of statues is typically hierarchical: the horizon is listed in first position, followed by attendant images and lesser deities. Japanese art historians typically cite the lack of contemporary inventories when explaining why the Nyoirin Kannon—their honzon of choice—is listed fourth in the Register, but given the protocol of eighth-century inventories, one might just as readily assume that the fist-listed Butsugen butsumo nyorai (Buddha) statue was the honzon. This point will be taken up again below.

In the case of the lecture ball, it is likely that the first-listed Buddha, the Butsugen butsumo, was intended as the main image or was part of an Esoteric grouping that did not feature a single image as honzon. It is also possible that the original honzon was the Miroku or Yakushi Buddha, listed second and third, respectively. One possible arrangement would comprise the first three Buddhas, with one at center as the honzon. The Nyoirin Kannon may have been one of a sculptural mandala, an attendant statue (waki-zō), or—most likely—a devotional image. It may have had a distinct function and status as a secret image without serving as horizon. It may have become a secret icon later in the temple's history. Other arrangements are also plausible. Despite a wealth of possibilities, most scholars and devotees alike believe that the Nyoirin Kannon alone was the honzon for the original lecture hall, and forr the monastery of Kanshinji.

Japanese art historians, save one, have all maintained that the Nyoirin Kannon was created before other statues in the lecture hall and that it was the horizon. This is refutable in light of its position in the inventory, the lack of similar horizon precedents, and the lack of supporting documentation—issues that would in many respects render it the least likely choice for a central icon.[79] Little mention is made in the literature concerning statues no longer extant, such as the Yakushi Buddha listed third in the Register. Advocates for the Nyoirin Kannon as honzon rest their claims on alleged corroboration between documents and aesthetic evahmtion, relying strongly on the latter to satisfy gaps in the documentation. It may well be true that the statue so esteemed today and in medieval times was the original honzon of the lecture hall altar. Its fine pigmentation and construction are notable. But appraisals of the only other extant Japanese works from the ninth century, the two Buddhas, are overshadowed by a range of extravagant claims about the Nyoirin Kannon. Although scholars agree that the two extant Buddhas correspond to those listed in first and second position in the 883 Register, their stylistic assessments typically date the two statues to a decade or more atier the Nyoirin Kannon, from about 850 to as late as the tenth century.[80] Yet there is no proof to support this position. The following discussion of technical and artistic features, in tandem with proposals for the disposition of icons on the altar, lays bare the hyperbole that too often accompanies analyses of the favorite icon.

The Nyoirin Kannon statue is 43 inches (109.4 centimeters) tall,[81] a height almost doubled by a lotus pedestal 42 5/8 inches (108.5 centimeters) high. The icon is backed by a striking red flame-ringed gold mandoria measuring 55 1/8 inches (140.3 centimeters) tall (Fig. 1).[82] Both the pedestal and mandoria are made of Japanese cypress (hinoki); the figure may also be made of cypress, although—in the absence of conclusive testing—many scholars assume that it is formed from kaya, a more precious wood.[83] The figure is constructed in the so-called single-woodblock technique (ichiboku zukuri) characteristic of the early Heian period (794–1185); the Nyoirin Kannon's head and main portions of the torso are made from a single piece of wood.[84] The statue is hollowed (uchiguri) at the core from three points (rear of head, center of back, and base). As was typical of wooden statues constructed during the ninth century in workshops linked to state sponsorship, the statue and pedestal are covered with a “dry lacquer” (kanshitsu) coating 2 to 9 millimeters thick, topped with primer, then decorated with pigments (saishiki) and, in the case of the drapery, cut gold-leaf (kirikane) patterns (Fig. 6). The exceptionally well-preserved motifs adorning the bodhisattva's draperies and the sequenced color patterns on the pedestal's lotus petals are among the finest examples of early Heian-period decoration (Fig. 7).[85]

The Butsugen butsumo and the Miroku statues are nearly identical in style and construction. Like the Nyoirin Kannon statue, the two Buddha figures are roughly three-foot(one-meter-) high figures made of hollowed wood and coated with lacquer (Figs. 10, 17). Both are constructed of Japanese cypress. The height of the Nyoirin Kannon is nearly identical to the height of the two extant Buddhas (Butsugen butsumo, 421/2 inches [108 centimeters], Figs. 8, 9, 17; Miroku, 43 inches [109.5 centimeters], Figs. 10, 11). The daises of the Buddhas (333/8 and 33 inches [84.5 and 83.6 centimeters], respectively), however, are about 10 inches smaller than that of the Nyoirin Kannon.[86] A visual comparison shows that all three figures have similar upper-body sashes (jōhaku) and probably wore similar crowns, suggesting a common date and workshop. It might also point to a triadic arrangement of the three images, or possibly, with the inclusion of the lost Yakushi Buddha icon of unknown dimensions, an arrangement of four images.

Konno Toshifumi recently proposed that the three extant works originally formed a triad following an iconographic conception of the three as transformational bodies (hengeshin) of the Dainichi Buddha, the central icon for the Shingon tradition.[87] Although a triad arrangement of images seems possible for images of this size, the iconographic basis he offers is highly unusual and the proposal seems forced. Moreover, like so many other researchers, Konno completely disregards lost works listed by the Register. Certainly the (lost) Yakushi Buddha listed third in the Register, before the Nyoirin Kannon, had an important place on the altar, as did the Chinese portrait statue, the pair of deva, and the Bishamontennō statues. Modern hypotheses that do not so much as speculate on the possible import of lost ninth-century works to the lecture hall altar are inherently flawed.

The Nyoirin Kannon's colorful designs and cut gold-leaf patterns have been carefully studied and roundly praised in the literature. It is easily forgotten, however, that the renown of the decoration arises largely from its extraordinary preservation, which is an effect of the statue's concealment as much as any intrinsic value. In comparing the relative quality of the Nyoirin Kannon bodhisattva and the two extant Buddhas, researchers stress that the application of colors and cut gold leaf on the Nyoirin Kannon constitute a highly labor-intensive process, while the Buddhas are decorated with gold leaf alone.[88] Mizuno Keizaburō and Nishikawa maintain that the quality of the Nyoirin Kannon's decoration was possible only with the support of a wealthy, probably imperial, patron.[89]

Nishikawa's research on the Nyoirin Kannon integrates stylistic and documentary analysis. He posits an early date for the Nyoirin Kannon, and sees it as followed by the creation of the two Buddhas on the occasion of Kanshinji becoming a state-supported jōgakuji in 869. He sees the Nyoirin Kannon as an accomplished work completed in the mid-ninth century, and he believes that its quality implies imperial patronage. Nishikawa further promotes the work to Kanshinji's main icon from the origins of the monastery. He cites an entry in the Register for Jōgan 16 (874).7.9 that records a land bequest by Dowager Empress (widow) Junna (809–879)[90] to Kanshinji for the repair of a worship hall referred to as “Saga-in Taikō taigō gogandō,” or “hall made according to the vow of Dowager Empress Saga.”[91] Empress Junna was Princess Seishi, daughter of Emperor Saga (786–842, r. 809–23), who became a nun after the death of her husband (and Saga's brother), Emperor Junna (786–840, r. 823–33). Empress Saga was Tachibana no Kachiko (785–850), the primary wife of Emperor Saga, who was the first imperial sponsor of the Shingon teachings and Kūkai's greatest patron. The fact that some time before her death in 850 Empress Saga dedicated a hall at Kanshinji, and that its upkeep was provided for in 874 by the subsequent Dowager Empress Junna imparts specific meanings to the monastery as a whole, as well as to the hall she endowed. Note, however, that neither of the two references to the endowed hall names which Kanshinji building received the imperial largesse. Nishikawa, equipped with the patronage record for an unnamed hall, points to the artistic and technical superiority of the Nyoirin Kannon and concludes, in a “chronological reversal,”[92] that the lecture hall received the dowager empress's patronage; in addition, he asserts that the Nyoirin Kannon statue in particular profited. In a metaleptic inversion, the effect (the dowager empress's largesse) is offered as proof of the cause (the alleged fact of the Nyoirin gannon's aesthetic and material value). Nishikawa, considering the hall endowed by the dowager empress to be the lecture hall, conjectures that the statue was made before Kashō 3 (850), the year of Dowager Empress Saga's death, probably around Jōwa 9 (842), when Emperor Saga died.

That the gilt-wood Buddhas may have been the object of similar largesse should not be overlooked. All three statues use a layer of lacquer between the wood and the exterior coat of gold leaf or pigment.[93] The extraordinary expense and skill required to construct images employing lacquer meant that only temples of notable means or patronage housed such works. Lacquer use for statues had declined by the mid-ninth century largely due to its cost; thus, lacquer application on the three surviving Kanshinji examples under discussion suggests that considerable expense was involved in their creation. It also links these works to artists trained in a state-supported atelier in the capital, or to the workshops themselves.[94] The style of all three statues has many affinities with the Esoteric statues made for the state temple Tōji about 839 (Figs. 16, 18, 24). At the same time, they resemble Chinese Esoteric works from the eighth century, a style that was conveyed through paintings, drawings, and small statues brought to Japan by Kūkai, among other ways. Despite these similarities, contemporary scholars judge the Butsugen butsumo and Miroku Buddhas to be artistically and technically inferior to the Nyoirin gamnon.

Furthermore, the labor involved in production and decoration was not always a criterion for the liturgical status of the work and certainly makes a tenuous substantiation: in many temple halls may be found a horizon (both Buddha and bodhisattva statues) decorated with gold leaf sharing the altar with finely polychromed statues, such as at the Tōdaiji Hokkedō or Tōji (Fig. 16). It may also be noted that surface decoration was largely determined by iconographic requirements. A gold surface was standard for Buddha images, as it counts among the most significant of his (Sanskrit) lakshanas, or auspicious physical marks. For this reason, polychromed Buddha examples are rare. Similarly, it is typical for bodhisattva statues and those of dena or lesser deities to be polychromed. The statue of Gōzanze myōō bosatsu of about 839 (Fig. 18) or the Shitennō, both at the Esoteric monastery of Tōji, are but two such examples. The Nyoirin Kannon is described as decorated by saishiki, or polychrome, by the Kanshinji Register, which also notes the lecture hall guardian statue, Bishamontennō, as having saishiki decoration. In addition, the patron for the guardian statue is noted, as it is for the Yakushi image, distinguishing otherwise unrelated images.

Given the notations of patronage for the Yakushi and Bishamontennō statues as well as the fact that both were important icon types in the early Heian period, these lost statues from the lecture hall warrant special attention. It should not be glossed over that no patron is noted for the Nyoirin Kannon statue by the Register inventory. If imperial patronage had been conferred on the Nyoirin Kannon, as modern researchers hypothesize, it would almost certainly have been noted by the Register. Scholars do not address this discrepancy in their discussions of patronage. Neither does the Register indicate that the statue is a hidden image, casting this possibility into doubt as well. The material (wood) and size (3 shaku) of the statue are provided by notations, which might suggest that these traits were not widely known because it was a hidden image, but the notations could have been made for other reasons. Lastly, there are many cases in which a particular statue was singled out by a patron for special production or treatment and was worshiped on particular occasions, but it was not necessarily the honzon of the altar or the temple. Even if the Nyoirin Kannon was the recipient of imperial largesse, this points to the statue's efficacy or the patron's affinity with the deity and does not necessarily indicate the icon's status at the temple. Thus, despite widespread assumptions to the contrary in the literature, there is little intrinsic or comparative evidence to support the Nyoirin Kannon as horizon during the temple's early period.

Yakushi Buddha was widely worshiped in Japan from the seventh century for its curative and magical properties. There exist both exoteric and Esoteric versions of sutras about Yakushi. The Kanshinji work, listed third in the Register, is noted as “danjiki,” that is, sandalwood—the material prescribed by several Yakushi texts. This aromatic and highly esteemed wood was, however, not native to Japan and China, so in these contexts the term danjiki typically refers to wood (and a mode of representation) that emulates sandalwood figures (Fig. 19 shows an early ninth-century example of an Esoteric form of Yakushi from Shinyakushiji). If the Yakushi statue were a Chinese work made of imported sandalwood, a notation would likely have described it as Chinese, as is the case for the Tang priest's portrait. Because the Register makes note of the missing mandorla of the Yakushi statue, this strongly suggests that the statue is an older icon, that is, a “guest” icon brought from another temple or already installed elsewhere at Kanshinji during its earliest history. This likelihood is supported by the mention ofYakuman, a donor, which provides further “history”—presumably to a knowledgeable reader of the time. The image is not described as a Seven-Buddha Yakushi, that is, the Esoteric form of the deity, so it was probably a non-Esoteric type. The incorporation of a preexisting or guest (non-Esoteric) Yakushi Buddha into Esoteric altar assemblies or Esoteric monastic contexts is seen at a number of Heian-period temples (the earliest among them are Jingoji and Tōji), where such images were typically in place before the temple became affiliated with the Esoteric Shingon sect.[95] When present in Shingon halls, Yakushi would be part of a mandala-like grouping (as at Zenrinji, Kyoto). The Yakushi at ninth-century Kanshinji may have been part of an altar group and not the main image. In light of the historical context (to be elaborated in the next section), it is possible that the Yakushi image was made as a curative vow for the health of an imperial patron. The Yakushi Buddha is also associated with relics, and typically holds a reliquary jar of the sort that connotes healing.[96] In any case, since the statue is lost and we know nothing of its visual character, it is imprudent to attempt a definitive hypothesis. At the same time, its possible implications for the original altar should not be overlooked.

Nishikawa and Mizuno propose additional evidence to bolster the worth of the Nyoirin Kannon. The contents of the temple storehouse as given by the 883 Register indicate a majority of ritual goods such as urns, bowls, censers, and other altar implements. The authors call attention to five gilt vessels and two white Tang Chinese porcelain bowls (both donated by the Buddhist master Eshuku), which are identified as objects “for the Kannon bosatsu.”[97] They assume this to mean the Nyoirin Kannon, although the Nyoirin Kannon is not the only Kannon represented at the temple: there are a Shō Kannon painting in the lecture hall inventory and the Nyoirin Kannon paintings, among others, in the Nyōhōdō; also, the noun Kannon bosatsu may be singular or plural. An additional notation lists one bowl in a set of ten for “the Kannon(s),” with the remaining eight for “the ritual hall” (Hōdō) and one “for incense.”

In other Register entries ritual goods are specified for use with the Chinese portrait statue (extant) and the pair of deva statues (now lost) on the same altar. It is suggestive that although these statues are virtually impossible candidates for the honzon or high canonical rank, precious ritual goods are also earmarked for them. Reference to the Kannon offerings are given by the authors as further proof that the Nyoirin Kannon must have been the horizon of the lecture hall by 883, when the Register was compiled.[98] In fact, the evidence offered does not indicate that the Nyoirin Kannon is the honzon so much as that particular goods and—in all probability—particular ceremonies were reserved for a “Kannon(s).” If the Nyoirin Kannon was a secret image by 883, special ceremonies would normally have marked its annual display or celebration; ritual goods would also be mentioned. It is also possible that the urns and bowls were donated by the priest Eshuku for particular purposes (such as rituals for healing) in the years or decades after the statue was made, but the existence of such goods is not proof of the Nyoirin Kannon statue's or any Kannon image's function for the temple, its status as a secret image, or its canonical rank on an altar.

Let us return to a consideration of the Butsugen butsumo Buddha statue in the early Kanshinji context. First, the Butsugen butsumo and Miroku statues are listed in first and second place in the 883 Register list, indicating possible iconographic primacy on the altar. As noted above, although the order of deities in the Register is not conclusive as regards the status of the images, eighth-century precedents suggest that the image(s) given first on a list are typically the main images. If the Register list merely indicated rank according to canonical deity type (Buddha first, then bodhisattvas, and so on), then the Birushana Buddha painting in the middle of the list should be nearer the beginning, following the three Buddhas listed in first, second, and third positions. Second, the Register literally represents a salient clue to the Butsugen butsumo's importance. Beneath the text and scattered across the surface of the paper, the official Kanshinji temple seal is stamped in red ink—about six hundred seals over seven scrolls (Fig. 13). The seal consists of two characters in a square (shown enlarged in Fig. 20, the red appearing as gray in the reproduction).[99]

The temple name, Kanshinji, comprises three Sino-Japanese characters, pronounced kan, shin, and ji. In their regular forms, the characters mean “temple (ji) of contemplation (or visualization, kan) [and] mind (shin).” The characters might also mean aspiration for enlightenment, bodhicitta, that is, “temple (ji) of bodhi (kan) citta (shin).” The temple seal imprinted across the pages of the Register uses the first two characters only and modifies the second of these. It shows the usual character (kan) in the first place, but for the second character (shin), a Sanskrit letter (Siddham), sri, is used in its place. The temple name is thus rendered by the seal as “Kan sri” (illustrated in standard type in Fig. 21). In Buddhist terminology sri means felicity or good luck (Japanese: kisshō or kōfuku). Siddham letters have special meaning and use in the Shingon Esoteric tradition that are key to understanding the seal.

Sanskrit is often used in Shingon ritual texts, and the teachings draw from many Indian elements. Sanskrit sounds and letters are used in mantra and letters are used as symbolic referents (or samaya forms) for deities, called shuji (“seed syllables”; Sanskrit: bija). The Esoteric symbolism of the Siddham sri, the letter used in the Kanshinji seal, is one of several shuji for the deity Butsugen butsumo Buddha.[100] Nishikawa and Mizuno mention the Esoteric meaning of the seal with the gloss, “there is a Butsugen butsumo statue in the Lecture Hall, suggesting that sri connotes felicitous meanings.”[101] I believe that the symbolism of the seal indicates a much more consequential relationship between the Butsugen butsumo Buddha statue/deity and Kanshinji temple than this.

To begin with, the first element of the seal, kan, means contemplation or visualization, as noted above. Kan etymologically corresponds to the Sanskrit letter loc, and loc is the root syllable for the Buddha Locanā, the equivalent of Butsugen butsumo Buddha in the Shingon Esoteric tradition. In short, the official ninth-century seal for Kanshinji should be understood as a double referent to the Butsugen butsumo Buddha. Additionally, the combination kan sri readily suggests a parallel reading of “meditation [or visualization] [102] on the Butsugen butsumo Buddha (Locanā).” The second letter is in Siddham script, which is the samaya form of the deity and the locus of meditation on the deity for initiates in visualization rituals, amplifying the resonance visually, virtually, and symbolically.

It is important to remember that within Shingon, as Kūkai explained in his essays, language—especially phonetic systems—effectively manifests the ultimate truth of emptiness.[103] The deity Butsugen butsumo Buddha has a primary role in the Shidō kegyō shidai (Precepts of the Four Stages of Prayoga) rituals central to Esoteric training.[104] In the rites of empowerment (kaji) within the training are the three central empowerments of body, word, and mind. In the second and fourth rites of the Shidō kegyō shidai a section called Butsugen butsumo empowerment appears before the Nyūga ganyū, Shonen-ju, and Jirinkan sections and helps the practitioner perfect his or her practice. As has been noted, Butsugen butsumo also has a central position in the Womb World mandala and is understood therein as the Buddha from which all Buddhas come, literally, the “mother.” This deity would have been important to the training aspect in Womb World rites (and remains so today). The Buddha Locanā is identified in the Mahāvairocana Sutra as vidyā-rājñā, the queen of Wisdom, consort of Mahāvairocana, and personification of prajñā (wisdom) as language—mantra.[105] Visualizing (the mind aspect of practice, eidetic meditation) Locanā (the mantra or language aspect of practice) is thus another possible interpretation of the temple seal. Language and mind are two of three essential components of ritual: gestures (mudra), language or incantations (especially mantra), and mind (kan, meditation). These three fundamental ritual expressions find their correspondences in the “three mysteries” (Sanskrit: trighuya; Japanese: sanmitsu) of a divinity. The three mysteries are understood as the means by which the Dharmakāya Buddha, Mahāvairocana, reveals his innermost enlightenment to the practitioner.

The Buddha Miroku, listed just after Butsugen butsumo, may represent the exoteric teachings and salvation in the future. Yakushi Buddha signals healing, and the Nyoirin Kannon, next in the list, embodies the power of the nyoiju jewel—associated with relics and the symbol of the Buddha Locanā, the proposed patron deity. This grouping of three strongly suggests Kūkai's faith in the mother of Buddhas, the Healing Buddha, and the Future Buddha, and his disciples' understanding of Kūkai's faith and promise to aid them after his death as an avatar of Miroku. The painting of Mahāvairocana (Birushana), listed next in the Register, may well be the ritual complement to the statue of Butsugen butsumo, Mahāvairocana's consort. Unlike most of the literature on Kanshinji, these interpretations consider contemporaneous practice and meaning in the Shingon tradition in analyses of the altar and the temple seal.[106]

The icon constitutes “the real” in a particular cultural structure, but it has value only as long as it remains viable (or present) within that structure. Icons are situated in a type of life cycle within the production-circulation context. When elements of that cycle are lost to history, analysis can super-impose its own cycle of commodity circulation.[107] We analyze both what we see and what we reconstruct, using tools and epistemologies we deem appropriate. The ninth-century icons designated by the Register have been analyzed thus far according to their iconography and symbolism; in the context of ritual function; in light of the meaning of the Kanshinji seal; according to known material features; in juxtaposition with a study of patronage; and against the contents of a body of literature that favors the Nyoirin Kannon as honzon. All these areas of inquiry indicate that the Butsugen butsumo Buddha is a likely candidate for the honzon of the image hall—or the temple—in its early history. This investigation further points to the function of the Nyoirin Kannon as the focus of individual rituals for the divinity in both the Nyōhōdō and the lecture hall (possibly for different reasons) by the time the Register was composed in 883. It also suggests that the Nyoirin Kannon was worshiped as a secret image by the eleventh century or, along with other newly created or newly secreted icons, by the fourteenth century. It is apparent that the deity Nyoirin Kannon was important to the Kanshinji priests and patrons during the ninth century. What is also evident is that other icons had a dynamic and primary role in the early history of the monastery. The following section examines the relationship of presumed imperial patronage for the Nyoirin Kannon and the honzon status of the Butsugen butsumo.

The following chronology of events may be ascertained from government and temple records for the period and is helpful to any consideration of the icons. Emperor Saga retired from the throne in 823 at the age of thirty-eight, the same year he awarded Kūkai the headship of the monastery Tōji for use as a Shingon training center in the Heian capital. His wife, Empress Saga, remained at court during the reign of Saga's brother, Junna, from 823 to 833. When Emperor Junna stepped down from the throne in 833, at the age of forty-seven, the son of the retired Emperor and Empress Saga ascended as Emperor Nimmyō (810–850, r. 833–50) at the age of twenty-three. In 842, during Nimmyō's reign, the retired Emperor Saga died at the age of fifty-seven. In the third month of the year 850, the reigning Emperor Nimmyō died at the age of forty. At that time Nimmyō's mother, Dowager Empress Saga, took the tonsure and entered a temple. She passed away less than two months later at the age of sixty-five. Each of these imperial court members was a practitioner and patron of Shingon.

As noted above, Nishikawa cites an entry for 874 in the Register of a land bequest to Kanshinji by Dowager Empress Junna specifically for the repair of a worship hall “made according to the vow of Dowager Empress Saga.” Nishikawa, the leading scholar for the history of Kanshinji, hypothesizes that the donation of funds by Dowager Empress Saga was prompted by the illness of her husband, retired Emperor Saga, around 839.[108] The date of the dowager empress's donation is not recorded; we know only that it was before her death in 850. The donation of funds to temples by members of the imperial family during this period were regarded as a means to secure religious merit, restore health, or ensure the safe passage to the Buddhist paradise of a deceased relative. Nishikawa selects the date of 839 because it supports already established conclusions about the Nyoirin Kannon. He concludes that the lecture hall and the Nyoirin Kannon were completed by the year 839 with the support of Dowager Empress Saga; he cites the record requesting that a bell be cast (seventh month, 840) as additional evidence. Having constructed an argument for the Nyoirin Kannon as the locus of imperial patronage in which canonization plays a supporting role, he suggests stylistic traits to support his conclusion that the two Buddhas were produced later, in about 850, unconnected to imperial support.

One might well argue that the strong female iconographic nuances of the Butsugen butsumo makes it an ideal candidate for the empress's patronage.[109] A second domain of doubt around Nishikawa's hypothesis is that there is no reason to assume that the dowager empress gave funds for the lecture hall rather than the Nyōhōdō. Given the strong female iconography and ritual elements present in the Nyōhōdō, including the Womb World mandala; links between the Godaison and activities at the imperial palace; the nyoiju jewel elements inherent in representations of the Five Great Storehouse bodhisattvas and the Nyoirin Kannon; and the association of the jewel with imperial relics (to be described below); in addition to the proposed function of the Nyōhōdō as a ritual initiation hall, that structure seems a stronger possibility for Dowager Empress Saga's patronage. A different interpretation of the record is that the hall in question was dedicated to the dowager empress rather than by. her, for which a likely date would be just after her death in 850.[110] Nishikawa's appraisal (and others that followed from it) arranges history on the drawing board to preserve the status of the Nyoirin Kannon from ancient to modern times as though history safeguards only the finest works of art.

That Nishikawa ascribes the date 839 to the dowager empress's gift limits the possibilities. If the funds were designated for the construction of a hall (and, presumably, its icons), it is unlikely that they could have been completed only a year later (recall that a completion date of 840 is strongly supported by a request to cast a bell that year). The construction of temples in mountainous regions was generally hampered by geographic and physical conditions, as was the case at nearby Kongōbuji, Mt. Kōya, even during Kūkai's lifetime. Moreover, contemporaneous Shingon monasteries both in the capital (Tōji) and beyond are on record as having developed slowly due to lack of financial support or official recognition.

Even if one allows that the Nyoirin Kannon statue was made around the year 839 to aid Emperor Saga's recovery from illness, this does not guarantee its status as honzon, nor does it preclude the possibility that other statues were simultaneously completed and worshiped. It is in fact difficult to imagine a “completed” hall lacking any of the three Buddhas listed first in the 883 Register inventory. If the funds were directed toward the lecture hall, the Yakushi statue recorded in the Register, now lost, would have been a more likely target for vows to cure the retired emperor than the Nyoirin Kannon. Such devotions had numerous precedents during the period.[111] The Miroku Buddha, Buddha of the Future, listed second in the Register before the Yakushi, could similarly suggest a vow for Emperor Saga or Nimmyō's after life in the Buddhist paradise, or for the aged empress herself. That Dowager Empress Saga and her son the reigning emperor Nimmyō both died early in the year 850 may also have prompted the placement or creation of devotional images. If the hall named in the Register was dedicated by Dowager Empress Saga, it may have been in the third month of 850, when Emperor Nimmyō died (and just before she passed away in the fifth month). These events and beliefs pertaining to Buddhist deities are valuable historical clues. There is no reason to compromise them for the sake of elevating the Nyoirin Kannon statue alone.

As I have stressed elsewhere, if the three original Buddha statues made for the altar did not already have special status (iconographic, aesthetic, liturgical, or other) at Kanshinji by the time the Register was completed in 883, but the Nyoirin Kannon was a honzon, as scholars assert, then one might expect the record to list the Nyoirin Kannon first. If, as Nishikawa maintains, Dowager Empress Saga had directed her gifts specifically toward the Nyoirin Kannon within the lecture hall by the year 850, followed by an endowment for the upkeep of the hall from Dowager Empress Junna in 874, temple records would in usual cases make specific note of such extensive patronage for an image or a hall, which they do not. In fact, lacking proof to the contrary, the Nyōhōdō, as the main hall of the Esoteric compound, would be the presumed recipient of patronage and would thus not necessarily require special mention in the Register other than the appellation “gogandō,” or “hall made by a vow.”

It might seem unlikely that a temple as remote as Kanshinji would be the recipient of imperial favor. The connection probably can be explained by the following. In 841, just a year before retired Emperor Saga died, Empress Saga received a lay Esoteric initiation (kechien kanjō) at Tōji from Jitsue, who was Kūkai's favored disciple as well as the teacher of Shinshō (Kanshinji's thunder) and a priest himself intimately connected to the early history of Kanshinji.[112] A year before Empress Saga's Esoteric initiation, we also find the record for the casting of the Kanshinji bell (840.7.27). It was during this same period that Jitsue was active at Mt. Kōya, anti—according to Shinshō's testimony—at Kanshinji. The empress may easily have become aware of Kanshinji through Jitsue, and she may have known Shinshō. This history supports my proposal that Kanshinji received the empress's patronage after Emperor Saga's illness was recorded (839.8.4), perhaps after her initiation. There is a good possibility that she funded the Nyōhōdō if it functioned as Kanshinji's initiation hall, in memory of her own conferral at Tōji.

Additional evidence links Kanshinji to Dowager Empress Saga through her son Emperor Nimmyō. According to entries in the chronicle Sandaijitsuroku, the construction of a set of five new Buddha statues to be dedicated to Kanshinji began under the direction of Shinshō in the year 854 (Saikō 1). Although neither the reason nor the sponsor for this project is clearly indicated, the document tells us that Shinshō's vow included these words: “I ask for the blessing of the late Emperor Nimmyō, whose virtue I am too ignorant and insignificant to appreciate.”[113] Nimmyō died in 850, the five Buddhas were begun in 854, and they were completed three years later. Itō Shirō believes, and I concur, that the record strongly suggests that the five new statues were commissioned for Emperor Nimmyō's reincarnation in paradise four years after his death.[114]

The evidence presented above, namely, the connections between Nimmyō's mother, Jitsue, and Kanshinji, in tandem with the dowager empress's sponsorship of a hall at the temple, strengthens the idea that these five statues were commissioned for the temple to honor and protect Nimmyō. An entry in the records of Emperor Montoku (r. 850–58) points further to the association between the court and Kanshinji. The document notes that on the seventh day after Emperor Nimmyō's death, Kasho 3 (850).3.27, envoys were dispatched to “the six temples connected to the sovereign.”[115] The envoy Michi-no-ō was sent to Hinoo-dera, the local name for Kanshinji. These links between Nimmyō and Kanshinji, indicated by the dedication of the Five Wisdom Buddhas and the envoy, suggest that Dowager Empress Saga's bequest for a hall at Kanshinji could have come at the time of Nimmyō's death or during a period of indisposition, placing the dedication of the “Saga-in Taikō taigō gogandō” hall in the same period, that is, the late 840s or early 850s.[116] The association between Kanshinji and the imperial palace relics rites during the fourteenth century described above may also find its origin in these earlier centuries.

The iconography of the five Buddhas made for Kanshinji under Shinshō's direction is germane to the present discussion. The Five Wisdom Buddhas, with Birushana (Dainichi) Buddha at center, are the central motif of the Diamond World mandala. This deity group is understood as the Diamond World and partner to the Womb World, the latter being the prevalent iconography of the Nyōhōdō. The impetus for an altar or hall featuring Diamond World imagery would not be surprising given the curious absence of a Diamond World mandala painting in the Nyōhōdō. Ritual halls for the Five Wisdom Buddhas, called Godaidō, were found at Esoteric temples from the ninth century. It is quite likely that a plan was envisioned (probably by Shinshō) to pair two halls at Kanshinji: the Nyōhōdō and a new hall for the Five Wisdom Buddhas, as manifestations of the Womb and Diamond Worlds, respectively.[117]

The five new statues, however, never realized their place in the proposed plan. Records state that the Five Wisdom Buddhas were made at Kanshinji; after their completion in 857, Shinshō had them shipped to Zenrinji, a Kyoto monastery built under Shinshō's direction in 853 on the estate of Fujiwara Sekio in the Higashiyama district of the capital.[118] If the plan for Kanshinji included a Godaidō hall to balance the iconography of the Womb World mandala—and the female nature of the proposed horizon Butsugen butsumo—and if that plan was not realized, then it is logical that a shift occurred over time in the liturgical and symbolic importance of the honzon proposed here, the Butsugen butsumo, to favor the Nyoirin Kannon.

It seems natural to question whether the death of Dowager Empress Saga in 850 contributed to changes at Kanshinji. An examination of land bequests to the temple in the Register shows that the peak of (recorded) support for Kanshinji occurred in the late 860s. In the seventh month of 874, Dowager Empress Junna made her land bequest to Kanshinji to pay for repairs to a hall previously dedicated by Dowager Empress Saga, as discussed above. By that time, judging from records of land grants and other contributions to the monastery, support for Kanshinji had declined. No scholar has as yet suggested that Dowager Empress Junna's land grant in 874 may have been made in honor of the priest Shinshō, abbot (bettō) of Kanshinji, who had passed away one year and two days before (in 873.7.7).[119] Given the importance of Shinshō's relations to the temple, and the established practice of marking the first year after death, such a proposal is reasonable.

With the group of Five Wisdom Buddha images intended for Kanshinji installed at Zenrinji from 857 and the Esoteric master Shinshō deceased, the iconography and symbolism of the Nyoirin Kannon may have taken on increasing importance to the activities at Kanshinji. It is also possible that the Nyoirin Kannon came to be associated with Shinshō after his death in 873. Paintings were frequently added to altar assemblies over time. The Register lists two Nyoirin Kannon paintings in the Nyōhōdō by 883. One possibility is that they were created for the hall after the Nyoirin Kannon in the lecture hall achieved greater prominence and as further complement to the female Womb World iconography of the hall. The Five Wisdom Buddha statues' relocation to Zenrinji in 857 opened an avenue for new meanings and icons at the Nyōhōdō, or the highlighting of existing trends. Thus, rather than serving to bolster the probability of the Nyoirin Kannon deity (and statue) as honzon through their iconographic affinity, the creation of two Nyoirin Kannon paintings for the Nyōhōdō may have been inspired by the increasing importance of the Nyoirin Kannon in the lecture hall context decades after it was made.

How would the Nyoirin Kannon have come to replace the Butsugen butsumo Buddha as honzon? Many art historians would point to its beauty as the answer. But even here we are on unstable ground. I have suggested that changes in worship practices and patronage at the temple contributed to its altered status. Below I will again consider the evolving context at Kanshinji, one in which the Nyoirin Kannon gained value and significance, offering evidence for a more complex nexus of meanings at the monastery in the ninth century and thereafter.

The name Nyoirin Kannon, literally, “Kannon of the Wish-granting Jewel and Wheel,” refers to two of the image's hand-held attributes: nyoiju (or hōju), a jewel (Sanskrit: cintāmani that grants all desires, and rin, a wheel (Sanskrit: cakra) that symbolizes the Buddhist teachings (both attributes are clearly visible in Fig. 2).[120] It seems likely that the Nyoirin Kannon had a special role at Kanshinji, especially as a Kannon of the wish-fulfilling jewel. Because two paintings and a statue of the deity existed in the temple by 883, it can be reasonably assumed that there occurred rituals dedicated to the deity. At some point in time, the Butsugen butsumo ceased to be regarded as honzon. This probably happened during the thirteenth century. The earliest reliable record to describe the Nyoirin Kannon as a honzon and as a secret image is a pilgrimage journal of 1378 by the priest Kenki, the Kanshinji sankei shodō junreiki (A pilgrim's account of a visit to the various halls of Kanshinji).[121] The Pilgrim's Account states that the Nyoirin Kannon honzon was placed behind a three-bay span of doors within the five-by-five-bay chancel (that is, seven-bay square) Hondō (that is, Kondō, or main hall), a “recently rebuilt hall.” The priest also notes that the statue was “carved by the master” (Kūkai). He explains that few persons were allowed to see this sacred image in times past, but on the occasion of the recent rebuilding of the former “Konpon godō” (another name for a main hall), he was able to view it for an instant.[122] He describes the polychromed statue as larger than life-size; he also refers to it as the “seven-star Nyoirin Kannon” (a reference to Ursa Major, to be considered below).

The fourteenth-century Pilgrim's Account notes that the earthen platform altar (jidan) “was not disturbed” when the Konpon godō was rebuilt.[123] The ninth-century Register does not mention an earthen altar. From this we can deduce that an earthen altar was created between 883 and 1378. It is unclear when the original ninth-century lecture hall was lost, but according to the Pilgrim's Account of 1378, the hall had recently been rebuilt, suggesting that at least one other hall (the Konpon godō) was built after the original lecture hall and before the structure of about 1378. The latter structure was lost and rebuilt at least once more before the present main hall (Fig. 22), which dates to 1489 (Eikyō 11).[124] The configuration of the hall, that is, with a five-bay-square chancel, is similar to the original lecture hall, except that the present building now has a raidō, or forehall, added to the facade for ceremonial use. [125]

From this document we know that the Nyoirin Kannon was already a secret image for some unspecified period of time prior to 1378. The text also indicates that the altar arrangement and equipment were very different from the context described in 883 but similar to the modern altar. The Pilgrim's Account also notes that the Nyoirin Kannon was flanked by “a Fudō myōō statue” and a “hōsho [that is, dharma nature] [miniature] pagoda,” both behind the span of doors and both “made by Kūkai.”[126] The latter two works, neither extant today, are not listed anywhere in the Register of 883. They have been replaced on the altar with fourteenth-century images of a Fudō and its iconographic relation, Aizen myōō Incidentally, both fourteenth-century images flanking the Nyoirin Kannon today were allegedly “made by Kūkai” and, like the ancient Nyoirin Kannon, have been enclosed within individual zushi since the fourteenth century (as in Fig. 3).[127] This context further supports my proposition that images associated with Kūkai took on increasing importance at the temple from around the time of Kenki's visit in 1378. His journal also suggests that the Fudō deity and the pagoda were important to the fourteenth-century temple or the main hall liturgy; Fudō was widely worshiped in medieval Japan, and the pagoda was symbolic of a reliquary and associated with relics, as well as the dharma. Except for the Nyoirin Kannon, not one of the statues or paintings inventoried for the lecture hall in 883 are mentioned in the Pilgrim's Account. Most curiously absent are the two extant Butsugen and Miroku Buddhas.[128]

A survey of medieval records allows us to trace some changes in the meaning and location of the two Buddhas from the late fourteenth century. A three-storied “jeweled stūpa” (hōto; modern, Tahōtō) or three-storied pagoda[129] was built by the year 1360 and was restored three times before being destroyed by fire in 1462 (Kansei 3).[130] The pagoda housed two or four Buddha statues (four Buddhas would be the standard iconographic grouping). It is unclear whether both the ninth-century Butsugen butsumo and Miroku were housed within that pagoda, but they were definitely installed as two of the four directional Buddhas in the new so-called Tatekake pagoda, completed between 1462 and 1633 and extant today, along with a Shaka and Yakushi Buddha of later dates.[131] The iconographic identity of the Butsugen butsumo and Miroku Buddhas changed around this time as well, to Miroku and Hōshō nyorai, respectively (today they retain these names in popular worship). This identity transformation was probably a result of their installation in the pagoda with its own special iconographic plan.[132] Miroku Buddha's new name, Hōshō (Sanskrit: Ratnasambhava), “One Arisen from the Jewel,” refers to a deity in both the Womb and Diamond World mandalas. Hōshō is also the principal deity in the imperial palace Latter Seven-Day rite established by Kūkai, in which relics from Tōji are the ritual locus.[133]

It is clear from a number of changes that Kanshinji became strongly associated with rites for the imperial relics—which are in turn closely tied to nyoiju jewel worship—no later than the fourteenth century. According to legend Kūkai buried a jewel (cintāmani or nyoiju) given to him by his Chinese master, Huiguo, in the mountains of Muroo (alt. Murō), to the northwest of Kanshinji, where the temple Murooji became an important Shingon center during the early Heian period. The jewel and relics of Muroo figure largely in the Latter Seven-Day rite at the imperial palace by the eleventh century. The presiding abbot of the rite, dispatched from Kūkai's Shingon monastery in the Heian capital, Tōji, was required to carry out the central ritual offering as he meditated on Hōshō, the buried nyoi of Muroo, and the Buddha's relic on the great ritual altar.[134] Kanshinji soon became an extension of this imperial rite. An imperial order of 1860 indicates that the relics from Tōji temple were presented at the Kanshinji main hall, “jewel pagoda,” Karitei (Sanskrit: Hariti, a goddess) altar, and offered to the Benzaiten deva.[135] These developments suggest the changing context at Kanshinji during the Northern and Southern Courts period, Nanbokuchō (1336–92), when one of two emperors ruled from Yoshino, south of Kanshinji.[136]

Another factor in the changes undergone at Kanshinji is its association with constellar traditions. Ancient (and current) belief holds that Kūkai followed the Seven Stars of the Northern Dipper constellation (Big Bear or Ursa Major; Japanese: Hokuto shichisei) deep into the mountains one night and designated the area to which its stars pointed as the site for Kanshinji. No later than the fourteenth century, the Nyoirin Kannon statue at Kanshinji was called the Seven-star Nyoirin.[137] Modern art historians do not comment on these connections, but the significance of Ursa Major is apparent from the veneration of its astral deities in most if not all the ritual texts of the Chūin lineage (that is, the Shingon branch associated with Mt. Kōya). The Kakuzenshō, an important Esoteric ritual manual of about 1218,[138] features a long section on the worship of the Northern Dipper and quotes a full passage from the Beidou qixing yanming jing (The Seven Stars of the Northern Dipper Longevity sutra), an apocryphal chinese sutra,[139] or a sutra of probable Japanese invention, the Sutra of the Seven-Stars Cintāmanicakra's Secret Essentials.[140] Grapard notes that rites focusing on Ursa Major were prohibited by the government as early as 706 for a variety of reasons, and that from the middle of the ninth century documents indicate increased interest in astrology on the part of the government and aristocrats.[141]

There is also unexpected material evidence of preferred treatment of the Nyoirin Kannon. When the icon is viewed frontally, it is evident that its flame-ringed wooden mandorla rests slightly off center (Fig. 1). Close examination of the joinery, wear marks, and measurements of the mandorla for the Nyoirin Kannon confirm that the mandorla does not precisely match. Further investigation proved that the mandorla was originally made for the Butsugen butsumo Buddha.[142] Presumably, when the mandorla for the Nyoirin Kannon was damaged or lost at some point in the temple's history, the nimbus for the Butsugen butsumo was “borrowed” to replace it, and it remains with the Nyoirin Kannon today. Neither the Butsugen butsumo nor the Miroku statues has a mandorla today.

It seems very likely that the significance of the Nyoirin Kannon was amplified, both at the iconographic and liturgical levels, due to the increasing attention given the jewel—in connection with relic worship—at court and at temples in the capital and mountainous regions. The existence of two Nyorin Kannon paintings in the Nyōhōdō by 883 indicates that Kūkai's disciples had knowledge of nyoiju worship or other rites for the Kannon and that Kanshinji was linked, historically or otherwise, to jewel or relics beliefs. I have suggested that given the increasing importance of the Latter Seven-Day rite and the legend of the cintāmani at Muroo from the mid-tenth century, noted above, legends about the founding of Kanshinji, including its historical association with Kūkai and the seven-star Ursa Major, may have developed, along with practices honoring the nyoiju and relics.[143] These legends were then bolstered by the increasing attention given Kanshinji by court and clergy during the Nanbokuchō era. The fourteenth-century pagoda housing the four Buddhas and the miniature pagoda on the main hall altar noted in the 1878 Pilgrim's Account were undoubtedly part of this complex evolution of meanings.

The proposition of a late tenth-century date for the increasing relevance of the Nyoirin Kannon statue to rites at Kanshinji is further substantiated by the existence of an alternate image of the Nyoirin Kannon at Kanshinji (Fig. 23). Such icons, known as maedachi, “to stand before,” typically serve as stand-ins for secret images.[144] Dating to the eleventh century, this smaller image was likely placed before shrine doors concealing the older Nyoirin Kannon. The creation of a maedachi for worship suggests that the original Nyoirin Kannon became a secret image during the eleventh century.[145]

Hidden images like the Nyoirin Kannon—hibutsu, literally, “secret Buddha”—are common to Esoteric temples from at least the tenth century.[146] As Elias Canetti writes, “secrecy lies at the very core of power.”[147] Such images are revealed once a year or more infrequently and, as in other cultural traditions, are thought to hold considerable power and efficacy. The idea of concealing and revealing a sacred object finds its origins in the worship of relics.[148] The aura and power of the hidden icon are maintained through denial of its direct gaze. Walter Benjamin wrote, “The essentially distant object is the unapproachable one. Unapproachability is indeed a major quality of the cult image. True to its nature, it remains ‘distant,’ however close it may be.”[149] His seemingly contradictory concept of “aura” allows that even the mechanically reproduced object (in the present case, the revealed and researched icon) retains the distance and sacrality of its original context.

The fanatic's assault on the Nyoirin Kannon statue in 1955 was an affirmation of the deity's continuing power—its aura. (One is reminded here of the story of the sculpture of Shīrīn in the Grotto of Tāq I Bustān, about which Ernst Kris and Otto Kurz note, “the statue was believed to be alive not only by the man who offered his love to it, but also by those who mutilated it to prevent others from falling in love with it.”[150] But the incident also exposed the hidden—it closed the distance that maintains sacrality. Finally, it unexpectedly revealed the icon to the eye of historians and art historians, an audience that wanted greater knowledge of its formal attributes.

The history of Kanshinji and its environs reveals how the Nyoirin Kannon was linked to the wish-granting jewel (or vice versa), contributing to the deity's favorable reception around the thirteenth century. The status of the Nyoirin Kannon and other images at the temple changed in tandem with Kūkai's renown and in keeping with their role in traditions concerning relics and the stars, and with imperial patronage. Considerations of efficacy are only inferential in my arguments, largely because efficacy is rarely documented. In the religious context one can usually assume that a cherished image is perceived as being efficacious. And there is certainly a mutually supportive structure among factors that aid canonization and perceptions of efficacy. The elevated status of a work maintains certain conditions and beliefs, while the context (for example, oral tales or the building of a new hall) is at once engendered by the work. Catherine Bell's observations on ritualization are useful in understanding this relationship—among many others. “Ritualization, the production of ritualized acts,” can be seen as “the strategic production of expedient schemes that structure the environment in such a way that the environment appears to be the source of the schemes and their values.”[151]

Perhaps the work of art, more so than ritual, can be readily appropriated and (mis)understood as fixed in meaning, so that the “production of expedient schemes” by and about the work appears seamless. In order to understand both ancient and modern meanings for the Nyoirin Kannon, it is important to accept that the strategies that structure and maintain certain Shingon beliefs are pervasive and self-perpetuating. As Bell further states, “ritualization does not see how it actively creates place, force, event, and tradition, how it redefines or generates the circumstances to which it is responding.… It does not see what it does in the process of realizing this end, its transformation of the problematic itself.”[152] Like ritual, canonization is a strategy that both validates and extends the schemes it embodies; these schemes are internalized in subsequent reception (ritual theorists might consider canonization a ritual practice). Modern scholarship on the Nyoirin Kannon is structured by the schemes and activities that have occurred around it and in which it participates. This setting, a nexus of activities, consists of religious practice, apocryphal tales, belief, historical research, aesthetic appraisal, and the work's affective order,[153] among many responses.

In this final section I will return to a discussion of the relationship between characterizations of Esotericism and evaluations of the Kanshinji Nyoirin Kannon. The stasis of dependency created by what Hans Robert Jauss calls the “horizon of expectations”[154] and generalizations concerning Esoteric Buddhist expression and artistic value secures the work in an exalted, canonized, and thus unresponsive and fixed position as regards evaluation: both the statue and the teachings whose character it ostensibly expresses are mutually predetermining. As for its place in the artistic canon, art historians similarly situate the Kanshinji statue within chronological and stylistic taxonomies according to its formal and aesthetic features. These are then applied to privilege the icon in a history of use, meaning, and representation without sufficient documentary or other evidence.

Sawa Ryūken praised the Nyoirin Kannon as “a work of Esoteric statuary well known for its exceptional expression of sensuality.”[155] In 1964 he wrote,

     Not only is the Nyoirin Kannon at the Kanshinji the
     greatest statue of this Kannon ever produced in Japan; it
     also represents the pinnacle of all Japanese Esoteric
     Buddhist art. Seated with one knee raised in the royal-ease
     posture, the trunk held straight and head slightly inclined
     to one side, the figure has a soft sensuality that is
     balanced by the dignity of the pose. Not long after this
     statue was completed, Esoteric art in Japan settled into
     formalization ….[156]

Here and elsewhere the literature on the Nyoirin Kannon posits it as a model of Shingon expression wherein the magical or sublime power of the image—its excess[157]—is countered by two intellectual maneuvers: its power is classed as visual and sensual rather than efficacious or magical, while the potential sexual aspect of its power or appearance is cloaked by claims of mysterious and complex doctrine.[158] Such reformulation masks undesirable elements while preserving them in a different, normalized, form.

Bernard Faure puts it another way:

     If unlike the devotional or ritualistic approach—and
     more than the traditional Buddhist emphasis on
     beauty—the modern aesthetic approach is essentially a
     strategy for containing the “impure” (sexual or
     magical) elements of cultural artifacts, we need to move
     beyond aesthetic discourse to consider the abundant, yet
     neglected, anthropological data regarding (and regarded by)
     the icon.[159]

The terms used to describe Esoteric icons frequently take on the same tone as the object they describe;[160] the detailed, complex, colorful appearance and the lively poses of figures in mandala paintings often elicit highly charged, dense descriptions. In the case of the Nyoirin Kannon, the delicate, indeterminate, caressing tone of the terms chosen by various authors mimics their appraisal (or vice versa): “minagiru nikutai no utsukushisa” (beauty constituted by overflowing sensuality),[161] “minagiru mikkyōtekki kan'nōtekina jōsokukan” (overflowing esoteric and sensual repleteness),[162] “maborokai ayashii made no nikkan no hyōgen ni taishite” (phantasmal, even strangely mystical expression of sensualism).[163] The icon is photographed with a dark background and glowing direct light on it—“boutique lighting”; when this effect is used in today's museum displays, Stephen Greenblatt finds that “[the] pool of light that has the surreal effect of seeming to emerge from within the object rather than to focus upon it from without—is an attempt to provoke or heighten the experience of wonder, as if modern museum designers feared that wonder was increasingly difficult to arouse.”[164] One is led to question, then, the intention behind giving the sacred Buddhist icon a heightened experience of wonder. Does the temple fear that devotional wonder has waned?

Descriptions that mention the sensual, sexual, or somatic qualities of the Nyoirin Kannon are posited largely by Shingon scholar-priests. These may arise from unrecorded ritual experiences and insights somehow akin to sexual experiences.[165] In China the rites of the stars had long involved sexual promiscuity[166] and were, as we have seen, a vital part of esoteric belief and practice—and of import to Kūkai before his esoteric training overseas. These rites, specifically those of Ursa Major, were suppressed by the Nara and Heian governments. The hibutsu status of the icon also contributes to a desire for the unseen or unobtainable that is analogous with sexual desire. The secret image is a double, an imagined body, of the deity. Secreting the icon helps to preserve its efficacious presence for the devotee. Sexual arousal or meditative experiences akin to arousal may begin with the gaze Looking, perception, visualization, and visual attention all find their genesis in an ontology of images that recognizes power in the absent image.

Comparing the Kanshinji work to the roughly contemporaneous Esoteric Shingon statues in the Tōji lecture hall (Figs. 16, 18, 24), Nishikawa writes: “for the first time we can see an originality of expression: there is an overflowing Esoteric and sensual repleteness in the expression and pose never before witnessed in previous periods.”[167] Sawa's research, like that of many other scholars, characterizes the Nyoirin Kannon as enigmatic, exotic, and sensual, but it omits a discussion of closely related and mutually constituting spheres: ritual activity and the icon's efficacy. Instead, they acknowledge the statue's status as a religious icon in terms that maintain an idealized characterization of Esoteric Buddhism and Esoteric Buddhist icons, almost exclusively through a description of a work's artistic merit, material or visual features, the historical record (especially that which leads to dating or a historically valued pedigree), and historical relativism. In Sawa's characterization, the Tōji statues show many of the same stylistic traits attributed to the “over-flowing Esoteric and sensual repleteness” of the Nyoirin Kannon (compare the Tōji bodhisattva, Fig. 24, with its full—although taut—form and gentle countenance).

Both the formal and somatic-religious approaches to the statue reflect general trends in scholarship for Buddhist sculpture during the late nineteenth century. Buddhist art was rehabilitated during the Meiji era as part of Japan's new cultural identity, its cultural partrimony. Research prior to World War II analyzed statues primarily according to stylistic traits based on formal analysis, or stylistic traits attributed to a religious sect's mode of expression (for example, Pure Land style, Shingon style, and so on). Sculpture of the Heian period (794–1185) was usually divided into early, middle, and late phases. Early Heian works (794–931) were characterized as belonging to the Nara style, Esoteric style, and exoteric (kengyō)style.[168] Postwar scholars generally prefer stylistic groupings specific to technique, such as wooden (mokuzō or junmokuzō chōkoku), faux sandalwood (danzō), and lacquer or lacquer-treated (kanshitsuzō or mokushin kanshitsuzō) statue types.

In the formal approach, stylistic, decorative, and other features of the Nyoirin Kannon, all laudatory, are said to point to the imperial court or similar patrons of ample means and artistic control. Careful comparisons of the Kannon's drapery folds, ear formation, surface modeling, and method of construction with those of other works provide valid evidence for linking it to ninth-century state-sponsored workshops. The data from both formal and technical studies generally provide the information for a linear typology and chronology of statuary and lead to larger taxonomic divisions of “native style,” “Chinese style,” “Esoteric style,” “exoteric style,” and so on, which are also useful categories in comparative studies. The articulation of such divisions, however, may rest on false claims situated outside the work. When a work is admired and canonized, contingencies such as time and place and function are pushed aside or reformulated to maintain other interests. In such cases, taxonomic divisions and models begin to limit our understanding of a work. Similarly, when the mysterious, profound, and foreign character of Esoteric Buddhism is discussed, examples such as the Nyoirin Kannon come to “prove” these very characterizations.

According to Smith, a canonized work is isolated because it “performs certain desired/able functions particularly well at a given time for some community of subjects, being perhaps not only ‘fit’ but exemplary.” Under such conditions it “will have an immediate survival advantage” and, by virtue of its resultant cultural reproduction, “will be more readily available to perform those or other functions for other subjects at a subsequent time.”[169] Appraisals of the Nyoirin Kannon support a normalized definition of Esotericism that inhibits exploration of many other images and image functions. If these evaluations are understood, in part, as a symptom of circumventing the work's powerful presence, their normalization speaks further to the suppression of that power.[170] At some point in their history, the two extant ninth-century Buddha statues at Kanshinji, the Butsugen butsumo and Miroku, were removed from the altar they once shared with the Nyoirin Kannon. Today, they are normalized: art historians consider the two artistically inferior to the Nyoirin Kannon and they are housed by temple administrators in a separate building for temple artifacts, not in any of Kanshinji's worship halls. Because the two extant Buddhas do not dovetail with the preferred modern characterization of Esoteric Buddhism as enigmatic or sensual, there is further impetus to discount their role at the monastery and in the history of Buddhist art. Removed from a liturgical context, the works have little opportunity to prove efficacious in the life of the practitioner.

    One of the determinations of the question of value is the
    predication of the subject.—Gayatri Chakravorty
    Spivak, 1985[171]

An epigraph typically functions as a rebus for an essay, providing a gloss or indicating the author's approach. Spivak's words serve as both a closing and self-censure. The Nyoirin Kannon has been the subject of many scholarly discussions, and in my attempt to decenter it I have only added to the literature. This essay nonetheless seeks to bring some balance to existing studies. Today's horizon is treated here as both the subject of study and as the object of history. I have done more than suggest that other interpretations are possible—and necessary—for the Kanshinji Nyoirin Kannon. Awareness of the unconscious biases of past and present has stimulated a fresh analysis. By considering the bases for prevailing interpretations, both etic and emic sources for canonization have been identified.[172] As part of this process new or recovered meanings have been suggested for both lost and extant statues and paintings at Kanshinji. Historical consciousness, Hans-Georg Gadamer explains, leads to “seeing historical movement not only in process, but also in understanding itself. Understanding is not to be thought of so much as an action of one's subjectivity, but as the placing of oneself within a process of tradition, in which the past and present are constantly fused.”[173] It is from within and outside a variety of traditions, canons, and histories that the Kanshinji icons should be understood.

This essay is dedicated to Professor John Rosenfield (Professor Emeritus, Harvard University), who sponsored my first study of the Nyoirin Kannon, presented at the Frick Symposium, New York, in 1985. He then patiently guided my interest in Buddhist art for a decade thereafter. Professor Mizuno Keizaburō (Professor Emeritus, Tokyo Geijutsu Daigaku) trained me to assess Buddhist statues and the documents of Japanese temple history and has supported my critical examination of his scholarship on the Kanshinji statues. I am indebted to the many Japanese scholars whose inestimable scholarship I evaluate in this essay, especially the late Nishikawa Shinji. My research and writing were sponsored by the J. Paul Getty Foundation; the Walter H. Simpson Center for the Humanities, University of Washington; and the Japan Foundation Endowment Fund, University of Washington.

I would like to acknowledge the generosity and expertise of friends and colleagues who commented on this essay at various stages, especially Ryūichi Abé, Ian Astley, the late Robert Boardingham, Jeffrey Collins, Allan Grapard, Christine Guth, Karen Kelsky, Charles Lachman, Donald McGallum, Samuel Morse, John Stevenson, Mimi Yiengpruksawan, and a very considerate anonymous reader for the Art Bulletin. The Rev. Eko Noble has kindly offered suggestions and shared her insights at key moments in my research. I am also grateful for input from students in a 2000 seminar on Esoteric art at the University of Washington, especially Heather Blair, and for suggestions from Alfred Acres, Barbara Altmann, Asai Kazuharu, Fujii Keisuke, Allen Hockley, Kihara Toshie, Konno Toshifumi, Mizuno Keizaburō, Brian Ruppert, Timon Screech, Robert Sharf; the late Nishikawa Shinji, Yukiko Shirahara, Jerome Silbergeld, Wada Keiko, Marek Wieczorek, and Catherine Vance Yeh.

Unless otherwise indicated, translations are mine. Transliterations of Japanese words, unless otherwise noted, are from MD. The artist is unknown for all works illustrated. Sanskrit terms may be used over Japanese equivalents if common usage dictates.

1. Literally. the “main object of veneration.” The Sanskrit equivalent is Sayadhi-devatah. The term probably derives from Tantric texts. In Japan, it quickly came to be used by both Esoteric and non-Esoteric traditions. In Japanese Esoteric practice the honzon can take three forms (ji, in, keizō): a verbal “seed syllable” (hr Japanese: shuji, Sanskrit [henceforth, Skt]: bija); a symbolic mudrā, or hand gesture; or a pictorial representation. Each of these is further subdivided into six groups, according to ritual texts. The term appears in the Dainichi-kyo (Mahavairocana sutra); see esp. the section honzon sanmai bon (sanmai, Skt: samādhi, meditative consciousness, concentration), T, vol. 18, text no. 848; see MD, 2068. In Notes on the Secret Treasury (Hizōki), a text by Kūkai featuring the oral instruction he received in China from the esoteric master Huiguo, he discusses the term honzon (KZ, vol. 2, 30).

2. A bodhisattva, literally, bodhi being, has voluntarily stopped the enlightenment process in order to help sentient beings on the path to attain enlightenment. Kannon (Skt: Avalokiteśvara), often called the bodhisattva of mercy, is the most popular form of the deity in east Asia. In Sanskrit, Nyoirin Kannon is Cintāmanicakra-avalokiteśvara.

3. Eighteen corresponds to the day set for Kannon worship by the “ten precept-observation days” (jūsainichi).

4. For an excellent discussion of Nyoirin Kannon iconography and iconology, numerous illustrations of both Japanese and Chinese images, and an extensive bibiography, see Inoue Kazutoshi, ed., Nyoirin Kannonzō, Batō Kannonzō, Nihon no bijutsu, no. 312 (May 1992): esp. 19–53, 96–97. See also MD, 1738–39; and Hamada Takashi, “Boston bijutsukan Nyoirin Kannon gazo kō,” Museum, no. 386 (May 1983): 23–33. For a stylistic and iconographic analysis of Nyoirin Kannon statues in English, see Sherry Fowler, “Nyoirin Kannon: A Chronological Analysis of Six-Armed Sculptural Examples from the Ninth through the Fourteenth Century,” M.A. thesis, University of Washington, 1989. Further references for the Kanshinji statues are provided below.

5. In Japanese, Henge Kannon, literally, “transformed-shape Kannon.” Kannon is one of many hodhisattvas, and the Nyoirin is one of many transmigrations or forms of Kannon. Transmigrations of the bodhisattvas and Buddhas feature prominently in Esoteric Buddhism—in both texts and imagery—and to a lesser degree in non-Esoteric traditions. A two-armed form of the deity is based on exoteric (that is, not Esoteric) texts. In the Victoria and Albert Museum, London, there is a Javanese four-armed gilt-bronze statue of this deity, roughly contemporary, with the six-armed Japanese example at Kanshinji (no. 459–1880, dated by the museum to the early ninth century). For Indian and Chinese examples, see Inoue and Fowler (as in n. 4).

6. The 1897 (Meiji 30) Protection of Old Temples and Shrines Law (Koshaji hozonho) designated numerous temple and shrine artifacts for government-funded conservation and exhibition in national museums. These were revised several times, and in 1951 (Shows 26) the number of Kokuho was drastically reduced. At that time the government designated Important Cultural Properties (Jūyō bunkazai), totaling 8,339 artworks and 1,753 structures, and National Treasures (Kokuhō, the highest level of importance), totaling 825 artworks and 207 structures. These same terms and designations remain in use today; new items are regularly added to both categories.

7. Left second hand (holding the lotus bud) and right third hand (holding the prayer beads). The burnt remains of the lotus bud was used as a core within the newly fashioned lotus, attesting to the perceived power and efficacy of the original piece.

8. A summary of the official record of the Dec. 1955 incident is kept by the Agency for Cultural Affairs, Tokyo.

9. Nishikawa Shinji, “Kanshinji Nyoirin Kannonzō ni tsuite,” Bijutsushi 22, vol. 6, no. 2 (Dec. 1956): 1–14.

10. Nishikawa Shinji, “Kanshinji no butsuzō (jō),” and “Kanshinji no butsuzo (ge),” Bukkyō geijutsu 119 (Aug. 1978): 61–68; and 121 (Dec. 1978): 86–106. One year before, Nishikawa S. collaborated with Nishikawa Kyotarō, Mizuno Keizaburō, and others in compiling pertinent documents, sources, technical analyses, and other relevant information on the Nyoirin Kannon statue and (to a far lesser degree) the two extant Buddha statues, in vol. 3 of NCSS-jsh. This volume is untitled but is described in the table of contents as “Nyoirin Kannon bosatsuzō, Osaka, Kanshinji.” Other significant research on the Nyoirin Kannon statue includes (in chronological order) Adachi Yasushi, “Kanshinji honzon to Kanshinji engi jitsurokuchō,” Kenchikushi 2, no. 3 (May 1940): 11–22; Tsuji Shind&omarc;, “Kanshinji Nyoirin,” Bukkyō geijutsu 10 (Dec. 1950): 74–75; Sawa Ryūken, “Kanshinji Nyoirin Kannonzō ni tsuite,” Bukkyō geijutsu 10 (Dec. 1950): 76–77; Ikawa Kazuko, “Akogare no hibutsu: Kanshinji Nyoirin Kannonz&omarc;,” Sansai no. 84 (Feb. 1957): 15–17; Kurata Bunsaku, Butsuzō no mikata (gihō to hyōgen) (Tokyo: Daiichi hoki shuppan kabushikigaisha, 1965), 134–39; Kuno Takeshi, “Kanshinji no Heianshoki butsuzō ni tsuite,” Kokka 22, no. 961 (1973): 15–24; idem, “Heian shoki no bosatsuzō,” in Heian shoki chokokushi, 2 vols. (Tokyo: Yoshikawa kōbunkan, 1974), 205–17; Ota Hirotarō, ed., Nihon kenchikushi kiso shiry&omarc; sh&umarc;sei, vol. 7, Butsudō IV (Tokyo: Chūō kōron bijutsu shuppan, 1975). 100–107, 189–92, 199–201; Fukuyama Toshio, “Kanshinji no sōritsu ni tsuite,” Bukkyō geijutsu, no. 119 (Aug. 1978): 53–60; Tanaka Megumi, “Kanshinji sosoki no zobutsu to Shinsho,” Iwate daigaku kyōikugakubu kenkyū nenpō 41, no. 2 (1982): 59–80; Tamura Ryusho, “Kanshinji to Kyōonji no butsuzō,” in Nihon koji bijutsu zenshū ed. Nagai Shin'ichi. vol. 7, Shitennōji to Kawachi no tera (Tokyo: Shūeisha, 1981), 117–24; Ito, esp. 115–20; Konno Toshifumi, “Kanshinji Nyoirin Kannonzō no fukei,” in Nihon bijutsu zenshū, ed. Maekawa Seiro and Okawa Naomi, vol. 5, Mikkyo jiin to butsuzō, ed. Mizuno Keizaburō, Konno Toshifumi et al. (Tokyo: Kodansha, 1993), 156–62; and Mizuno Keizaburō, “Heian jidai zenki no chōkoku,” in ibid., 147–48.

11. On this subject, see Smith, esp. chap. 3. Viewed from another theoretical stance, Pierre Bourdieu situates the image in a field of cultural production. See his essays “Outline of a Sociological Theory of Art Perception” and “The Field of Cultural Production, or: The Economic World Reversed,” in The Field of Cultural Production: Essays on Art and Literature, ed. Randal Johnson (New York: Columbia University Press, 1993), 29–73 and 215–37, respectively.

12. I cannot deal here in depth with the efficacy of the image for the devout, largely because that aspect of the icon's history has not typically been recorded. Questions of efficacy are certainly of importance to devotees.

13. A Chinese statue of a priest also survives, which can be dated to the ninth century (Fig. 15). It is 15 in. (37.8 cm) tall and constructed of Chinese cherry wood, with touches of pigment remaining. For a photograph and brief discussion of the priest figure, see the exhibition catalogue Danzō: Byakudanbutsu kara Nihon no mokuchōbutsu e: Tokubetsuten (Nara: Nara kokuritsu hakubutsukan, 1991), 148–49 (cat. no. 66); and Mae and Nagashima, 136–37, pls. 47, 48. See also Mainichi shinbunsha, ed., Jūyōbunkazai, vol. 6 [chōkoku VI] (Tokyo: Mainichi shinbunsha, 1975), 113, no. 277.

14. The journal is dated Eiwa 4. Excerpts appear in NCSS-jsh 3 (shiryō 6, line 5), text section (unless otherwise noted, pages refer to text pages and not pages of a preceding section of plates), 45. The journal notes that the image, which the priest was allowed to view, had been a secret image from “earlier times” and had rarely been seen by visitors. The record will be discussed in detail below in “The Life of Images.”

15. Only Kuno Takeshi dates the two extant Buddha statues earlier, to about 834–48 (Jōwa era), and the Nyoirin Kannon to about 850. See Kuno, 1973 (as in n. 10). See also Kuno 1974 (as in n. 10), 211–17. All other studies disagree with Kuno's dating, citing stylistic criteria.

16. Martin Jay, Downcast Eyes: The Denigration of Vision in Twentieth-Century French Thought (Berkeley: University of California Press, 1993). For an excellent summary of the history of visuality, see Robert Nelson, “Descartes's Cow and Other Domestications of the Visual,” introduction to Visuality before and beyond the Renaissance: Seeing as Others Saw (Cambridge: Cambridge University Press, 2000), 1–21. In a manuscript in progress I will attempt to “see as others saw” in the 9th century; the current essay primarily addresses how modern visuality affects possibilities of other visualities, to borrow current terminology.

17. Harold Bloom, The Western Canon: The Books and School of the Ages (New York: Harcourt Brace, 1994), 1, 3. There is a significant body of work in the field of literary criticism on the concept of the canon. Useful are Barbara Herrnstein Smith; “Canons,” Critical Inquiry 10 (1983): 1–36; Leslie Fiedler and Houston Baker, eds., English Literature: Opening Up the Canon (Baltimore: Johns Hopkins University Press, 1981); Herbert Lindenberger, The History in Literature: On Value, Genre, Institutions (New York: Columbia University Press, 1990); and Paul Lauter, Canons and Contexts (New York: Oxford University Press, 1991), among many others. Discussions of a canon of literature or art should consider the relative meaning of value, “a term that straddles the material and spiritual realms” (Lindenberger, xvii). Recent examples on the artistic canon and value include Ann Gibson, “Recasting the Canon: Norman Lewis and Jackson Pollock,” in Modern Art and Society: An Anthology of Social and Multicultural Readings, ed. Maurice Berger (New York: IconEditions, HarperCollins, 1994), 216–30; Diane Hill, “The ‘Real Realm’: Value and Values in Recent Feminist Art,” in Interpreting Visual Cultures: Explorations in the Hermeneutics of the Visual ed. Ian Heywood and Barry Sandywell (London: Routledge, 1999), 143–61; and Joseph Leo Koerner and Lisbet Koerner, “Value,” in Critical Terms for Art History, ed. Robert Nelson and Richard Shiff (Chicago: University of Chicago Press, 1996), 292–306. There is very little discussion in the literature on the notion of value in Buddhist imagery. Of relevance are Richard Davis, Lives of Indian Images (Princeton: Princeton University Press, 1997); and Stanley K. Abe, “Inside the Wonder House: Buddhist Art and the West,” in Curators of the Buddha: The Study of Buddhism under Colonialism, ed. Donald S. Lopez Jr. (Chicago: University of Chicago Press, 1995), 63–106.

18. See, for example, Norman Bryson, “Semiology and Visual Interpretation,” in Visual Theory: Painting and Interpretation, ed. Bryson, Michael Ann Holly, and Keith Moxey (New York: IconEditions, HarperCollins, 1991). 61–73.

19. Sherwood Moran, “Early Heian Sculpture at Its Best: Three Outstanding Examples,” Artibus Asiae 34 (1972): 155. Although Moran was not formally trained as an art historian, his work was respected. As considered below, other Western scholars echoed his view.

20. Excessive valorizing, artistic, or aesthetic descriptions suggest that something has been left out. On disavowals of power and sexuality, see David Freedberg, The Power of Images: Studies in the History and Theory of Response (Chicago: University of Chicago Press, 1989), 345–77, 429ff. He writes: “the time has come to acknowledge the possibility, that our responses to images may be of the same order as our responses to reality; and that if we are to measure response in any way at all, then it is to be seen and judged on just this basis” (438).

21. Nishimura Kōchō, “Omoi no mama ni takara o kureru hotoke-san,” in Nyoirin Kannon, ed. Nagasaki Hashio et al., vol. 15 of Miwaku no butsuzō, ed. Mainichi Shinbun (Tokyo: Mainichi shinbunsha, 1987), 37. Figs. 1, 2, and 12 are reproduced in this publication; Mae and Nagashima; and others. As a Buddhist sculptor and conservator of many Japanese works of art, including National Treasures, and an Esoteric priest of the Tendai school, Nishimura Kōchō has special access to icons such as the Nyoirin Kannon—and special control over their interpretation.

22. Sherry Fowler, “Nyoirin Kannon: Stylistic Evolution of Sculptural Images,” Orientations 20 (1989): 58, 60. This article was written at an early point in Fowler's career; the author would probably approach the topic differently today. The passage nonetheless demonstrates the pervasiveness of canonizing assessments of the type I have been describing, as well as of the influence of Japanese scholarship on subsequent appraisals.

23. The latter has been the received understanding of the Zen tradition. On the evolution of “reverse orientalism” or Zen “occidentalism,” respectively, see Bernard Faure, “The Kyoto School and Reverse Orientalism,” in Japan in Traditional and Postmodern Perspectives, ed. Charles Wei-hsun Fu and Steven Heine (Albany: SUNY Press, 1995), 245–81; and Robert H. Shaft, “The Zen of Japanese Nationalism,” in Lopez (as in n. 17), 107–65, and an earlier version in History of Religions 33, no. 1 (1993): 1–43. See also Shafts essay “Visualization and Mandala in Shingon Buddhism,” which addresses the understanding of mandala and the reconstitution of the public and sacerdotal traditions of Shingon practice as highly individual psychoexperience, in Living Images: Japanese Buddhist Icons in Context, ed. Elizabeth Horton Shaft and Robert H. Shaft (Stanford: Stanford University Press, 2001). I am grateful to Dr. Sharf for providing me with a copy of his essay prior to publication, and for his consultation.

24. The term Protestantization (of Buddhism) was coined by Gananath Obeyesekere to describe the new form of Buddhism that developed in British-dominated Sri Lanka in the late 19th century. See Obeyesekere, “Religious Symbolism and Political Change in Ceylon,” Modern Ceylon Studies 1, no. 1 (1970): 43–63 (the term is defined on 46–47). For an excellent discussion. see also Gregory Schopen, “Archaeology and the Protestant Presuppositions in the Study of Indian Buddhism,” History of Religions 31 (1991): 1–23. Schopen writes, “it is possible that a sixteenth-century Protestant polemical conception of where ‘true’ religion is located has been so thoroughly absorbed into the Western intellectual tradition that its polemical and theoretical origins have been forgotten” (22). The American wife of early 20th-century Japanese Zen teacher D. T. Sozuki, Beatrice Erskine Lane (1878–1938), studied Shingon Esoteric Buddhism in Japan and published on the sect during the 1920s and 1930s in the journal Eastern Buddhist. See also Beatrice Erskine Lane and D. T. Suzuki, Impressions of Mahayana Buddhism (Kyoto: Eastern Buddhist Society; London: Luzac. 1940). On Buddhist studies in Europe and America, see J. W. de Jong, A Brief History of Buddhist Studies in Europe and America, Bibliotheca Indo-Buddhica, no. 33, 2d ed. (Delhi: Sri Satguru, 1987). Donald Lopez has termed the early period of Buddhist studies in the West the “European construction of an original Buddhism”; Lopez, Elaborations on Emptiness: Uses of the Heart Sutra (Princeton: Princeton University Press, 1996), 99. The research of James Edward Ketelaar, Of Heretics and Martyrs in Meiji Japan: Buddhism and Its Persecution Princeton: Princeton University Press, 1990), is also extremely helpful in understanding issues of modernization and religion in Japan. On Western attitudes toward religious experience, see Wayne Proudfoot, Religious Experience (Berkeley: University of California Press, 1985). I have drawn on these works and that of Robert Sharf, above, for the theoretical framework of my analysis of Esoteric reception.

25. Smith, 49. Hans-Georg Gadamer deals with these same issues in his discussions of objectivity and “the classical.” See Gadamer, Truth and Method, trans. Sheed and Ward Ltd. (New York: Seabury Press, 1975), 253–58; originally published as Wahrheit und Methode (Tübingen: Mohr Sieberk, 1960). Smith discusses “the classical” on 50–51. Theodor Adorno reminds us that none of these statuses, however, are fixed: “If each work is in a condition of equilibrium, each may yet once again enter into motion. … What art works say through the configuration of their elements in different epochs means something objectively different, and this ultimately affects their truth content.” See Adorno, “Toward a Theory of the Artwork,” in Aesthetic Theory, ed. Gretel Adorno and Rolf Tiedermann (Minneapolis: University of Minnesota Press, 1997), 193–94. For Pierre Bourdieu, actions such as essentialization and canonization presume a less mediated relationship to the object, a relationship presumed and secured by class privilege. Sharf, 2001 (as in n. 23). and idem, “Experience,” in Critical Terms in Religious Studies, ed. Mark C. Taylor (Chicago: University of Chicago Press, 1998), 94–116, suggests that the application of a “hermeneutics of experience” falsely constructs Shingon meditation as a personal, experiential practice. On the pernicious effects of Shingon sectarian scholarship on the study of Esotericism in eastern Asia, see Charles D. Orzech, “Seeing ‘Chen-yen’: Traditional Scholarship and the Vajrayana in China,” History of Religions 29, no. 2 (Nov. 1989): 87–114. On the impact of Shingon scholarship on our view of both Esotericism and Esoteric art, see Cynthea J. Bogel, “A Matter of Definition: Japanese Esoteric Art and the Construction of an Esoteric Buddhist History,” Waseda Journal of Asian Studies 18 (1996): 23–39.

26. Sawa Takaaki [Ryuken], Art in Japanese Esoteric Buddhism, trans. Richard L. Gage (New York: Weatherhill; Tokyo: Heibonsha, 1972), 56; originally published as Mikkyō no bijutsu, vol. 8 of Nihon no bijutsu (Tokyo: Heibonsha, 1964), 56. The translation gives an accurate equivalent of the Japanese except that “nihonjin no shūkyōkan ni yotte” (purely Japanese feelings) is better rendered as “Japanese religiosity.”

27. Ibid., 56, 80. For many years Sawa's work was the only material on Esoteric art available in English other than that dealing with mandala paintings (narrowly defined). Although he is respected more today as an iconography expert than an art historian, his work was nonetheless very influential and his response to the Nyoirin Kannon typical of enduring attitudes.

28. White, 9. As White further notes, this definition must be modified according to its contexts. An excellent study of Buddha systems in Tantric Buddhism is Yoritomi Motohiro, Mikkyō butso no kenkyū (Kyoto: Hōzōkan, 1990), with a useful English summary, 691–716.

29. Kūkai studied on a scholarship from the government that was meant to keep him in China for twenty years. The best source on early Shingun history in English is the recent study by Abé. On Kūkai and his writings, see Yoshito S. Hakeda, Kūkai: Major Works (New York: Columbia University Press. 1972); and on Shingon practice and thought, Taikū Yamasaki, Shingon: Japanese Esoteric Buddhism, trans. Richard and Cynthia Peterson (Boston: Shambhala, 1988). In his book on early Shingon history, Abé convincingly argues that the ritual language of mantra was the basis for Kūkai's dissemination of Mikkyō. In Japanese the best recent sources are Kushida Ryōkō. Shingon mikkyō seiritsu katei no kenkyū (Tokyo: Sankibō busshorin, 1964); and idem, Kūkai no kenkyū (Tokyo: Sankibō busshorin, 1981); Matsunaga Yūkei, Mikkyō no rekishi (Kyoto: Heirakuji shoten, 1969); and Takagi Shingen, Kūkai shisō no shoshiteki kenkyū (Tokyo: Hōzōkan, 1990). The best sources on Kūkai in English are Abé; Hakeda; and David Lion Gardiner, “Kukai and the Beginnings of Shingon Buddhism in Japan,” Ph.D. diss., Stantord University, 1994.

30. Literally, “mental device or instrument of thought,” mantra is “an acoustic formula whose sound shape embodies the energy-level of a deity: a spell, incantation or charm employed in Tantric ritual or sorcery” (in White, 629). The definition of mantra varies considerably. Mantra recitation can be used for personal, worldly, and other effect and has been understood as having religious, linguistic, magical, ritual, and somatic functions, among others. In English, see Harvey P. Alper. ed., Mantra (Albany: Suite University of New York Press, 1989); and Fritz Staal, Rules without Meaning (New York: Peter Lang, 1989).

31. In this essay Esoteric (capitalized) will be used when referring to the systematized esoteric tradition, that is, Shingon and, occasionally Tendai, and esoteric (lowercase) will be used for pre-Shingon or non-Shingon esoteric elements. The process of systemization continued after Kūkai. For 8th-century pre-Shingon esoteric statues and practices, see Cynthea J. Bogel, “Ritual and Representation in Eighth-Century Japanese Esoteric Buddhist Sculpture,” Ph.D. diss., Harvard University, 1995. On 8th-centnry esotericism in Japan, see Kushida. 1964 and 1981 (as in n. 29); Ishida Mosaku, Shakyō yori mitaru Narachō bukkyō no kenkyū (Tokyo: Tōyō bunko, 1931); and Horiike Shunpo, “Nara jidai bukkyo no mikkyoteki seikaku,” in Kūkai, Nihon meiso ronshu, 3, ed. Wada Shūjo and Takagi Shingen (Tokyo: Yoshikawa kōbunkan, 1982). The Shingon school wits established by Kūkai, but during his lifetime he did not often identify his Shingon teachings as a “school” (Shingonshū) so much as a category (for example, the shingon secret treasury or secret vehicle). The term Shingonshū was used in Emperor Junna's decree of Kōnin 14 (823); this coincided, however, with the granting of certain privileges to the Tendai sect and a broader atmosphere of political interest in these two new Buddhist teachings. See Bogel (as in n. 25), 37–38, and Abé, esp. 189–204. Later generations stressed Shingonshu over other designations and emphasized terms such as pure esotericism (junmitsu) in contradistinction to miscellaneous esotericism (zōmitsu). Bernard Faure has discussed the Tachikawa-ryū, a subschool of Shingon, which equated sexual bliss with Kūkai's doctrine of “becoming a Buddha in this very body” and gave Shingon practice a much more sexualized coloring. See Faure, “Japanese Tantra, the Tachikawa-ryu, and Ryobu Shinto,” in White, 543–56.

32. The validity of what Kukai brought was not accepted unquestioningly. Apocryphal texts (that is, sutras. or scripture, composed in China) and insufficiently learned monks returning from abroad were thoroughly reviewed by the monastic hierarchy; apocrypha were often accepted as “original sutra.” Kūkai's writings and ideas and the new texts he imported were debated and discussed by clergy in the temples of Nara.

33. Hereafter referred to as sutra(s) and mandala(s). According to White, 629, a mandala is a “‘circle’; an idealized circular model of the cosmos, with the source of cosmic or temporal power located at the center, and deities or beings representing lesser powers or energies radiating outward toward the periphery, the limits of the system.” Kūkai carried huge polychrome paintings that both represented and embodied (or radiated) this concept. Original Buddha nature dwells in the mandala, and it manifests the practitioner's potential for spiritual awakening. It is also a part of the primary rituals of Shingon Buddhism. Mandalas may be geographic, architectural, sculptural, or linear/painterly; they are form, concept, and practice. On the subject of mandala in Japanese, see Toganoo Shoun, Mandara no kenkyū (Kōyasan, Wakayama-ken: Kōyasan Daigaku Shuppanbu, 1927); and Ōmura Seigai, Sanbon ryōbu mandarashu (Tokyo: Bussho Kankōkai, 1913). In English, see Elizabeth ten Grotenhuis, Japanese Mandalas: Representations of Sacred Geography (Honolulu: University of Hawai'i Press, 1999); David L. Gardiner, “Mandala, Mandala on the Wall: Variations of Usage in the Shingon School,” Journal of the International Association of Buddhist Studies 9, no. 2 (1996): 245–79; Lokesh Chandra, The Esoteric Iconography of Japanese Mandalas (New Delhi: Jayyed Press, 1971); Adrian Snodgrass, The Matrix and Diamond World Mandalas in Shingon Buddhism, 2 vols. (New Delhi: Aditya Prakashan, 1988); and Ulrich H.R. Mammitzsch, The Ryobu Mandara of Shingon Buddhism in Religious and Historical Perspective, vol. 8 of Ajia Kenkyujo kiyo (Tokyo: Ajia Daigaku Ajia Kenkyujo, 1981).

34. The “universal three secrets”; Skt: triguhya; Jpn: sanmitsu. The yogic techniques are described in many Esoteric texts, especially in chaps. 19–23 of the Mahāvairocana sutra. See Abé, 120–33.

35. Ski: tathatā. Traditionally translated as “thusness” (Jpn: shinnyo), it conveys a Mahayana conception of the true reality underlying all phenomenal discrimination, the absolute source of all.

36. Skt: rūpa (Jpn: shiki). Form or matter, that which is capable of disintegration.

37. Goshōraimokuroku (Inventory of Imported Items), written by Kūkai in 806, and found in KZ. vol. 1, 95. This passage is from the preface to the section listing imported Buddhist images, specifically, mandala paintings and patriarch portraits. For a complete translation in English, see Hakeda (as in n. 29), 140–50. For provocative comments on the topic of imagery and Buddhist practice, see Roger Goepper, “Some Thoughts on the Icon in Esoteric Buddhism of East Asia,” in Studia Sino-Mongolica, Festschrift für Herbert Franke, ed. Wolfgang Bauer, Münchener Ostasiatische Studien, vol. 25 (Wiesbaden: Franz Steiner, 1979), 245–54. The work of Hans-Georg Gadamer is of relevance here for his understanding of the relationship between the image (the represented) and the original, which he posits as mutually constituting. The represented image is art “ontological event” that affects the original and “shares in what it represents.” See Gadamer (as in n. 25), 127, 125.

38. Kūkai's Hizōki, in KT, vol. 2, 40–41, trans. Abé, 129–30.

39. Dated Gankyo 7 (883).9.15, that is, the seventh year of the Gankyō era (or 883). ninth month, fifteenth day (see Fig. 20). The Register is in the traditional handscroll format. For a complete transcription, see NCSS-jsh 3 (shiryo 4), 35–44, which includes photographs of sections of the original manuscript, consisting of one scroll 11 in. (27.7 cm) in width and 42 ½ in. (108 cm) in length. During the Nara and early Heian periods, inventories called shizaichō or shizai rukichō usually were required of officially sponsored temples (typically) called jōgakuji, “fixed-stipend temples” (see n. 43 below) or daiji, “important temples.” These temples received funds for standard equipment such as oil and candles and for repairs in exchange for submitting inventories of their assets to the state. See Hiraoka Jōkai, Nihon jiinshi no kenkyū, kodai-hen (Tokyo: Yoshikawa kōbunkan, 1981), 417–52.

40. The temple's modern address is Osaka-fu, Kawachi Nagano-shi, Teramoro 475. Kawachi is an ancient province, now part of Osaka Prefecture. It is situated along tile ridge of the Kongō-Katsuragi mountain range within a picturesque valley extending north from Nagano-shi. The Izumi mountain chain runs to the southwest and the Kongo (alt. Kongo-Katsuragi) range to the east; Mt. Kongō rises 3,648 ft. (1,112 m) and in the southeast Mt. Katsuragi stands at 3,150 ft. (960 m). Today Kanshinji may be reached by train, then bus and foot.

41. As the NCSS-jsh 3 authors point out. Tenchō 2 (826) is an error for Tenchō 3 (827), based on the calendrical designation hinoui uma, heigo noted in the text. For an excellent discussion of the founding history, see also Fukuyama (as in n. 10), 53–60. Many later histories state that the site was first used by Jitsue and called Unshinji (Cloud-mind temple), the name being changed to Kanshinji in 837.

42. On Shinsho, see Mochizuki Shinkyō, ed., Bukkyō daijiten, 10 vols. (1933–36; reprint, Tokyo: Sekai seiten kangyō kyōkai, 1974–77), vol. 3, 2063. His biography is based primarily on the Gonshōsozu Shinshō fuzokujō excerpted in NCSS-jsh 3 (shiryō 2), 35; and in Takeuchi Rizō, Heian ibun (Tokyo: Tokyodō shuppan, 1965), 55ff. Shinshō was made bettō (superintendent) of Tōdaiji, Nara. in 840 and rose to higher ranks in subsequent years, ending with the administration of Kanshinji and Zenrinji in Kyoto. Jitsue was named the head disciple in Kukai's Last Testament, the Goyuigō (attributed to Kūkai but apparently written in the tenth century). Jitsue administered the building and artistic activities at Tōji and Kongōbuji on Mt. Kōya after Kūkai's death in 835. The Gonshōsozu Shinshō fuzokujō records that it was Jitsue who began Kanshinji, in 827, with Shinsho continuing its development thereafter. Gonshōsōzu Shinshō fuzokujō, cited in NCSS-jsh 3 (shiryō 2), 35. On Jitsue. see Mochizuki, vol. 2. 1932. Despite conflicting versions of the temple's origins, it can be surmised from these accounts and others that the site of Kanshinji was Shinshō's private hermitage (probably recommended by Jitsue) from the year 827.

43. Literally. “set-amount temple,” that is, a temple on a fixed stipend. Such monasteries, officially recognized by the state, received a fixed subsidy from the government (see n. 39 above), derived revenue from their own landholdings (rather than depending entirely on state support), and were allowed nenbundōsha, or annually appointed ordinands, for each sect. See Hiraoka Jokai, “Heian bukkyo no seiritsu to hensen,” in Ronshū nihon bukkyōshi, Heian jidai, ed. Hiraoka Jōkai (Tokyo: Yūzankaku, 1986), 3–30; and idem, 1981 (as in n. 39), 417ff.

44. NCSS-jsh 3 (shiryo 4), 39. The land grant, to become part of the temple's estate, is dated 869.6.9, from the office of Public Affairs (Minbusho); the jōgakuji was awarded on 869.6.13; see NCSS-jsh 3 (shiryō 4), 36 and 39–40, respectively. Shinsho provided a capsule history of the temple in his petition, which is excerpted within the Register, stating that Jitsue established the Kanshinji for the benefit of the nation and that he recognized the local administrator of the ancient Kawachi Province as the temple's head, or bettō.

45. Established routes to the west and north had long led traffic between the major Inland Sea port of Naniwa and the 7th-and 8th-centnry capitals at Asuka and Nara along the valley above which Kanshinji is situated. By the 9th century there were at least three known routes through the Kawachi district, leading to and from the Yamato plain, and from there north to Kyoto; or south to mountainous Mt. Kōya in Kii Province (modern Wakayama Prefecture) via ancient Izumi Province (modern Ōsaka-fu) through the Kimi Pass; or to Osaka bay. These same routes could be used to reach the important religions worship sites at Mt. Kimpu (Yoshino Shrine), Mt. Makino, and Nachi. Although not mentioned in the Register, the founding of Kanshinji is traditionally associated with the miracle-working 7th-century mendicant monk En-no-Gyōja (En no Ozuna), who probably did travel the area. See Tamura (as in n. 10), 117–22, esp. 118. The hermitages (and, later, temples) along the Kanshinji route served as lodgings for travelers. At the same time, the region has rich associations with the sacred dragon spring (noted as the north border of Kanshinji in the Register, above) and its indigenous (Shintō) gods, called kami. In waterfall, also mentioned in the boundaries, was also a kami worship site.

46. Kūkai states in a letter to Emperor Saga that he discovered Mt. Kōya during his time as an ubasoku, that is, the period between his leaving the central university at age twenty-four, in 797, and his departure to study in China in 804. See KZ, vol. 3, 524 (Seireishū, fasc. 9). Kukai's historical relationship to the region is clouded by centuries of legend.

47. Because of an influx of Korean immigrants in the 5th century the Kawachi district was a commercial, cultural, and Buddhist center by the 6th century. The ancient chronical Kojiki notes that Kawachi Asuka flourished during the reign of Emperor Kenzō (485–87). On the region and its history, see Naoki Kōjirō, “Kawachi Asuka no rekishi,” and Sakai Takashi, “Kawachi Asuka no tera,” in a special issue on Kawachi and its temples, Bukkyō geijutsu 119 (Aug. 1978): 11–21 and 116–33. The richness of popular lore in the region attests to mixed religious practices. Kūkai acknowledged and supported the combined worship of kami and the Buddhist divinities; he documented the ascent of Mt. Futara (Nantaizan, Nikkō) by the priest Shōdō (735–817) to pray to the kami of the mountain and seek enlightenment. The text has been translated by Allan Grapard in “Kūkai: Stone Inscription for the sramana Shōdō, Who Crossed Mountains and Streams in His Search for Awakening,” in The Mountain Spirit, ed. Michael Tobias and Harold Drasdo (New York: Overlook Press. 1978), 50–59. On syncretic beliefs and shrine-temple histories, see several works by Kuroda Toshio, especially Jisha seiryoku (Tokyo: Iwanami shoten, 1980) and Ōbō to buppō (Kyoto: Kōzōkan, 1983). In English, see Allan Grapard, Protocol of the Cods (Berkeley: University of California Press, 1992).

48. Pilgrimages devoted to Kōbō Daishi (Kūkai) developed in the areas around Mt. Kōya and on the island of Shikoku from the 12th century. On the latter, see Ian Reader, “Legends, Miracles, and Faith in Kōbō Daishi and the Shikoku Pilgrimage,” in Religions of Japan in Practice, ed. George Tanabe (Princeton: Princeton University Press, 1999), 360–69.

49. NCSS-jsh 3 (shiryo 5 4), 36. A bay is one ken, an ancient linear measure equaling 6 shaku, or 71 ½ in. (1.82 m; see n. 72 below). Documents for this period often describe a structure as a certain number of bays with aisles. The bays in this case refer to the length and width of the mōya, or chancel; the calculation of the total size depends on whether there are aisles all around (four sides) or only on one or two sides. The Nyōhōdō is three bays (that is, its chancel) with an aisle on all four sides, making it a five-bay-square structure; the lecture hall is five bays (its chancel) with an aisle on all four sides, and thus a large seven-bay-square structure; the fire-ritual hall, given as six bays with an aisle on one side, is thus six by seven bays. The number of doors is also noted in some cases. The nonstandard long ō for Nyō in Nyōhōdō is correct, according to Abé (the term is not found in MD).

50. NCSS-jsh 3 (shiryō 4), 35–38.

51. Fujii Keisuki, Mikkyō kenchiku kūkanron (Tokyo: Chūō kōron bijutsu shuppan, 1999), 47. Nonetheless, given its remote location, the plan is considerable.

52. First mentioned by Nishikawa Shinji in his 1956 article (as in n. 9), two different documents cite a request by the Kanshinji sangō (clerical council) that a bell be cast for the temple. The petition is noted in the Kanputō hennen zasshū (Collected government annals) and the Zoku Kōbōdaishi nenpu (Chronology of Kōbō Daishi, continued) in Jōwa 7 (840).7.27. For the fullest description, see Nishikawa S., Aug. 1978 (as in n. 10), 65–66.

53. Typically, the main hall (Kondo), lecture hall (Kōdō), sutra repository, and dormitories of a monastery were constructed first. Pagodas were also important structures but Kanshinji had a banner instead, probably due to the expense and complexity of building a pagoda, also possibly due to iconographic requirements. At Tōji, the main hall and several dormitories were completed initially, followed much later by the lecture hall and pagodas.

54. Main halls existed at the 9th-century Esoteric temples of Tōji and Jingoji prior to their designation as Shingon sect temples. Both subsequently retained traditional Healing Buddha (Yakushi) statues as the main icon.

55. The first abhiseka hall (Kanjōdō) in Japan was established at the important Nara monastery of Tōdaiji in 822. That same year Kūkai initiated the abdicated emperor Heizei, probably as part of the opening ceremony for the hall (for the document, Heizei tennō kanjōmon [alt. Kanjōbun], see KZ, vol. 2, 117–45); trans. Allan G. Grapard, “Precepts for an Emperor,” in White, 146–64. Later Esoteric monasteries sometimes had a mandala hall (Mandaradō). On 9th-century Esoteric halls, see Fujii (as in n. 51), esp. 15–61. In English, see Nancy Shatzman, “The Mizong Hall of Qinglong Si: Space, Ritual, and Classicism in Tang Architecture,” Archives of Asian Art, no. 49 (1991): 27–50, which discusses a Tang Chinese precedent for early, Japanese initiation halls.

56. A hall of the same name was built on Mt. Hiei, at the chief monastery of the Tendai school, Enryakuji. Possibly related to its genesis are practices such as the Nyōhōkyō and the copying of the Lotus sutra, although these do not appear to have been conducted at Esoteric halls or the Kanshinji structure. See Mochizuki (as in n. 42), vol. 9, 4140–41.

57. There is a bronze Shaka among a group of four such icons in the collection of Kanshinji dating to the second half of the 7th century. According to Japanese scholars the Shaka is not documented in the temple's history and thus cannot be equated to the Shaka image listed in the Register. See Tamura (as in n. 10), ill. and text, 122; Tamura dates it by style to the Tenchi or Tenmei era and notes the lack of documentation about the icon. It would not be uncommon to house an image made in the region at an earlier date in a newly built temple hall, nor would its function as a small votive icon in a relatively small hall filled with paintings (typically brought out only for particular rites) be at odds with other such images in early Japan.

58. A fuku is a traditional Japanese measure for cloth (width). In the 9th century I fuku measured 21 to 22 inches (54.0 to 56.2 cm). Fujii (as in n. 51), 47, notes that such a large painting in a relatively small hall is notable (without suggesting why).

59. The sutras typically refer to the Nyoirin by this appellation.

60. The myōō are bodhisattvas and they achieve both group (typically five) and individual deity status within the Esoteric tradition (for a myōō bosatsu statue, see Fig. 18).

61. Central to these was the abhiseka. Kūkai performed three abhiseka as part of his relatively brief training in China. There are several types of initiation, including that in which the lay public may participate (such as kechien kanjo, or “binding karmic affinity”). Abé 124, discusses the three levels identified in Subhakarasimha's Commentary on the Mahāvairocana Sūtra. The kechien kanjō is the first of three levels of initiation rites, at the close of which the participant is given the mantra for his or her own personal deity. The intermediate level is the “studying of the dharma,” or gakuhō kanjō after which the priest is taught to ritually invoke his deity, through mantra, mudra, and eidetic meditation, and in some cases to study yogic exercises for additional deities of the pantheon, allowing him to visually construct the mandala. The highest level of initiation rite is the denpō kanjō (alt. denkyō kanjō), “transmitting the teaching.” For an excellent summary in English, see David L. Gardiner, “The Consecration of the Monastic Compound at Mount Koya by Kukai,” in White, 119–30.

62. Abé, 122. The Womb World mandala is thought to be based on chap. 2 of this sutra. The term kai (world) is not used after the term taizō (womb, or matrix) in the Mahāvairocana sutra, but Taizōkai came into usage among Tendai Esoteric priests during the second half of the 9th century and spread into common usage. The Womb World deities do not appear to have had great currency in Tang China, according to extant texts and the archaeological record. Although Kukai brought back to Japan the Two Worlds concept and mandala, and they were almost certainly transmitted to him by his Chinese teacher, Huiguo, the Chinese history of mandala use remains unclear.

63. The Sanbōinryu (branch or lineage) of Shingon originated with the priest Shōbō (832–909) at Daigoji and, along with the branch associated with Mt. Kōya (and Tōji), the Chūinryu, constituted the earliest subsect of the tradition. Both enjoyed imperial patronage. Examples of Sanbōin lineage Nyoirin texts include the Nyoirin hōshuhō, Nyoirin kanjisaibosatsu nenjuhō, Kanjisaibosatsu Nyoirin nenjugiki, and the Nyoirin yugahō, among others; see MD, 1739–40. In the Sanbōin lineage of Shingon practice, the Shō Nyoirin Kanjizai bosatsu nenjū shidai is the first of four important precept rites (Shidō kegyō shidai) and features the Nyoirin Kannon as its visualized honzon. Over time, the sources for these two primary Shingon lineages have blurred, making a conclusive evaluation impossible. It is not clear which rites were practiced at Kanshinji during the 9th century, although later the temple and its rites would be associated with the Mt. Kōya lineage (Chūin), not the Sanbōin and its rites. For the Sanbōin lineage of Esoteric rituals, see Takaii Kankai, Mikkyō jiso taikei—tokuni Sanbōin kenshinpō kichō ta shile (Kyoto: Takaii Zenkashu shakka kankōkai, 1952); for the Chūin lineage, see Toganoo Shoun, Himitsue jisō no kenkyū, vol. 2, Toganoo Shōun zenshū, ed. Kōyasan Daigaku mikkyō bunka kenkyūsho (Kyoto: Rinsen shoten, 1935).

64. From his reading of the iconography of the paintings or the reading of the characters for Nyōhō, Konno (as in n. 10), 157–58, presents several hypotheses, but none is particularly convincing, nor are there known occasions of these rites in the early Heian period. Going by the prominence of Womb World imagery, Konno believes that the rite of the Womb World (Taizōkaiho) may have been performed in the hall, or perhaps a rite called the Nyōhō taizōkaihō, which could have featured the Nyoirin Kannon, although his suggestions lack the support of other examples. Rites featuring wish-fulfilling jewels and related deities, including the Nyoirin Kannon, abound in the Shingon tradition. Some originate in the 10th century with the Shingon master Shunnyu, who was associated with Ishiyamadera and its Nyoirin Kannon. See the Ishiyama shichi shu, 143b–c, and the Yōsōn dōjo kan, 49c–50b, as noted by Brian D. Ruppert, Jewel in the Ashes: Buddha Relics and Power in Early Medieval Japan (Cambridge, Mass.: Harvard University Asia Center, 2000), 147 n. 18, 428. Ruppert's work is an excellent discussion of jewel iconography and worship as it relates to rites for relics.

65. The relevant rite, Fugen enmei-ho (alt. Fugen enmy ō-hō), is based on the Kongō jumyō darani-kyō, T, vol. 20, text no. 1134B, and is described in the Kongō jumyō darani nenjuhō. T. vol. 20, text no. 1133, and the Kongō jumyō darani-kyōbō T. vol. 20, text no. 1134A, as noted by Abé, 354, 526 n. 86; and Allan G. Grapard, “Religious Practices,” in The Cambridge History of Japan, vol. 2, Heian Japan, ed. Donald H. Shively and William H. McCullough (Cambridge: Cambridge University Press, 1999). 542. These rites were popular throughout the Heian period, increasingly in the private sphere. See Mimi Yiengpruksawan, “In My Image: The Ichiji Kinrin Statue at Chusonji.” Monumenta Nipponica 46, no. 3 (autumn 1991): 329–30.

66. See Misaki Ryōshū, Taimitsu no kenkyū (Tokyo: Sōbunsha, 1988), as noted by Grapard (as in n. 65), 532 (he cites the Fugen-e dambo).

67. There are no records before the 10th century that provide information about the ritual components of this annual rite; by the 10th century, we know that relics worship was central to it. See Ruppert (as in n. 64), esp. 102–41.

68. The Jimmyōōin shows (right to left) Fdō myōō, Gōzanze myōō, Hanyaharamitta bosatsu, Daiitoku myōō, and Shozanze myoo (the latter is not one of the Godaison).

69. Kokūzō bosatsu nōman shogan saishoshin darani gumonjihō, T, vol. 20, text no. 1145.

70. In ancient times, one hundred days; today, fifty days. For a description of the rite in English, see Yamasaki (as in n. 291, 182–90.

71. NCSS-jsh 3, 36, 43. no. 30–5.

72. A traditional linear measure. One shaku is just under a foot (30.3 cm).

73. Three fuku equal about 65 in. (165 cm). The Chinese characters for Birushana (alt., Rushana) typically denote Vairocana Buddha of the Kegon-kyō (Avatamsaka sūtra). It may refer here to Dainichi or the Esoteric Vairocana of the two primary Esoteric Shingon texts, the Sarvatathāgatatattva-samgraha or Vajrasekhara sūtra (Jpn: Kongōchō-kyō or Diamond Peak sutra), T, vol. 18, text no. 865; and the Mahāvairocana sutra (Dainichi-kyō), T, vol. 18, text no. 848. Without visual evidence or a known iconographic plan for the altar, the identity of the deity is unclear.

74. The statue of Butsugen butsumo Buddha is known popularly as Miroku bosatsu, and the Miroku Buddha is known popularly as Hōshō nyorai. These attributions may have begun in the 14th century. Our identification of the statues is based on early iconographic sources (for the Butsugen butsumo, see Takakusa Junjirō and Ōno Genmyō, eds., Taishō shishū daizōkyō, Zuzō [Iconography]. 12 vols. (Tokyo: Daizō Shuppan kabushiki kaisha, 1932–341, vols. 3–4, for example, the Kakuzensho, Shosonzuzo, Daigojibonzuzō, Shikashozuzō, and, for Miroku, the Shikashozuzō) and on an inscription found on the base of Butsugen butsumo statue.

75. The Fudō myōō (right facing) is illustrated in Mae and Nagashima, pl. 21, description, 119–20. It ia a seated figure 37 in. (93.9 cm) tall. The seated Aizen myōō (left facing) is illustrated in the same volume, pl. 21, description 120–21, and is 42 5/8 in. (108.5 cm) tall. Both are designated Important Cultural Properties (Jūyō bunkazai). See also Bunkacho and Mainichi shinbunsha jūyō bunkazai iinkai jimukyoka, eds., Jūyō bunkazai, 32 vols. (Tokyo: Mainichi Shinbun, 1972–77. vol. 3 [chōkoku III], 98, no. 445, and 118, no. 540.

76. Mae and Nagashima, pls. 49–52; and Jūyō bunkazai (as in n. 75), vol. 4 [chōkoku IV], 5 (nos. 54–56). They are made of wood in the single-woodblock technique and range from 56 ½ to 59 ½ in. (144 to 151 cm), standing. All are designated Important Cultural Properties.

77. Mizuno; Nishikawa S.; Tanaka; and Tamura (all as in n. 10).

78. Except for that of Anshōji, in which the arrangement of Five Wisdom Buddhas (Gochi nyorai) is orthodox.

79. Only Kuno, 1973 (as in n. 10), dates it later, and Itō allows that the Nyoirin Kannon may have been part of a honzon group. Ito believes that the Nyoirin Kannon was made between 850 and 869, and that the other statues, including the Miroku and Butsugen butsumo, were completed at the same time.

80. Kuno, 1974 (as in n. 10), 214–17, is among the few authors to discuss the Butsugen butsumo and Miroku Buddha statues in his research. He dates them to 834–48. Shimizu Zenzo (discussion with author, May 1990) believes the two extant Buddha statues date to the 10th century. Other scholars date them to the second half of the 9th century, a decade or more later than the Nyoirin Kannon.

81. Three shaku, 6 sun, 1 bu, given in traditional measurements.

82. The lotus pedestal is the original mate for the statue, but the mandorla (kohai—although it dates to the 9th century—is from another statue.

83. Hinoki is Chamaecyparis obtusa; kaya is Japanese nutmeg, Torreya nucifera, commonly called the California nutmeg tree. Mizuno Keizaburō noted in a conversation with author (Mar. 1996) that conclusive wood tests have not been conducted on the Nyoirin Kannon statue, although most sources, including the definitive technical study of the work by Mizuno and Nishikawa Shinji, in NCSS-jsh 3, state that the material of construction is kaya. The latter is a rarer wood usually used for special types of Buddhist icons.

84. The lower arms, outermost leg and foot sections, and panels covering the hollows are made from separate pieces of wood, but all original wooden parts of the statue appear to be from the same tree. For a complete technical description, see NCSS-jsh 3, 7–13, and for diagrams on the statue's construction, 22–28. Ichiboku zukuri is a slightly misleading term to the nonspecialist when, as here, more than one piece of wood is used. Technically, if the torso of the statue, excluding the feet, arms, or even the head, is made front a single block of wood, the term may he applied.

85. The color combinations (ungen saishiki) and applied gold patterns show little retouching. The pigmentation has been documented in NCSS-jsh 3, 8–13; see also colorpls. 6–21. Also original to the statue and well preserved are a wooden crown made of hinoki covered with lacquer and gold paint. Only the wooden hand-held attributes, second left hand and third right wrist, byakugo (iconographic mark on the forehead), and some of the bracelets and arm jewelry are later replacements. On the repaired and replaced sections of the statue, see NCSS-jsh 3. 18–19.

86. A section of the dais for each of the two Buddhas has been replaced since the 9th century, but without affecting the original height or structure. NCSS-jsh 3, 29–31.

87. Konno (as in n. 10), 158.

88. Until recently, color photographs showing the two Buddhas were not easily obtained, further indicating the modern scholar's lack of interest in them.

89. NCSS-jsh 3. 16; Inoue (as in n. 4) concurs.

90. Wife of Emperor Junna (786–840, r. 823–33).

91. This record is first noted in NCSS-jsh 3 (bikō), 20, which Nishikawa co-edited, and is further discussed in his articles of 1978. The land bequest was in nearby Furuichi-sho, Kawachi.

92. Sec Friedrich Nietzsche, Werke vol. 3 (8041, as discussed by Jonathan Culler, On Deconstruction: Theory and Criticism after Structuralism (Ithaca, N.Y.: Cornell University Press, 19821, 86.

93. Some sources refer to the technique, as mokushin kanshitsu, or wood-core dry lacquer, but recently the use of this term is increasingly limited to works in which the wood core is unfinished in many areas, relying on lacquer for modeling. The wooden Kanshinji statues were nearly completed to the surface, then coated with lacquer.

94. It is uncertain where the Kanshinji images were constructed. There may have been a workshop on the monastery grounds (although the Register does not mention one) or they may have been made in the capital and transported to Kanshinji. The fact that a set of Five Wisdom Buddhas was sculpted at Kanshinji around the year 847 (to be discussed below) points strongly to the former. It is notable that despite its remote location such a workshop was in place.

95. Recent research on the Yakushi Buddha at Jingoji and other jingūji, or state temple-shrine complexes, suggests that this type of Yakushi may have been made for sites with deep links to indigenous kami worship. See Nagasaka Ichiro, “Shoki jingūji no seiritsu to sono honzon no imi,” Bijutsu kenkyū 354 (Sept. 1992): 1–18. Kanshinji, as a mountainous site, would have had such links. The honzon of the Mt. Kōya Kongōbuji main hall is today a Yakushi. In the 9th century (although the date of completion is debated, with the earliest record indicating 968) the main icon of the lecture hall (today's main hall) was Ashuku, the Esoteric equivalent of the Buddha of Healing.

96. Many sources identify this Buddha's attribute as a “medicine jar,” but its fundamental form and meaning indicate it is a reliquary that has the power to heal.

97. NCSS-jsh 3, 37. The two bowls were probably Chinese Xing or early Ding ware; in China porcelain was only just beginning to be considered a material precious enough for Buddhist ritual use, as we know from bowls sealed in the year 880 beneath Famensi pagoda, about 95 miles from modern Xian, China, as part of the equipment used in relics rituals, in English, see Roderick Whitfield, “Esoteric Buddhist Elements in the Famensi Reliquary Deposit,” Asiatische Studien (Études Asiatiques) 44, no. 2 (1990): 247–57; idem, “The Significance of the Famensi Deposit,” and Zhu Qi-xin, “Buddhist Treasures from Famensi: The Recent Excavation of a Tang Underground Palace,” Orientations 21 (May 1990): 84–85, and 77–83; and Patricia Eichenbaum Karetzky, “Esoteric Buddhism and the Famensi Finds,” Archives of Asian Art 157 (1994): 78–85. The finds were first reported in Wenwu, no. 10 (1988): 1–56.

98. NCSS-jsh 3, 20.

99. We do not know the origins of this seal, only that it dates to some time before the Register (883). Mizuno, who has seen the original document, states that it is beneath the writing (conversation with author, Apr. 1997). The seal is illustrated in Kokushi daijiten henshu iinkai, ed., Kokushi daijiten, vol. 3 (Tokyo: Yoshikawa Kobunkan, 1983), 850.

100. The common liturgical sources list sri, ga, gam, bu, and kham. I hope to further research ritual texts on Butsugen butsumo, for if sri is used in a particular rite a more exact analysis of the character of Kanshinji and its deities might be possible. The Nyohōbutsugenhō rite may be of relevance here. See MD, 1748.

101. See NCSS-jsh 3 (biko), 20.

102. On the nature of yogic meditation practices and the meaning of kan, see Shaft, 2001 (as in n. 23).

103. This idea is set forth in Kukai's The. Meanings of Sound, Word, and Reality (Shōji jissōgi), The Meanings of the Word Hūm (Unjigi), and Attaining Enlightenment in the Very Existence (Sokushin jōbutsugi). For translations of the texts, see Hakeda (as in n. 29), 234–45, 246–61, 225–33. Abé's title translations differ.

104. The most comprehensive description of these rites available to the researcher is found in Toganoo (as in n. 63), 33–96. They consist of the Eighteen-Stage rite (Jūhachido nenju kubi shidai), Diamond World rite (Kongōkai nenju shidai), Womb World rite (Taizōkai nenju shidai), and Goma (fire) rite (Sokusai goma shidai) to the deity Fudō. The mantra of Butsugen butsumo begins and ends the Sannenju (Additional recitation) sections of the four rites, which points to the significance of the deity's generative function and context. This is not necessarily the case with other major and more advanced ritual practices like those in the Rishukyō; thus, Butsugen butsumo functions in a spiritually particular and important way in the Shidō kegyō text. The Nyoirin Kannon features in these rites as well (see n. 145 below).

105. T, vol. 18, text no. 848, and the commentary, vol. 39, text no. 1796.

106. A similar interpretation (with the appearance of some of the same deities) is offered by Abé in his analysis of mantras used in one component of the Latter Seven-Day rite at the Shingon chapel in the imperial palace (Abé, 351–53).

107. As Michel de Certeau wrote, “The situation of the historiographer makes study of the real appear in two quite different positions within the scientific process. … On the one hand, the real is the result of analysis, while on the other, it is its postulate”; Certeau, The Writing of History (New York: Columbia University Press, 1998), 35.

108. This condition is recorded in the eighth month of 839 in the Shoku Nihonkōki, entry, for Jōwa 6 (839).8.4. Kuroita Katsumi, ed., Nihon kokushi taikei, vol. 3 (Tokyo: Yoshikawa kobunkan. 2000).

109. Inoue (as in n. 4), 29, suggests that the female appearance of the Nyoirin Kannon and the fact that it holds a nyoiju jewel (of importance to Kukai) attracted the empress's attention.

110. No one has as yet suggested this reading for the phrase “Saga-in Taiko Taigo Gogandō.”

111. See Hayami Tasuku, Heian jidai kizoku shakai to bukkyō (Tokyo: Yoshikwa kōbunkan, 1975).

112. The consecration took place in Jōwa 8 and is noted in “Hot Saga kanjōbun,” in Tōbōki, vol. 4, as noted by Ito, 119.

113. Shinsho, in the Montoku jitsuroku, as quoted by Ito, 119.

114. Ibid.; and NCSS-jsh 3, 50.

115. In the Montoku jitsuroku, as quoted by Itō, 117.

116. If one or more of the lecture hall statues was completed as a vow around the time of Nimmyo's death in 850, the five new Buddhas in 857 would constitute sustained art-making activity for the temple over more than a decade. This may be more plausible than the comparatively erratic chronology presented by other scholars: the Nyōhōdō and its images, the lecture hall and the Nyoirin Kannon (in addition to others?) by 840; then, ten years later, the Miroku and Butsugen butsumo statues along with other numerous lecture hall statues; and finally, the five new Buddhas in 857.

117. This possibility and iconographic correlation have not been noted previously in the literature.

118. Sandai jitsuroku, in NCSS-jsh 3 (shiryō 1), 35. Zenrinji became a state-sponsored jōgakuji in 863, six years before Kanshinji achieved that status. Shinshō bequeathed it to his disciple Shūei in 867. It may have some bearing that on the statues' completion, Shinshō wrote, “the temple [Kanshinji] is so secluded in the depths of the mountains that it is difficult to maintain a residence there. I am concerned that it will fall to ruin in later generations.” Shinshō's explanation is possibly an excuse for moving the statues created for Kanshinji to a temple nearer the capital. Shinshō's affiliation with Zenrinji grew strong during the late 850s, when he made it his own training center. Zenrinji also had strong ties to the powerful Fujiwara nobility.

119. He died in 873.7.7, and the grant was made on 874.7.9. This proposal is strengthened if we recall that a bell was cast in honor of the deceased priest Shinshō at Jingoji in the following year, Jogan 17 (875).8.13, as noted in NCSS-jsh 3, 50.

120. The deity also holds prayer beads and a lotus bud. The carved wooden attributes postdate the statue but presumably correspond in form to the original items used.

121. Kanshinji sankei shodō junreiki (henceforth, Pilgrim's Account), by the priest Kenki, noted above and as quoted in NCSS-jsh 3 (shiryō 6). 45.

122. Hyperbole about special icons is common in pilgrimage accounts, and it is advisable to read the account with this understanding.

123. it was formerly called the Konpon godō (main hall) and was “recently” rebuilt and called the Hondō (main hall). NCSS-jsh 3 (shiryō 6), 45.

124. According to Adachi (as in n. 10), 102. This date is based on a roof tile inscription. It is possible that portions of the present main hall were lost and the tiles reused, as there is a temple fire recorded in 1462 (Kansai 3). Most literature dates the hall to the Nanbokuchō era (1332–92) based on earlier evidence, but 1439 is correct.

125. See Yamagishi Tsuneto, Chūsei jiinshakai to butsudō (Tokyo: Kōshobokan, 1990), 130–32; and Alexander Coburn Soper III, The Evolution of Buddhist Architecture in Japan (Princeton: Princeton University Press, 1942; reprint, New York: Hacker Art Books, 1978), 249–51.

126. NCS-jsh 3 (shiryō 6), 45.

127. For the 13th- and 14th-century history of Kanshinji, see Nagashima Gyozen, “Kanshinji no rekishi,” in Mae and Nagashima, esp. 91–108.

128. Neither is the 9th-century Chinese priest portrait statue mentioned in the Pilgrim's Account; today it is enclosed within a zushi dating to the 15th century in the temple's museum.

129. The Sino-Japanese term makes no distinction between stūpa (henceforth, stupa) and pagoda. The Tahō tō is the Prabhūtaratna-stūpa. According to Ruppert (as in n. 64), 66, 402 n. 80, the construction and worship of these “jeweled stūpa” in Shingon and Tendai temples proliferated by the 10th century and proved to be among the most important ritual influences on the development of the offering of Buddha relics. The building of stūpa (pagodas), including both stone monuments and wooden-construction pagodas, for karmic merit was promoted in texts brought to Japan by Kūkai.

130. In a record concerning Emperor Gomurakami from the year 1360 (Shōhei 15), the Gomurakami tennō rinshi, in the Tōjichōja Raii bussharihōnō hōgyōjō, a section dealing with the priest Raii and the rite of relics, it is stated that the relics from Tōji were presented at the Kanshinji Hondō, Hōtō (pagoda), Karitei altar, and to the Benzaiten deity; from this we know that the pagoda existed by 1360. The 1378 Pilgrim's Account indicates that it is a three-story pagoda. The 1462 fire destroyed the whole temple. See NCSS-jsh 3, 31.

131. The 1378 Pilgrim's Account notes “Amida Miroku” as the honzon of the pagoda. A 1669 record notes “Amida” as the honzon of the pagoda. A 1733 record notes that “four directional Buddhas” were installed. The Shaka nyorai Buddha, a wooden statue, dates between the 10th and 11th centuries; the wooden Yakushi Buddha to the 13th century. All four Buddhas are currently housed in the Kanshinji treasure hall (museum). The name of the pagoda, Tatekake—literally, “under construction”—is, according to temple legend, a reference to the interruptions that occurred during its construction.

132. I have been unable to locate records of the Buddhas' installation in the pagoda.

133. It is odd that the Butsugen butsumo Buddha received the new appellation of Miroku (as it is commonly known today) and not Hosho, “One Arisen from the Jewel,” since the two share the nyoi jewel as their symbolic attribute. Further investigation of the (late medieval) traditional appellations of the two Buddhas might add clues to the changing meanings of the temple icons over time.

134. Paraphrased from Abé, 349. On the Latter Seven-Day rite, see Abé, esp. 344–55; and Ruppert (as in n. 64), esp. 130–35.

135. The document is the Gomurakami tenno rinshi, in the Tōjichōja Raii bussharihōnō hōgyōjō (see n. 130 above), dated 1360 (Shōhei 15), cited in NCSS-jsh 3, 31.

136. On the Nanbokucho history of Kanshinji, see Kokushi daijiten (as in n. 99), vol. 3, 850; and Nagashima (as in n. 127).

137. The first recorded mention of this event and appellation is found in the Kanshinji engi jitsurokuchō (copied 1394, with a date of 827) within a spurious record dated Jōwa 4 (837). See also Kenki's 1378 Pilgrim's Account, in NCSS-jsh 3 (shiryō 6), 45–46 (term found on 45). The phrase “Seven-star Nyoirin” is also used in the 1669 Kanshinji garan jiyakusōbōhōshiki hikae (in Kanshinji monjo, 556), excerpted in NCSS-jsh 3. 16 n. 7: and the 1733 Hinooanzoki, in NCSS-sh 3 (shiryō 6). 46–48 (phrase on 46).

138. For the Chūin lineage, see n. 63 above. The Kakuzenshō a work in 316 chapters. Reproduced in Takakusa and Ōno (as in n. 74), vols. 4–5.

139. 7; vol. 21, text no. 1307. The Chinese text is of uncertain date. The best discussion of Seven Stars worship and its history in east Asia is Henrik H. Soerensen, “The Worship of the Great Dipper in Korean Buddhism.” in Religions in Traditional Korea, SBS Monograph, no. 3 (Copenhagen: Seminar tor Buddhist Studies, 1995), 71–105. esp. 74–79. As Soerensen (75) explains, because the Chinese (apocryphal) sutra is mentioned in both the Kakuzenshō and the later Esoteric ritual compendium the Hakuhōkushō (Book of the white precious mouth), by Ryōzen (1258–1341), it probably dates to the Tang dynasty.

140. T, vol. 20, text no. 1091. The text is based on an edition from 1801, but the work is undoubtedly of earlier origin; see Soerensen (as in n. 139), 77. None of the literature on the Kanshinji statue mentions this text, although connections between the two are likely. Also of possible significance is the ritual for the nyoiju (jewel) and Ursa Major, the nyhōhokutoku (with the diety Kinrin buchō at the north position); see MD, 1748.

141. Grapard, 1999 (as in n. 65), 555; he discusses celestial worship and interpretations of celestial irregularities throughout this essay.

142. NCSS-jsh 3, 9. The authors do not mention it, but it is possible that the mandorlas of the two Buddhas were removed when they were housed in the pagoda. Why the Nyoirin Kannon required a new mandorla comes less readily to mind.

143. It may also be of interest that among the set of relic boxes found beneath Famensi pagoda (see n. 97 above) the fourth of seven nested caskets in the rear chamber features a six-armed Cintamanicakra-avalokiteśvara (Nyoirin Kannon); the smallest of the caskets contained the precious relic. Of great interest here is that the seven principal stars of the Northern Dipper are carved on tile interior surface of the innermost casket (a point that is not discussed in the Famensi literature). On the back central panel of the same casket is Vairocana; on the left side is Bhaisajyaguru (Yakushi)—both deities in the lecture hall inventory at Kanshinji. See Karetzky (as in n. 97), fig. 11, for an illustration of the Cintamanicakra-avalokiteśvara. We have already noted the paintings of the Five Great Storehouse bodhisattvas inventoried for the Nyōhōdō that may have been used in rituals associated with the stars. While many researchers interpret associations between the site or the statue and the constellations as local tradition of little relevance to Kanshinji's ancient history, it is profitable to consider the contents of such tradition, their textual sources, and the light they may cast on earlier, albeit unndocumented, beliefs or populations.

144. For a photograph, see Jūyōbunkazai, vol. 3, 58, no. 168. This maedachi (alt., maedate) today is called the “kokoromi no saku” (literally, “test model”), meaning that it was modeled after the earlier Nyoirin Kannon.

145. Neither the explanation provided here regarding the relationship between medieval relics and jewel worship with changing meanings for the Kanshinji statues nor that concerning the maedachi statue and the Nyoirin Kannon as hibutsu is suggested its the literature on the icon.

At first glance it would seem unlikely, then, that this variant of the rite was practiced at Kanshinji, even if it had existed in the 9th century. It is worth noting, however, that Emperor Daigo (r. 897–930, patron of Daigoji) was very active as a donor in the Kawachi region during the 10th century (Kokushi daijiten [as in n. 9], vol. 3. 850); the Daigo (Sanbōin) lineage may have been influential. It is also possible that the Sanbōin-style ritual was subsequently adopted because the Nyoirin Kannon had already achieved a canonized status by the 10th century. The late 9th century was a period of dispute between Tōji and Mt. Kōya; about 912 the priest Kangen (853–925) was appointed bettō simultaneously at Tōji, Kongōbuji, and Daigoji, which further increased the likelihood of shared practices and transmissions. The literature on Kanshinji to date has not explored these potential links between Esoteric ritual practices and the Nyoirin Kannon's popularity.

• 146. Contrary to popular belief, no hibutsu are recorded prior to the 10th century, although the practice may have begun before that time. One likely secret image from the mid-8th century is the Shūkongōjin statue in Todaiji's Hokkedō Hall. On hibutsu, see Mochizuki, 4326; and MD, 2068.
• 147. Elias Canetti, Crowds and Power, trans. Carol Stewart (New York: Farrar Straus Giroux. 1962), 290.
• 148. See several essays on relics and stupas by Gregory Schopen, Bones, Stones, and Buddhist Monks: Collected Papers on the Archaeology, Epigraphy, and Texts of Monastic Buddhism in India (Honolulu: University of Hawai'i Press, 1997); see also Kevin Trainor, Relics, Ritual, and Representation in Buddhism (Cambridge: Cambridge University Press, 1997); and Bernard Faure, “Relics, Regalia, and the Dynamics of Secrecy in Japanese Buddhism,” in Rending the Veil ed. Elliot R. Wolfson (New York: Seven Bridges Press, 1999), 271–87. The Vajrasekhara sutra, a central text in the Shingon Esoteric tradition, describes the secreting of the image Ākaāsagarbha (Jpn: Kokāzō bosatsu).
• 149. From Walter Benjamin's famous essay on the aura of objects, “The Work of Art in the Age of Mechanical Reproduction,” in Illuminations, ed. Hannah Arendt, trans. Harry Zohn (New York: Schocken Books, 1969; Frankfurt: Suhrkamp, 1955), 243 n. 5. More recently, Stephen Greenblatt calls this “wonder.” See Greenblatt, “Resonance and Wonder,” in Exhibiting Cultures: The Poetics and Politics of Museum Display, ed. Ivan Karp and Steven D. Levine (Washington, D.C.: Smithsonian Institution Press, 1991), 42. For a discussion of concealment in the temple, see Edmund Leach, Culture and Communication: The Logic by Which Symbols Are Connected (Cambridge: Cambridge University Press, 1976), 84–88.
• 150. Ernst Kris and Otto Kurz, Legend, Myth, and Magic in the Image of the Artist: A Historical Experiment (New Haven: Yale University Press, 1929), 72.
• 151. Catherine Bell, Ritual Theory, Ritual Practice (New York: Oxford University Press, 1992), 140.
• 152. Ibid., 109, the following sentence is paraphrased from 110.
• 153. Michael Ann Holly, Past Looking (Ithaca, N.Y.: Cornell University Press, 1996). I thank Alfred Acres for first directing me to Holly's work, enabling me to further develop my understanding of a work's affective qualities.
• 154. Hans Robert Jauss, “Literary History as a Challenge to Literary Theory,” in Toward an Aesthetic of Reception, trans. Timothy Bahti (Minneapolis: University of Minnesota Press, 1982), 3–45. Others have pointed out that Jauss's account presumes a degree of faith in the historical transcendence of value judgments; see Janet Wolff, Aesthetics and the Sociology of Art (Ann Arbor: University of Michigan Press, 1999). 35. The term is useful here, however, in understanding how the ascribing of value to a work turns on its encounter with the expectations of each period in which it is evaluated.
• 155. Sawa (as in n. 10), 47–48.
• 156. Saws, 1972, trans. Gage (as in n. 26), 147; idem, 1964, original (as in n. 26), 150–51, more precisely translated: “… tile six-armed figure has a relaxed form with a sensual beauty.… ”
• 157. As recent theoretical works on literature have noted, excesses and surpluses form the boundaries of our rhetorical and analytic concepts. Jacques Derrida explains in Of Grammatology trans. Gayatri Chakravorty Spivak (Baltimore: Johns Hopkins University Press, 1976), that social convention hides différence and excess; the excess disguised by social convention is also preserved by it. See also the work of Georges Bataille, for example, “The Notion of Expenditure,” in Visions of Excess: Selected Writings, 1927–1939, ed. and trans. Allan Stoekl (Minneapolis: University of Minnesota Press, 1985), 116–29.
• 158. The “sexual” element of Tantric or Esoteric cultivation differs from that of ordinary experience. Essays in White provide very helpful discussions about the energies and empowerments of Esoteric practice; see especially his introduction, esp. 9–18, the essay by Richard K. Payne, “Ritual Manual for the Protective Fire Offering Devoted to Manjusri, Chūin Lineage,” 489–508, and that by Gavin Flood, “The Purifications of the Body,” 509–20. Kūkai considered ritual experiences to be private and sacred. Although a sexual experience may be the only corollary for a description of the empowerment generated by ritual practices, the Esoteric experience is, according to practitioners, different in quality and duration.
• 159. Bernard Faure, “The Buddhist Icon and the Modem Gaze,” Critical Inquiry 24, no. 3 (spring 1998): 778.
• 160. Holly (as in n. 153) and others would argue that the response is (in part) organized by the work's structure and constitutes an affective order.
• 161. Nishimura Kōchō (as in n. 21), 37. Minagiru means “swollen, overflowing.”
• 162. Nishikawa, Aug. 1978 (as in n. 10).
• 163. Sawa Ryūken, Mikkyō no bijutsu, Nihon no bijutsu, vol. 25 (Tokyo: Kōdansha, 1964), 70. Sawa's choice of the word ayashii is interesting because the foremost meaning of ayashii is “doubtful or suspicious,” although it is also used in the sense of “mystical or mysterious” (thus, my rendering of the single word as “strangely mystical”). The particle made, “even” or “as much as,” modifies it.
• 164. Greenblatt (as in n. 149), 49; for a brief explanation of “wonder,” see n. 149 above.
• 165. Sexuality and sexual experiences figure in the Buddhist tradition to a degree that would surprise many readers. On the topic, see Bernard Faure, The Red Thread: Buddhist Approaches to Sexuality (Princeton: Princeton University Press, 1998); and White, 15–18. In the Esoteric tradition, ritual experiences are not usually recorded. Monks and nuns whom I questioned suggested that practices for the Nyoirin Kannon may result in an empowerment with sensations that find a descriptive (but not somatic) corollary in sexual orgasm, as above. This explains, perhaps, the many characterizations of the deity, by Shingon scholar-priests employing sexually charged vocabulary, as well as the abundance of such descriptions in the general literature on the Kanshinji Nyoirin Kannon and other Esoteric icons. Another way to understand the experience might be in the Kundalini tradition with its experiences arising from the increase in energy, flow along bodily channels.
• 166. See Grapard, 1999 (as in n. 65), 549 (no source given).
• 167. Nishikawa, Aug. 1978 (as in n. 10), 68. In Japanese. “sono hyōjō ya shitai ni minagiru mikkyōtekki kan'nōtekina jūsokukan wa. …”
• 168. The final date for “early Heian sculpture,” according to Mizuno, 1993 (as in n. 10), should correspond to the end of Daigo's reign (923–31). For two contrasting views, see Kuno, 1974 (as in n. 10), and the ongoing series of articles by Konno Toshifumi entitled “Heian chōkoku no seiritsu,” in Bukkyō geijutsu, beginning with pt. 1, no. 175 (Nov. 1987): 27–40. On the designation of Buddhist icons as “Buddhist art,” see John M. Rosenfield, “Japanese Buddhist Art: Alive in the Modern Age,” in Buddhist Treasures from Nara, ed. Michael R. Cunningham (Cleveland: Cleveland Museum of Art, 1998), 232–44.
• 169. Smith, 48.
• 170. By symptom I mean an inadvertently conveyed sign, in the Peircean sense. See Douglas Greenlee, Peirce's Concept of Sign (The Hague: Mouton, 1973).
• 171. Gayatri Chakravorty Spivak, “Scattered Speculations on the Question of Value” (1985), in The Spivak Reader, ed. Donna Landry and Gerald MacLean (New York: Routledge, 1996), 109. On 110 she asks, “What subject-effects were systematically effaced and trained to efface themselves so that a canonic norm might emerge?”
• 172. For a recent discussion of the terms etic (outside, of different origin) and emic (inside, sectarian), see Kenneth L. Pike, Emics and Etics: The Insider/Outsider Debate, Frontiers of Anthropology (Newbury Park, Calif.: Sage Publications, 1990). See also Gudrun Ekstrand, Developing the Emic and Eric Concepts for Cross-Cultural Comparisons (Malmö Sweden: Dept. of Educational and Psychological Research, School of Education, 1986).
• 173. Gadamer (as in n. 25), 258.

Abé Ryūichi, The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse (New York: Columbia University Press. 1999).

Itō Shiro, “Shingon mikkyō chōzōron,” in Jingoji to Murooji, vol. 8 of Shinpen Meihō nihon no bijutsu, ed. Ota Hirotaro et al. (Tokyo: Shogakkan, 1992). 99–141.

KZ: Inaba Yoshitake et al., eds., Kōbō daishi zenshū, vols. 1–3, 3d ed., rev. (Kōyasan, Wakayama Pref.: Kōyasan daigaku, Mikkyō Bunka Kenkyūjso, 1968–78).

Mae Toshio and Nagashima Gyōzen et al., eds., Kanshinji, vol. 2 of Koji junrei, Saikoku, ed. Inoue Yasushi and Sawa Ryūken (Tokyo: Tankōsha, 1981).

MD: Mikkyō Jiten Hensankai, ed., Mikkyō daijiten, 6 vols. in I (1931; reprint, Kyoto: Hōzōkan, 1983).

NCSS-jsh 3: Nishikawa Shinji, Nishikawa Kyotarō, and Mizuno Keizaburō, eds., Juyō sakuhin hen, vol. 3, Nihon chōkokushi kiso shiryō shūsei, ed. Maruo Shōsaburō et al. (Tokyo: Chūō kōron bijutsu shuppan, 1977).

Smith, Barbara Herrnstein. Contingencies of Value (Cambridge, Mass.: Harvard University Press, 1988).

T: Takakusu Junjiro and Watanabe Kaigyoku, eds., Taishō shinshū daizōkyō (the Buddhist canon), 100 vols. (1922–32; rev. ed., Tokyo: Taishō issaikyō kankōka. 1969).

White, David Gordon, ed., Tantra in Practice (Princeton: Princeton University Press, 2000).

PHOTO (COLOR): 1 Nyoirin Kannon bosatsu (bodhisattva), Heian period, mid-9th century. Kanshinji (photo: courtesy Kanshinji and The Mainichi Newspapers, Publishing Division)

PHOTO (BLACK & WHITE): 2 Nyoirin Kannon bosatsu, detail (photo: courtesy Kanshinji and The Mainichi Newspapers, Publishing Division)

PHOTO (BLACK & WHITE): 3 Kanshinji, main hall, with the Nyoirin Kannon bosatsu shrine at center, ca. 1950 (photo: courtesy Kanshinji)

PHOTO (BLACK & WHITE): 4 Kanshinji and surrounding hills (photo: courtesy Asuka-en)

PHOTO (BLACK & WHITE): 5 Kanshinji, main hall, with the Nyoirin Kannon bosatsu shrine open during annual viewing rite (photo: courtesy Kanshinij and Tankōsha)

PHOTO (COLOR): 6 Nyoirin Kannon bosatsu, detail, right leg (from NCSS-jsh 3, pl. 2; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō koōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 7 Nyoirin Kannon bosatsu, detail, petal of lotus pedestal (from NCSS-jsh 3, pl. 14; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō koōron bijutsu shuppan)

PHOTO (COLOR): 8 Butsugen butsumo nyorai (Buddha), commonly known as Miroku bosatsu, Heian period, mid-9th century. Kanshinji (from Nihon no Kokuhō [Tokyo: Shūkan Asahi hyakka, 1996]; photo: courtesy Kanshinji and Shūkan Asahi hyakka)

PHOTO (BLACK & WHITE): 9 Butsugen butsumo (from NCSS-jsh 3, suppl. 1, ill. 4; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 10 Miroku nyorai (Buddha), commonly known as Hōshō nyorai, Heian period, mid-9th century. Kanshinji (from NCSS-jsh 3, suppl. 2, ill. 1; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 11 Miroku (from NCSS-jsh 8, suppl. 2, ill. 4; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 12 Nyoirin Kannon bosatsu, detail (photo: courtesy Kanshinji and The Mainichi Newspapers, Publishing Division)

PHOTO (BLACK & WHITE): 13 Official Register and Inventory for Kanshinji (Kanshinji kanroku engi shizaichō), 883, section on the lecture hall (from NGSS-jsh 3, 43, ill. 30–5; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 14 Womb World mandala, one of the Two Worlds mandala, hanging scroll, ink and color on silk, 72% x 64 ½ in. (183.6 x 164.2 cm), Heian period, second half of the 9th century. Kyōōgokokuji (Tōji) (photo: courtesy Tōji)

PHOTO (BLACK & WHITE): 15 Portrait statue of a priest, Tang dynasty, China, 9th century. Kanshinji (photo: courtesy Mizuno Keizaburō)

PHOTO (BLACK & WHITE): 16 Kyōōgokokuji (Tōji), lecture hall, statues of the Heian period, ca. 839 (from NCSS-jsh, vol. 1, 154; photo: courtesy Tōji, Mizuno Keizahurō, and Chūō kōronsha)

PHOTO (BLACK & WHITE): 17 Butsugen butsumo (from NGSS-jsh 3, suppl. 1, ill. 1; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 18 Gōzanze myōō bosatsu, Heian period, ca. 839. Kyōōgokokuji (Tōji) (from NCSS-jsh, vol. 1, 13; photo: courtesy Kanshinji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 19 Yakushi nyorai butsu (Buddha), Nara, Nara period, late 8th century Shinyakushiji. (photo: courtesy Asuka-en)

PHOTO (BLACK & WHITE): 20 Official Register and Inventory for Kanshinji, detail showing seal (from Nihon no Kokuhō, vol. 3, 184; photo: courtesy Kanshinji and Shūkan Asahi hyakka)

PHOTO (BLACK & WHITE): 21 Kanshinji Register, transcription of seal (from NCSS-jsh 3, 20; altered by author with permission of Mizuno Keizaburō and Chūō kōron bijutsu shuppan)

PHOTO (BLACK & WHITE): 22 Kanshinji, main hall, 1439 (photo: courtesy Kanshinji and Asuka-en)

PHOTO (BLACK & WHITE): 23 Maedachi Nyoirin Kannon bosatsu, Heian period, 11th century. Kanshinji (photo: courtesy Kanshinji and Mizuno Keizaburō)

PHOTO (BLACK & WHITE): 24 Kongōsatta bosatsu, Heian period, ca. 839. Kyōōgokokuji (Tōji) (from NCSS-jsh, vol. 1, 1; photo: courtesy Tōji, Mizuno Keizaburō, and Chūō kōron bijutsu shuppan)

~~~~~~~~

By Cynthea J. Bogel

Cynthea J. Bogel (Ph.D. Harvard University, 1995) teaches Japanese art and architectural history at the University of Washington. She has published on Buddhist art, coauthored Hiroshige: Birds and Flowers (1988), and curated exhibitions of contemporary textiles and ukiyoe prints. She is writing a book on Esoteric image, and icon [Division of Art History, School of Art, Art History and Design, University of Washington, PO Box 353440, Seattle, Wash. 98195-3440, cjbogel@u.washington.edu].