Getting Away from ‘Religion’ in Medieval Japan

Abstract

The concept of ‘religion’ as modern, European-derived, and therefore problematic in premodern and Asian contexts is well established, but leaves us with a problem: if the church/state sacred/secular dynamic is a modern misconception even in England, as Fitzgerald argued, then how should we go about examining the central place of specific institutions, behaviour, and belief in the workings of medieval Japanese society that have formerly been classified or understood as ‘religious’? Abandoning ‘religion’ as a separate field of study from the ‘secular’ in Japanese history has the paradoxical effect of drawing attention to the pervasive centrality of activity, performance, mentality, and observance to every aspect of medieval life. Elements of practice, performance, and the sacred were essential, core, components of the functioning of public and private governance from the imperial system to local landholding. The great temple shrine complexes of the medieval period were centres of organisation, authority, and legitimacy, which are best understood not as ‘religious’ complexes which were also ‘economic’ and ‘political’ powers, but as institutions whose authority cannot be separated out into separate (modern) categories of ‘economic’, ‘judicial’, or ‘religious’ authority. Such distinctions cut across the deeply interconnected nature of law, landholding, family, lineage, place, and belief in the period, the networks and systems by which medieval life was ordered, but they also cut across the way that they were perceived by those living within them: the ways in which people thought, behaved, and interacted with each other. In order to understand the workings of what we think of as medieval Japanese society, we must understand these connected systems as composed of elements that might look ‘religious’ or ‘secular’ to modern eyes, but which were complementary, indivisible, even, in the period.

1. Introduction

Medieval Japanese history cannot be understood without understanding the fundamental importance of religion to society, the state, and medieval culture.

As a statement, this seems straightforward—despite expressing something important—but does so by proceeding from a shaky conceptual foundation which obstructs our understanding of the period. The problem is one of categorisation, but also of language. Discussing ‘religion’ as having a role in society or the state presupposes that these are separate categories, and that ‘religion’ exists as a concept which can be independently and perhaps universally defined. This, too, is a relatively un-contentious statement, as the pitfalls, even invalidity, of ‘religion’ as a concept have been increasingly explored this century in the global context, with recognition that ‘religion’ as a discrete category of human behaviour is a fundamentally modern and European-derived concept expressed in the work of Timothy Fitzgerald (2007), Daniel Duboisson (2019), and others.

The idea of religion as an entity, defined in opposition to the idea of corresponding ‘secular’ activity, is inherently modern and steeped in the development of modern academic disciplines in the context of European and American history. Put simply, the category of religion is a conceptual tool which was invented for a particular job, namely the investigation and understanding of societies in a certain time and place. As such, it is often simply not an accurate or helpful framework to apply to times and places outside this context. One must use the right tool for the job, and the category ‘religion’, as developed in the popular and academic understanding, is a tool which is fundamentally unsuited to the study of medieval Japan. This, too, is an easy statement to make, but does nothing to solve the problem: we are constrained by patterns of language as well as the history of disciplinary academic study in the modern period. Whether we reach for the word ‘religion’ or ‘宗教’, we are making use of a term which is inseparable from that European- and Christian-derived concept outlined above. This means that it is not sufficient to say that ‘religion’ and ‘government’ or ‘religion’ and ‘economy’ were not separate in medieval Japan, as we would still be reliant on the terminology of binary difference in attempting to explain the lack thereof. Admittedly, as Christoph Kleine (2013, p. 6) argues, ‘we cannot conclude from the lack of a semantic equivalent (the signifier) in a given language, that the thing or concept that the term refers to (the signified) is or was completely unknown’. This approach does not sidestep the problem entirely, however, as the concept itself remains one which is derived from a culturally-bounded context, and to apply it to premodern Japan risks once more casting it as universal (Horii 2021, pp. 217–18). Furthermore, even if one might reasonably use ‘religion’ as an analytical category even without a contemporary semantic equivalent, the effect of the linguistic and conceptual structures which did exist in medieval Japan on the conception of society by those living in it must be considered. As Jason Josephson contends, ‘[the] point is not that words have etymologies but that the emplacement of certain categories transforms their members in demonstrable ways’ (Josephson 2012, p. 2), or as Clinton Godart put it when talking about the transformations of ‘religion’ in the Meiji period, ‘categories are not simply containers of thought: they have an effect on the contents,’ (Godart 2008, p. 71). If medieval actors did not perceive their actions as ‘religious’ or non-religious, then how did they understand their actions and the interactions of institutions?

Debate has tended to centre on the extent to which the concept—and terminology—of religion can be employed before the late nineteenth-century, and while convincing arguments for (Pye 1990) and against (Fitzgerald 2007; Grapard 1992, p. 2) have been made, even these are primarily concerned with questions of early modern conceptions of religion, an entirely different kettle of conceptual fish to the worldview of the medieval period. The interdependence of aristocratic and monastic institutions as parts of a mutually supporting system is well established (Kuroda 1975, 1983; Adolphson 2000), as is the permeation of Buddhist conceptions of rule and the central role of monastic ritual within the imperial court in the classical and medieval period (Abe 1999; Conlan 2011). The potential for whether the secular (and therefore the religious) can be discerned in classical and medieval Japan has been explored by Kiri Paramore, who sees the public and private power of the aristocracy and the Buddhist clerical institutions as ‘institutionally, culturally, and often biologically linked realms’, but that the very nature of the close interactions between these realms served to reinforce the differences between them (Paramore 2017, p. 24). Paramore’s argument is that there was a clear conceptual difference between the institutions of the court and the temples, and that this necessitates a conceptual awareness of their functions as separate fields—the presupposition of distinct spheres necessitated by the negotiation of the boundary between them, in Kleine’s (2013, p. 27) terms. However, while we may draw a distinction between institutions, it is worth recalling Neil McMullin’s classic outline of the fundamental influence of ‘religious’ notions in the mental frame of reference of classical Japanese statehood itself (McMullin 1989, p. 13). Even if we are to accept an understanding of distinction in organization between the aristocratic and monastic, as Paramore suggests, this does not automatically imply an understanding of ‘religion’ versus ‘secular’ or ‘political’ organization, but rather the recognition of a distinction between Buddhist institutions and the state: this in itself does not necessarily imply that one of those was a ‘religious’ form and the other ‘political’.

The question, therefore, is how we are to understand the workings of medieval Japanese society, a society in which the presence of the gods and Buddhas and the correct performance of ritual were fundamental to the functioning of the state and society, without becoming enmeshed in the linguistic and conceptual problems of ‘religion’ and its corresponding ‘secular’ analogues. One possible approach is to consider the intersections of practice, institution, performance, and authority, as they interacted at specific places, and so to investigate the overlapping or un-differentiatable roles of the actors who engaged, interacted, or competed with each other in those spaces. By avoiding labelling individuals or actors solely as ‘warriors’ (implicitly ‘political’ or ‘military’) or ‘monks’ or ‘temples’ (implicitly religious), we are then able to consider medieval interactions holistically, which is essential in the understanding of the complex exchanges, tensions, and disputes between individuals and institutions in medieval Japan. Considering medieval interactions through the lens of place and authority provides a framework which supports the investigation of actions and intentions without having to ascribe ‘political’ or ‘religious’ motivations.

Points of tension and disputes are a vital window into the workings of medieval society, and the friction in the interactions at the heart of these disputes, and the means of their resolution, inform our understanding of how medieval Japan ‘worked’ and of the vitally non-differentiable nature of what we might otherwise have thought of as ‘religion’ throughout society.

As a demonstration of this, I present a single incident which occurred up a mountain in central Japan in 1283. This incident was minor and, as far as is known, no-one died. It concerned procedural matters of jurisdiction and authority, and the causes of the dispute and the means of its resolution are representative of the society and justice systems of the Kamakura period. At the same time, this rural incident was directly connected to what in modern terms we would call the ‘religious’ and ‘political’ defence of Japan against the Mongol Invasions, the final institutional and doctrinal split between the orthodox and reformist branches of Shingon esotericism, the existence of any kind of organisational relationship between different Shingon temple complexes, the relationship between public government and private resolution, and the functions of the imperial family as exercised through governmental agencies and shrines and temples. Therefore, not only was this seemingly minor incident significant as a connecting point for several major interactions and processes, but is a demonstration that, by considering disputes of this kind through a holistic understanding of the levers of influence and the construction of legitimate authority and jurisdiction, we can uncover the sinews of medieval society and gain a better overall understanding of it. The heart of this incident was a question of how different forms of authority came together—in conflict—at a specific place, and how the multiple valences of that place as a national/public ritual site, family shrine, administrative office, pilgrimage point, and network node all influenced the contest for control of the site. This cannot be understood by reducing it to only an economic dispute, or only a religious dispute, or even a simple contest for prestige between local rivals: only by considering how all these aspects were interlinked can we understand what happened and why.

2. Rights to Rites: Finding/Getting Away from ‘Religion’ with the Holy Horses of Amano

A characteristic example of the non-differentiable nature of authority in Japan’s medieval society is provided by a dispute and its resolution, which occurred in the late thirteenth century at the Amano Shrine, in what was then Kii Province (modern Wakayama Prefecture).

The dispute concerned who had the right to administer the gifts of horses to the deities of the shrine, and, by extension, who held authority over the performance of the ceremony and administration of the shrine area, and with it, local territorial influence and income. It would be easy to cast this dispute as a struggle between “secular” warrior power and “religious” monastic control, but to do so would be to miss the true nature of the dispute and its participants: if the participants had truly been representing entirely different structures and systems of authority, then neither the dispute nor its resolution would have unfolded as they did. The “warrior” on one side of this dispute was no more or less a secular figure than were the monks on the other, not just because the modern religious–secular dichotomy is meaningless in a medieval context, but because both sides were vying for control over the shrine on the basis of claims which were fundamentally not dissimilar from each other, and which were firmly enmeshed in a shared legal and political-religious system.

Following several incidents between the rival claimants to these ritual rights, a resolution was brokered by the abbots of Ninnaji, itself a representative of both public and private authority, esoteric lineage, and imperial legitimacy. The dispute itself can therefore only be properly understood through a holistic understanding of authority and legitimacy in medieval Japan and, as this dispute will show, this cannot be achieved when considering aspects of it as belonging to the “political”, “legal”, or “religious” history of Japan, rather than a connected whole.

The Amano horse dispute flared up at several points during the mid- to late-thirteenth century, but in order to understand how it came about, it is first necessary to elaborate on the history of the Amano Shrine and its relationship with the great temple complex, Kōyasan. Both sites had been important loci of pilgrimage, land administration, worship, and interaction, with governmental authority for several hundred years, and their relationship with each other had been entwined since the founding of Kōyasan by the monk Kūkai in the ninth century.

2.1. Background

Kūkai received imperial permission to found a temple in the mountains as a place of practice in 816, and his ‘Temple of the Adamantine Peak’, Kongōbuji, was consecrated in 819. The legends surrounding the choice of the mountain plateau Kōyasan for this monastic site are informative. Kūkai conceived of the site as a sacred space on two levels. He envisaged the mountains themselves as a projection of the combined esoteric mandalas, which he had studied in China and brought back with him to Japan in 806. The name he gave to the central temple site, Kongōbuji, identifies it with the kongōkai (vajradhātu) or the Adamantine/Diamond Realm, associated with the esoteric Kongōchōkyō (Vajraśekhara Sutra), while the surrounding mountain peaks were conceived of as the eightfold petals of the lotus assembly at the centre of the taizōkai (gharbadhātu, Womb Realm) mandala, which represents the Dainichikyō (Mahavairocana Sutra).1 The physical mountain landscape itself was therefore mandalised, a projection of and realisation of the nonduality of the two realms of esoteric scripture (Hakeda 1972, p. 50). This provided a ‘universal’ claim to the sanctity of the space, but Kūkai also recognised the role of the existing local deities.

The legend, as retold in later chronicles such as Kōya shunjū hennen shūroku, is that, as Kūkai departed from China on his return journey to Japan in 806, he had hurled a vajra, a three-pronged ritual implement, across the sea. A decade later, and searching for the ideal location to found his ‘temple of the adamantine peak’, Kūkai explored the mountains of the Kii peninsula while looking for the point at which the vajra had landed. As he searched, he encountered a hunter and his two dogs. The hunter was the local deity Kariba (‘hunter’), and he and his dogs guided Kūkai to the vajra. When Kūkai reached the site and found his vajra, a god, the mountain-king (sannō), appeared and conveyed a vast swathe of mountain lands to Kūkai. This is identified in Kōya shunjū hennen shūroku as a tale passed down by the Niu priestly family of the Amano Shrine.2 This is important, as, along with the goddess Niutsuhime, Kariba (also called Kōya) was one of the two primary deities enshrined at the foremost local ritual site, Amano. Kūkai subsequently enshrined the pair of deities at the heart of his planned temple site, with the miyashiro of Niutsuhime and Kariba/Kōya forming the western side of the new temple’s central area. Kūkai was thus welcomed to the area by a pre-existing focus of worship or, to put it another way, he chose a pre-existing sacred space in which to found his temple, and then cemented the link between the Amano deities and the new temple by bringing them into the centre of the ritual space of Kōyasan, itself the centre of the projection of the two esoteric mandalas onto the landscape.

The temple site at Kōyasan did not initially prosper after Kūkai’s time: conflagrations and the harsh winter climate reduced the site to near-ruin by the mid-Heian period. However, the site’s fortunes began to improve from the late tenth century to the early eleventh century onwards, thanks to the development of a third layer of sacred influence. The mausoleum of Kūkai gradually became a focus for elite pilgrimage, encouraged by the developing narrative that Kūkai had not died but had instead entered near-eternal meditation at the eastern edge of the Kōyasan valleys, the point deepest into the surrounding ring of mountains. The growth of pilgrimage to Kōyasan was significant in transforming the status of the site, its economic foundations, and its relationship with Amano, which would have had a direct bearing on the medieval shrine dispute.

During the Heian period (794–1185), there was a gradual transformation of the economy from a centralised bureaucratic model in which government ministries, temples, and the aristocracy were supported by payments of tax revenue to an increasingly decentralised structure in which this income was gradually replaced by privately-administered taxation. The ability and willingness of the government to directly fund the maintenance and repair of temples declined over time, compelling temple administrators to develop private income, primarily through the acquisition of rights within corporately-held private estates (shōen). The early development and nature of shōen has been described more ably and in greater detail elsewhere, but it is worth noting again that we are not looking at neatly-demarcated, mutually-exclusive spheres of operation: temples, their administrators, and their patrons were all shareholders within an emerging system of the privatised administration and taxation of land. A temple did not ‘own’ land outright, nor was it ‘religious’ or ‘glebe’ land, any more than a court noble directly ‘owned’ the land as secular or demesne land. The estate was a corporate entity in which numerous individuals or institutions could hold shares with associated rights and obligations, the product of long processes of legal exemptions and geographical demarcations, and the development and articulation of personal and institutional relationships. The legal bases for these entities made use of the legal provisions made to support the operation of temples and the governmental and administrative role of the aristocratic elite, and investment in shōen rights and donations and the transfers of rights to temples were a key means for noble families to build and retain wealth and influence. The nexus between cultivators, administrators, temples, and nobles in the estate system was thus central to the economic structure which developed in the Heian period, and cannot be reduced to ‘religious’, ‘secular’, ‘economic’, or ‘political’.

After experiencing near-ruin in the generations after Kūkai, Kōyasan’s viability was gradually restored by the growth of interest in the site, accompanied by the transformation of public land into private estates and the donation of estate rights to the temples and halls which made up the temple complex. The first major donation was a swathe of former public land called the Six Villages, or six estates (Rokka-gō or Rokka-no-shō), centred on Amano. Control over this parcel of land was donated in 994 by Higashi Sanjō’in (Fujiwara no Senshi). Higashi Sanjō’in (961–1002) was one of the most significant figures in Court, was the sister of the great politician Fujiwara no Michinaga (966–1028), and was the mother of the reigning emperor, Ichijō (980–1011, r. 986–1011).3 Her donation was made in order to finance the reconstruction of Kōyasan following its destruction by fire and the construction of a new ‘Buddhist’ structure at Amano, the Sannōin. The monastic community of Kōyasan was obliged to remain in residence at Amano for some two decades while the temple site was being rebuilt. Control of the estate was given to the Kōyasan monk Gashin, and the proprietorship of the estate was officially held by the Amano Shrine itself.4 This cemented the connection between the two ritual sites: Amano, at the centre of a string of villages in enclosed valleys half way up the Kii mountain range, and Kōyasan, in its valley network at the top of the mountains. Following the donation of this core area, Kōyasan began to acquire a portfolio of estates across the sanctified mountains, and the ancestral domain conveyed on Kūkai by the Amano gods began to overlap with the income production which supported the temple. One way to express this would be to say that the sacred and economic geographies of the northern Kii province grew toward each other as the land under the administrative control of Kōyasan’s temples expanded in tandem with the development of the mythology of the Amano gods and of Kūkai in the mountain space. At the time that Kōyasan was being reconstructed through the income received from the Amano estates, Kūkai’s connection to this mountain domain was becoming articulated as one of continuing presence: the development of veneration of Kūkai as remaining on the mountain in permanent meditation further imbued the space with immanence and connected the land to the temple site, while at the same time encouraging elite pilgrimage and the donation of estates within that same space. This transformation of Kūkai into a continuing presence was part of the transformation in the interaction of laypeople with mountain sites (Satō 2009, p. 96), but the ‘religious’ activity of the courtly elite in making pilgrimage to Kōyasan and donating rights to local estates was also part of the process of the development of the estate system itself. The growth of the Kōyasan domain was therefore a process in which the sacred and economic functions of the landscape were fundamentally connected.

For this reason, control over the Amano Shrine became a vital political lever for influence within the monastic community at Kōyasan and for control of the land and people around it. From the eleventh century onwards, the monastic community at Kōyasan began to grow beyond the original cluster of buildings which made up Kūkai’s planned Kongōbuji, leading to the consecrated area at the centre of the mandalised mountain landscape becoming a highly contested space, doctrinally, economically, and politically.

By the thirteenth century, there were several major organisational groups at Kōyasan vying for administrative and ritual rights with each other, and with more distant linked temples such as Tōji and Ninnaji. While these institutions contested the control over the physical and ritual space of Kōyasan, ceremonial status, and doctrinal differences, they were in turn also vying with the people who actually lived on the surrounding land: the warrior families who served as estate administrators, and the yeomen and peasantry who resisted being administrated. This was complicated by the multi-layered public and private legal systems on which landholding depended. Just as the ‘religious’ and ‘economic’ functions of this wider landscape were inseparable from each other, the society and organisation of the wider rural community was closely entwined with the organisation and politics of the temple complex. The monastic community was not a separate overlord to the local population and did not form a class ‘above’ the cultivators and local administrators: by the early medieval period, the monastic population of Kōyasan was largely drawn from that same local population, and so tensions and alliances within that population were as much reflected within the Kōyasan community as were any conflicts or cooperation between the mountain and the villages (Wada 1987, pp. 79–90; Yamakage 1988, pp. 323–25).5

In the early medieval period, the Kōyasan area was home to diverse groups of practitioners with a considerable variety of doctrine and organisational structures. The interactions between these groups and provincial society were multilayered, as were the understandings of the physical and sacred space they projected onto the mountain countryside. Three main institutions dominated the central space of Kōyasan: the spatial and ritual centre of Kōyasan was the danjō garan, the space which housed the primary temple buildings, the lecture hall, the central and western pagodas, the image hall, the shrine to the Amano gods Niutsuhime and Kariba, and the central gate. The cloisters which grew up around this area collectively formed the orthodox grouping known as Kongōbuji after Kūkai’s original foundation. The second group was the institutional grouping of the Daidenbōin and Mitsugon’in, cloisters founded in the 12th century by the monk Kakuban. These were associated with Kakuban’s reformist esoteric practice, and they were also well-funded but protocol interlopers in the eyes of the older cloisters. The third major temple was Kongōsanmai’in, in Odawara Valley on the southeastern side of Kōyasan. Kongōsanmai’in grew to importance in the early thirteenth century through its connections to Hōjō no Masako, widow of the first shogun and effective ruler of the Kamakura government. These three main power blocs thus represented three different eras in the establishment of monastic institutions on Kōyasan.

Kongōbuji was the oldest, most complex, and most fragmented of the three blocs in the thirteenth century, but it would eventually emerge as the preeminent organisation on the mountain and the dominant landholder in the region. In the twelfth century, Kongōbuji’s constituent cloisters were dependent on the income from a number of estates donated by members of the aristocracy, before making significant additions in the thirteenth century. In particular, a small group of subtemples which were heavily involved in land administration came to be increasingly central to the overall temple’s decision-making process, and, from the thirteenth century to the sixteenth century, near-monopolised the abbotship of Kōyasan. The internal story of Kōyasan in the twelfth and thirteenth centuries is primarily one of rivalry and conflict between the Kongōbuji cloisters and Kakuban’s Denbōin, later Daidenbōin, over ceremonial rights and precedence, control of the abbotship, and control of land.

The tension between the Daidenbōin group and the Kongōbuji faction rested on doctrinal disagreement and an argument over precedence within the Kōyasan community, but it was also fuelled by income and administrative rights. Daidenbōin and Mitsugon’in were funded by the donation of a parcel of valuable estates made in 1132 by Kakuban’s patron, the retired emperor, Toba. As Mikael Adolphson has pointed out, funding new cloisters within established religious centres was a common tool for political leaders hoping to gain a foothold of influence at these sites, as well as any personal spiritual motivation held by the original (Adolphson 2000, p. 83). Toba’s donation gave Daidenbōin income from Iwade, Hiroda, Yamazaki, Okada, and Sandō, downriver in western Kii Province, and from Ōga and Shibuta estates on the upper reaches of the Ki River nearer Kōyasan (Tanabe 1998, p. 47). These estates were richer and more numerous than Kongōbuji’s limited twelfth-century possessions, rapidly changing the wealth dynamic on the mountain.

In contrast to the clusters of local estates held by Kongōbuji and Daidenbōin, the third major institution, Kongōsanmai’in, was primarily supported by estate holdings in neighbouring provinces, primarily in the nearby Kawachi Province, but also further afield in Settsu and Harima. In contrast to the core of Kongōbuji’s estates, which came from noble donations, Daidenbōin’s, which came from imperial support, Kongōsanmai’in received its estate holdings courtesy of its links to the military administration at Kamakura.

2.2. The Origins of the Dispute

The institutional and doctrinal tensions within the Kōyasan community and the expansion of private land rights in the Kii province were intertwined, drawing questions of local standing and family rights into the orbit of the institutional strife between Kongōbuji and Daidenbōin in the one hundred and fifty years between the establishment of the Daidenbōin cluster and its eventual departure from Kōyasan.

Key to this was the class of local administrators and enforcers, which constituted the leading families of rural society. This class could be described as provincial warriors, but to do so is to define them in terms of military function and therefore only describe one facet of their power base. The families which dominated society at the estate level did indeed have access to military organisation and equipment, but their position also depended on the holding of official or private administrative roles and the control of ritual rites. As estate officials, such individuals were the administrative link between monastic proprietors and the revenue-generating cultivators, but they were also rivals for control of the mechanisms of control and administration at the estate level.

In the Amano area, the preeminent family was the Sakanoue family. The Sakanoue family had been resident in the Kinokawa valley since at least the mid-Heian period, predating the enclosure of most of this land into private estates, and had from time to time served as minor government officials with jurisdiction over Ito District, in which most of the villages of the Rokka estate were located (Wada 1965, p. 15). The family and their followers had opened and developed land between Amano and the Kinokawa (River Ki), across a swathe of land connected to Kōyasan and to the Amano Shrine. The relationship between the monastic community and the Sakanoue family was typically ambivalent: the family were banished from Kongōbuji’s primary estate, Kanshōfu, in 1089 or 1090, after the head of the family, Sakanoue no Tsunezumi, murdered a Kongōbuji administrator.6 The main line of the Sakanoue family resettled nearby, on land which would go on to become Ōga estate.

In the twelfth century, Ōga became a personal holding of the Shingon reformer Kakuban, founder of Daidenbōin, before being confirmed as a holding of Kakuban’s Mitsugon’in by the retired emperor, Toba. When the estate was transferred to Mitsugon’in, the head of the Sakanoue family, Toyozumi, wrote to Kakuban and provided documents and a genealogy which showed that his ancestors were the original administrators of the land and that Toyozumi’s dependents had developed the land extensively (Noritake 2002, p. 7). These documents established that the Sakanoue line was the preeminent family in the estate, and that they held pre-existing organisational responsibilities within it. Kakuban therefore appointed Toyozumi as geshi, a senior in-estate administrative role. The Sakanoue family therefore became an essential link in the economic support for Daidenbōin’s Mitsugon’in as estate administrators, while the family’s own position was strengthened by this direct connection to an imperially-sponsored organisation.

By the Kamakura period (1185–1333), the Sakanoue family had also grounded their local influence through the administration of ritual at the Amano Shrine. The head of the Sakanoue family also claimed the title of Amano uji-no-chōja, a role which expressed the authority of the head of the family in terms of ritual responsibility. The uji-no-chōja, the ‘clan elder’ or head of family, was the most senior position within extended family structures in the Heian and Kamakura period, and derived its prestige from its responsibility for the administration of rites to the family’s tutelary deity. The origin of the Amano uji-no-chōja role is unknown; however, from mentions of it in thirteenth-century documents, it can be surmised that one of the Amano gods was considered to be the Sakanoue family’s tutelary deity. Wada Shūjō has posited the interesting suggestion that the origin of the Sakanoue link to Amano goes right back to the original presence of the Sakanoue in Kanshōfu estate and the construction of the Sannōin at Amano by the Abbot Shinga: his hypothesis is that Shinga contracted leading local families in Kongōbuji’s primary estate to accomplish the construction at Amano and Kōyasan, appointing the Sakanoue family as uji-no-chōja to gain their assistance and tie the shrine closer to Kōyasan (Wada 1965, p. 16). Given the two-hundred-year span of time between Abbot Shinga and the first surviving records of the Amano uji-no-chōja role, this remains an intriguing but speculative theory. Possession of the uji-no-chōja title directly linked the Sakanoue family, as local landholders, administrators, and warriors, to the Amano Shrine as a sacred and administrative space (Niutsuhime Jinja-shi 2009, p. 86). The assertion of control over the management of ceremonies at the shrine was a matter of weighty significance, which cuts across the modern conceptions of ‘political’, ‘economic’, and ‘religious’ authority. The local position of the Sakanoue family was dependent on the assertion of organisational power and social status over the cultivators and lesser landholders, and this status and capacity to organise was directly bolstered by the possession of the right to organise and administer ceremonial activity at the foremost shrine in the region. The hierarchical structure of this ‘warrior’ family and its followers and dependents likewise depended on the head’s position as the primary worshipper of the family’s tutelary deity: control over Amano was thus necessary for the maintenance of Sakanoue status and local control. In addition to this, control over the administration of shrine activity was a vital lever for the administration of the estate land surrounding it, to which the family held extensive rights. The development of the association between this local family of warrior administrators and the shrine to the local gods is therefore impossible to separate into ‘economic’, ‘political’, ‘familial’, or ‘religious’ aspects: the nature of the relationship can be approached through each of these facets, but only fully comprehended by understanding their indivisibility in the Kamakura period.

The geographic, administrative, and ritual connection between the Sakanoue family and the Amano Shrine drew the family into conflict with the Kongōbuji group at Kōyasan in the thirteenth century, and, by its nature and the nature of the multiple connections between land, sacred space, administration, and patronage, this dispute became part of a much larger conflict between Kongōbuji and Daidenbōin, with lasting effects on not just the economic control of the Amano Shrine, but also the schism in the esoteric doctrine, which bifurcated medieval Shingon Buddhism.

2.3. Amano in the Thirteenth Century

In the thirteenth century, Amano consisted of several clusters of buildings and areas which were the representatives of different interested parties in this sacred space. There were four primary shrines: those of the original local deities Niutsuhime and Kariba/Kōya, plus two more gods, Kibi and Itsukushima-myōjin, whose shrines were added in about the year 1200 (Niutsuhime Jinja-shi 2009, p. 66). Each of these shrines was administered by a priestly family, headed by the Niu family of the Niutsuhime shrine, who also held the post of sō-kannushi, head of the shrine priests. Outside the shrine area stood the Sannō’in, the temple established at the end of the tenth century and affiliated with Kongōbuji. The four Amano deities were represented in combinatory form within the Sannō’in, with a recitation of the Lotus Sutra performed for their benefit by the resident monks in the fourth and fifth months each year.7 Sannōin was established as part of the reconstruction of the primary Kōyasan complex through Higashi Sanjō’in’s donation, and thus linked the administration of the land (Rokka Estate) and its use in supporting the upkeep of Kōyasan’s buildings with the role of the Amano gods as the original possessors of a sacred space which encompassed that land and the protectors of the temple complex. Rather than understanding this as the overlap of two different concepts, the economic function of the land and administrative function of Sannōin on one level and the protective role of the deities over Kōyasan and the role of the Sannōin monks in ministering to them on another, this connection between Amano and Kōyasan must be considered as a connected whole.

Adjacent to the Sannō’in was a longhouse (nagatoko) used by a group of practitioners of mountain asceticism, the Amano Nagatokoshū. The Nagatokoshū were an example of the liminal groups who held nominally low status and subordinate connections to established institutional centres, but which were significant to the Kamakura-period economy and the operation and management of the networks upon which the major temple and shrine sites depended. The significance of such groups, and their role as networks, has been discussed by Yoshihiko Amino (2007, pp. 209–10; 2012, pp. 76, 158–60), Kazuo Yamakage (2002, pp. 185–91), and Yoshitaka Yamamoto (2010). As mountain practitioners, the Nagatokoshū moved across the landscape of the Amano and Kōyasan region, their paths extensively crisscrossing the area (Wakayama-ken and iinkai 2005, p. 77). This made the group an important network of communication and organisation, one which connected the sacred sites to their surrounding landscape and the scattered communities within the estates to these centres. The Nagatokoshū also held administrative rights in some of the land, across which their peregrinations took them: they held the proprietorship of Iwabashiri Village in Kongōbuji’s Makuni Estate, to the southwest of Amano, and were extensively involved in disputes against local warrior-managers and the networks of rival shrines for control of the land.8 The Nagatokoshū were connected to Amano by location and practice, to Kōyasan through Amano’s connections to Kongōbuji and role within Kongōbuji’s land administration, and were considered part of the broad category of the lowest-status members of the overall monastic community (the Rokubanshū, the sixth group), and they were also connected to the temple Ninnaji in Kyoto through Ninnaji’s own interests in land administration in Kii and to the control of the Amano Shrine.

There were also two managerial titles connected to the administration of the shrine area: that of Amano uji-no-chōja, the title claimed by the head of the Sakanoue family as described above, and Amano inju, supervisory head priest. The authority to appoint both of these managerial roles changed hands over the span of the eleventh century to thirteenth century, held sometimes by individuals and at others by institutions, primarily Kongōbuji at Kōyasan and Ninnaji in Kyōto. Ninnaji exercised extensive influence over the administration of several Kongōbuji estates in the thirteenth century, and was the original patron institution to the Amano Nagatokoshū through its strong connection to the Amano Shrine. The influence of this geographically-distant temple, therefore, ran through the causes and resolutions of the periodic disputes.

Ninnaji’s original connection to Amano was as a result of a chain of transfers of land rights in the late tenth to eleventh century. Higashi Sanjō’in had donated the land around Amano to finance the reconstruction of Kōyasan, with the patron right (honke-shiki) of the surrounding Rokka Estate held by the Kōyasan kengyō, the de facto abbot of the community. The abbot of the time, Shinga, had then passed on the title to the land to his disciple, who had then attached it to a temple he founded near the capital, and this temple and the land title were then commended to Ninnaji. The abbot of Ninnaji at the time was the imperial prince Shōshin, who had himself received an esoteric initiation from Shinga’s disciple, thus linking land rights with ritual legitimacy (Niutsuhime Jinja-shi 2009, pp. 67–68). This meant that Ninnaji held the highest title in the hierarchical chain of shares and rights to the land around Amano, though Kōyasan held the administrative rights.

The importance of Ninnaji to the local dispute in Amano can only be understood if we consider the full significance of the temple as an institution in the medieval period. It is also an excellent example of the unprofitability of attempting to separate medieval institutions into “religious”, “political”, or “economic” entities. Ninnaji was, like all major temples, an institution with responsibilities which straddled all three of these modern categories, but most significant was its role as an institution that was both monastic and imperial: an organ within a diffuse framework of government in which sovereign authority was similarly diffuse. The temple was founded directly by Emperor Uda (867–931) in the year Ninna 4 (888) as an ‘imperial vow temple’ (chokuganji), and was central to the development of a nexus between monastic and imperial lineage in the following centuries (Bauer 2012, p. 241). Mikael Bauer has identified the founding of Ninnaji as central to this expression of ‘shared sovereignty’, the site of Uda’s residence and place of his death, a temple led by monks who were both his bodily descendants and initiates into the esoteric-exoteric Hossō-Shingon lineage he established (Bauer 2012, p. 244). Ninnaji was also linked to Kōyasan’s Kongōbuji and Daidenbōin as a Shingon temple. Whilst it is an anachronism to talk of a single organised or hierarchical Shingon school in the period, Ninnaji was an influential node within the network of Shingon centres, with considerable influence extending to other major nodes. Ninnaji was therefore a monastic and a governmental institution, in which the public authority of the imperial line gave legitimacy to the private adjudication of its prince-abbots. This means that, while attempting to avoid the snares of modern categorisations, it must also be considered as a legal institution and dispenser of justice, and not directly a part of the public offices of the imperial court, or of the law courts of the Kamakura government, but a ‘private’ adjudicator with the mantle of ‘public’ imperial authority.

These buildings and groups of practitioners reflected the fluid balance of power at Amano. The families of shrine priests were connected to the Amano’s early history, and were hereditary custodians of the shrines whose status was closely connected to the gods of the shrine itself (Niutsuhime Jinja-shi 2009, p. 66). The Sakanoue uji-no-chōja post linked the organisation of the shrine with the organisation of people and land, and were a key link between institutional proprietors and local administration. The Amano Nagatokoshū and the Amano inju post were also links between the local and the national. In the context of the doctrinal, ceremonial, and jurisdictional dispute between Daidenbōin and Kongōbuji at Kōyasan, it also meant that both factional groupings had affiliates at Amano (see Figure 1). The intensification of hostilities between the two groups at Kōyasan coincided with a rise in Amano’s national prestige and a multi-generational feud between heads of the Sakanoue family and Kongōbuji, setting the scene for a significant confrontation in 1283.

The increased prominence of Amano and the strengthened hand of Kongōbuji in its affairs were the result of the defence of the nation against the second Mongol invasion of Japan in 1281. Shortly before the invasion, the Amano Shrine increased its national profile and income when it produced an oracular statement from the deity of the third shrine, Kibi, announcing that the Mongols were coming and that the gods of Japan would lead the defence (Yamakage 2006, p. 111). The military government at Kamakura accordingly sent tribute to Amano in the form of a bow, sword, and cloth, and would go on to reward the shrine further by donating an estate in the neighbouring province. Amano was also elevated to the status of the ‘first shrine’ (ichinomiya) of Kii Province (Miyake 2000, pp. 304–5). This gift was a vital part of the Kamakura shogunate’s preparation for the war, as central to preparations to repel the Mongols as was the mustering of its human troops: both the Amano oracle and the promise of godly war in parallel with human efforts were accepted as truth by the Kamakura government (Yamakage 2006, p. 112).

Kongōbuji also contributed to the mobilisation against the invasion, with its Nan’in cloister sending its primary object of worship (honzon), ‘Wavecutter-Fudō’, to the front line in northern Kyūshū, led by the monk Kenryū, where, as promised by the Amano oracle, the fire of the esoteric deity Fudō myōō was seen to fall on the Mongol ships. The Mongol invasion fleets were, of course, destroyed by violent weather as well as human defences. Before the dreadful and awesome power of wind and waves, it is easy to see how great the power of the earth must have seemed in comparison to the capacity of human civilisation, and it is therefore no wonder that the defence of the nation through ritual/spiritual power made celebrities of those involved. The Kōyasan kengyōchō, the register and chronicle of the temple’s medieval abbots, expressed this as “talk of Kenryū’s dharma power rang out, and Abbot Seiben resigned in favour of Kenryū”.9 This means that, in the 1280s, Amano and Kongōbuji were both in favourable national positions, and the abbotship of Kongōbuji was held by a vigorous rising star, Kenryū.

2.4. The Holy Horses Incident

On the morning of the 25th day of the third month of 1283, the priests of the four primary Amano shrines were due to hold the shiki onmatsuri spring ceremony, at which horses for the gods would be presented to the shrine. The presentation of shinme, horses for the gods’ use, had occurred several times previously in the thirteenth century, with gifts to the shrine being made by the princely monk Dōjin of Ninnaji and by the retired emperor Kameyama in 1237 and 1280–1281. Each of these significant offerings had led to outbreaks of violence over the rights to organise each ceremony and the question of precisely who held the rights to conduct the organisation of the event and to present the horses to the shrine. When Dōjin of Ninnaji made his pilgrimage to Kōyasan and presented horses to Amano, the Sakanoue uji-no-chōja Tomozumi was murdered by the Kongōbuji faction in the ensuing fracas and the Kōyasan abbot Shōshin was compelled to resign. The next year, the murdered Tomozumi’s son, Morizumi, was confirmed in post as the next Amano uji-no-chōja by Ninnaji and further sought the support of the Amano priestly families in 1266.10 When Kameyama attempted to offer horses as part of the preparation for defence against a second Mongol invasion, the Kōyasan abbot again attempted to assert control by sequestering the horses, but backed down when Morizumi travelled to Kyoto to seek a favourable ruling from the court (Yamakage 2006, p. 116). Each of these confrontations led to shifts in the possession of rights, prestige, legitimacy, and income for each of the parties involved, and the incident in 1283 appears to have marked a decisive resolution.

The heads of the Sakanoue family maintained that the administration of the ceremony was their right as uji-no-chōja, while the leader of Kongōbuji (kengyō—the de facto abbot) claimed the title of Amano inju. This placed Kongōbuji in direct opposition to its rival Daidenbōin’s proxy, but it also drew in Ninnaji due to its connections to the Rokka Estate and Amano area and control over the confirmation of the uji-no-chōja and its sometime control over the appointment of the Amano inju.

In 1283, it appears that rumour spread that uji-no-chōja Sakanoue Morizumi’s men and Kongōbuji abbot Kenryū’s troops were about to fight each other at the shrine over the matter of the administration of the ceremony. Exactly what happened is unclear, but, in effect, both sides claimed that they had heard that the other side were going armed, and so brought a military force of their own to defend themselves, with the obvious result that everyone showed up armed to the teeth. According to the subsequent lawsuit brought against the uji-no-chōja by Kongōbuji’s monks, Sakanoue Morizumi “mustered his men, donned helmet and armour, and intended to assault the shrine ceremony”.11 However, according to Morizumi’s defence statement, it was as much the fault of the abbot Kenryū, and he asserted that the Kongōbuji abbot “feared that he [Morizumi] would seize the horses for himself, made preparations for a fight, and awaited Morizumi’s arrival”.12

What happened when—or if—armed men showed up at the ceremony is unknown, but, according to Kongōbuji’s statement, this putative confrontation then triggered an escalation of tensions within the Kōyasan community. The lawsuit alleges that Sakanoue Morizumi took advantage of this dispute to plot together with Daidenbōin to attack and burn Kongōbuji with a combined force of Daidenbōin monks, Morizumi’s followers and the residents of Ōga estate, and ‘bandits’ from neighbouring provinces (Noritake 2002, p. 10). This alleged plan was widely known in Kōyasan, therefore Kongōbuji summoned its own estate officials and set them to guard the buildings under Kongōbuji’s control. The third major institution on Kōyasan, Kongōsanmai’in, was also drawn into the stand-off, as the rumours also claimed that Morizumi’s ‘bandits’ intended to break into Kongōsanmai’in’s warehouse, causing this temple to also summon its own estate officials (Noritake 2002, p. 10). This situation persisted throughout the fourth and fifth months of 1283, until the situation became untenable. According to the Kongōbuji statement, the concentration of the temple’s estate managers cum military forces on the mountain meant that these same officials were therefore not able to conduct the management of their estates, and so no work was being done.13

Kongōbuji’s complaint that the tension had not only “consumed the whole mountain with worry”, but that it affected the communities below the mountain to the extent that “old and young left their occupations and the peasants forgot their agricultural labour”, gives a sense of the temple complex as the centrepoint in a broader system (see note 13). Tensions between Kongōbuji and Daidenbōin were rising ever higher in the 1280s, and this dispute was, at its heart, a matter of doctrine and ceremonial order. At the same time, the two groups were institutional landholders, and their control over local land was exercised through resident managers. Those managers were men such as Sakanoue Morizumi, the leaders of prominent local families, whose social positions depended on their connections to the landlord temple, to local society, and to the land, as understood through patronage and control over shrine rights. Thus, the arguments over philosophy and praxis which divided the Kōyasan community were at the same time territorial disputes backed by warrior families. The estate officials—managers, warriors, local worthies—were summoned to defend their patrons, but their participation in the Daidenbōin–Kongōbuji confrontation was not merely a matter of quasi-feudal service, but an active pursuit of their own local interests. For Morizumi, collusion with Daidenbōin served to secure his position in estate society through an alliance with Daidenbōin’s own manpower, and also through its access to legal mechanisms and connections at the national level. The men summoned to defend Kongōbuji and Kongōsanmai’in were of the same background as Morizumi, each reinforcing their own local jurisdiction and influence by drawing in under the legal umbrella and military resources of the proprietary temple.

To return to the central point, this confrontation is a good demonstration of the difficulty in finding or separating ‘religion’ from medieval society, but it also exposes the impossibility of answering the traditional essay question of to what extent ‘Buddhism’ was significant to the general population in the early medieval period. If the larger conflict between Kongōbuji and Daidenbōin is deemed to be ‘Buddhist’ or ‘religious’, then does that mean that the estate managers were participating in religion when they joined the military build-up at Kōyasan in 1283? surely not. One might say that these were just simple warriors serving as guards, an interpretation which seems reasonable, even if it is implicitly buying into the notion that ‘religion’ belongs to the domain of thought and doctrine, while the mechanics of administering the places of religion are essentially ‘secular’. However, this assertion is untenable once one remembers that Sakanoue Morizumi’s participation was in defence of his own ritual order. We can go further than this, however: if we call these local men ‘warriors’, or even ‘estate managers’, there is a risk of treating them as ratiocinated ‘secular’ forces, as if they were not as deeply embedded in a ‘religious’ understanding of the land around them and their own lives. As Kuroda argued, shōen society was dominated by agriculture, and agricultural communities were, in his words, “thoroughly entwined with and overshadowed by thaumaturgy and polytheism”, with warriors and local lords as connected to land, ritual, and belief as the cultivators (Kuroda 1996, pp. 288, 298).

More importantly, the quasi-feudal obligation which Kongōbuji put on its estate managers to render this military service was itself expressed in language drawn from esoteric thought and practice. As Kongōbuji’s control over its proprietary estates strengthened in the late Kamakura period, it compelled many of its resident managers to sign written oaths (kishōmon), many of which included a vow requiring that their immediate attendance was required when summoned by the temple, and that, if the signatory himself were too ill to travel, then he must send his son in his stead.14 The powers invoked by the signatories are indicative. These estate managers swore by “Bonten, Taishaku[ten] and the four Great Heavenly Kings, the great and small shrines of the country of Japan, the protective gods of the four sacred sites of Amano, [Kōbō] Daishi (Kūkai), and the various Buddhas of the [esoteric] Diamond Realm and Womb Realm”.15 This list includes Indian deities incorporated into esoteric Buddhism, the myriad Buddhas and Bodhisattvas of the Mahavairocana Sutra and Vajrasekhara Sutra and the figure of Kūkai himself, as well as the deities enshrined at Amano. This was therefore a statement which unequivocally linked the ‘warrior’ managers to their patron temple through the deployment of the terrifying power of the esoteric pantheon and the local gods who had become protectors of the temple site. It should also be remembered that the land that these local ‘warriors’ sought to maintain influence over was the original domain of the Amano gods, as entrusted to Kūkai, and was land whose peaks Kūkai had identified with the projections of the mandalas of the Diamond and Womb Realms. It is clear, therefore, that the interaction between the temple proprietor and local families cannot be understood by an explanation of their ‘administrative’ or ‘economic’ or ‘legal’ relationship without that relationship being understood as saturated by ‘religious’ belief and practice. Equally, any discussion of the religious practices or beliefs of the provincial population cannot take place without an understanding of the administrative relationship and hierarchical rights which defined the estate system.

What was the outcome of the Amano Holy Horses incident in 1283? Despite the allegations made by both Sakanoue Morizumi and Kongōbuji, each accusing the other of bringing armed men to the ceremony in the third month, and despite the months of tension between opposing armed camps within Kōyasan, the resolution of the dispute appears to have been achieved without further violence. The reason why we know of the existence of the dispute itself is because both parties set out their grievances in writing. These documents are titled in accordance with the practice of law in the Kamakura period: the Kongōbuji clergy made a statement of grievance (shūjō), and this was met with a defence statement (chinjō) from Morizumi.16 The legal documents were then presented, accompanied by oral statements, also in accordance with the expectations of the judicial system of the period. However, the venue for resolution was not the Imperial Court nor the legal offices of the shogunal government at Kamakura, but Ninnaji. Ninnaji’s authority to attempt to resolve the case was derived from three overlapping sources: its historical connection to the Amano Shrine through possession of the patronage of the Rokka Estate (honke-shiki), which gave it a role within the customary law of the estate system (honjo-hō); its position as a senior Shingon temple, with connections to both Kongōbuji and to Daidenbōin; and its longstanding affiliation with the imperial family and the imperial court at Kyōto.

The adjudicator, Hōjo of Ninnaji, attempted to negotiate a settlement between the complainant and defendant: the details of his proposed resolution are unclear, but his frustration at the difficulty of inducing the rival sides to agree to it is palpable. At the end of the month, he lamented that the positions and claims of Kongōbuji and Morizumi were ‘like fire and water’ in their utter opposition, and that he had not truly been able to get to the bottom of their claims.17 By the end of the sixth month, it was clear that the attempt to settle privately at Ninnaji had failed, and Hōjo recommended that the case be passed from the private hearing at the temple to the court of the Retired Emperor. Retired Emperor Kameyama also seems to have been puzzled by the claims and counter-claims in the lawsuit, but issued an edict ordering that the terms of the mediation at Ninnaji be adhered to.18

This demonstrates the multiple layers of law in the early medieval period. First, the two parties of the dispute presented their argument and defence at Ninnaji—the patron of the shrine at the centre of the dispute, a suitable arbiter on matters pertaining to the issue, and also an institution which conveniently lay between ‘private’ and ‘public’ and between the modern categories of ‘religious institution’ and ‘government’. As discussed earlier, Ninnaji had been founded by Emperor Uda in the ninth century, and was an eminent centre of Shingon ritual. What is significant is that Ninnaji was the vehicle through which Uda continued to exercise sovereign authority after his retirement as emperor, and the temple continued to have members of the imperial family as abbots for much of the Heian period (Adolphson 2007, p. 218; Conlan 2011, pp. 83–84). This created an enduring legacy of involvement in both private and governmental affairs, giving Ninnaji the mantle of governmental legitimacy while retaining the ability to assess legal cases in the private sphere. The monk who attempted to achieve a settlement in the Amano dispute, Kaiden jugō Hōjo, was not an imperial scion, but was the son of Kujō Michiie, the powerful politician, imperial regent, and father to the fourth Kamakura shōgun, and so was directly connected to the highest of political lineages (Adolphson 2000, pp. 189–93). Therefore, if it was a mistake to see the arbitration at Ninnaji as representing a purely religious or internal attempt to resolve the dispute, it would equally be inaccurate to see the Retired Emperor’s judgement as representing a purely civil or public resolution. The institution of retired, or cloistered, emperors had emerged in the late Heian period in close connection to the establishment of temples by retiring emperors, and their frequent assumption of titles of monastic authority, and thus the tradition of the Courts of the Retired Emperor, was one which was freighted with religious symbolism, as much as it made use of the retention of the retiring emperor’s sovereignty.

3. Aftereffects and Conclusions

The outcome of the 1283 dispute over the ceremonies at the Amano Shrine is poorly attested, but appears to have resulted in a decisive shift in favour of Kongōbuji. The Kōyasan kengyōchō laconically says of the Kongōbuji leader Kenryū: “in this abbot’s time the Amano uji-no-chōja post was abolished. This was because of the Holy Horses dispute. This was all due to the abbot’s power”.19 This record gives no further information on whether Kenryū was able to accomplish this elimination for the Sakanoue claim beyond that it was due to his vigour or power, but a few suggestions have been made. Kaji Hiroe has suggested that Kongōbuji accepted Ninnaji’s proposed settlement, and whatever compromise it entailed had the effect of weakening the uji-no-chōja’s claim to the horses, making it possible for Kongōbuji to eliminate the post after this (Yamakage 2006, p. 117). Yamakage Kazuo takes this further and suggests that perhaps Ninnaji and Kongōbuji colluded to the effect that, if Kongōbuji accepted the terms of the settlement, then Ninnaji would abolish the Amano uji-no-chōja role in return. Whatever the diplomacy required to accomplish this, the long-term result was that the Amano Shrine came under the direct control of Kongōbuji, with the Sakanoue, and hence Daidenbōin, interest eliminated and Ninnaji’s own influence terminated. This was a key stepping stone towards Kongōbuji’s domination of Kōyasan and the expansion of its landed domain in the late Kamakura period. In the years immediately following the Amano dispute, Kongōbuji expanded aggressively: In 1285, it annexed parts of three estates which fell within the ancient boundaries of the sacred domain of the Amano gods, invading them with a force of worker monks from Kōyasan and possibly the Amano Nagatokoshū; moreover, in 1288, the Daidenbōin group left Kōyasan entirely, abandoning the mountain for Kakuban’s temple Negoroji (Tashiro 1977, pp. 122–23; Kaizu 2002, p. 122; Garrett 2015, pp. 94–95), resulting in the domination of the mountain and surrounding estates by Kongōbuji. The elimination of Daidenbōin/Negoroji influence within the bounds of the sacred domain of the Amano gods as granted to Kūkai was completed in 1333, with Go-Daigo’s recognition of Kongōbuji as the unitary holder of Shibuta estate and the portion of Ōga estate which fell within the divine boundary (Koyama 1998, pp. 68–71).

The outcomes of the dispute are thus as resistant to categorisation as ‘religious’ or ‘political’ as its causes. The final schism between the ‘New Shingon’ lineage of Kakuban’s Daidenbōin and the orthodox ‘Old Shingon’ which established dominance at Kōyasan had real effects in the development of practice and doctrine at these centres—an animosity which began in the twelfth century, having grown into philosophical disagreement by the time of the Amano dispute in the thirteenth century (Van de Veere 2000, pp. 39–40). However, the development and intensification of that split can only be understood by taking into consideration the multiple layers of the relationships between the Daidenbōin and Kongōbuji factions with the land around them, from the complex connections between ‘warrior’ families and temples, to the cultivation of land from its ‘economic’ organisation to its ritual underpinnings and supernatural guarantors, to the role of the temples and of the Amano Shrine as parts of a national network of defence, itself a combination of military, economic, and ritual factors.

Early studies, such as that of Uwayokote Masataka, considered the Sakanoue family as a ‘warband’ (bushidan), and approached the interaction between it and Kōyasan as an adversarial political/economic relationship, in which the expansion of Kōyasan’s estate control threatened the Sakanoue family’s administrative grip on cultivators and land, pushing them towards becoming rebels or bandits (akutō, as the term was understood in the mid-twentieth century) (Uwayokote 1956; Ōae 1958). While this approach was superseded by the revolution following Kuroda Toshio’s kenmon theory and more recent studies which have contextualised the Sakanoue–Kongōbuji competition over Amano as part of the territorial conflict between Daidenbōin and Kongōbuji (Atsuta 2004, p. 145; Yamakage 2011), these have treated the incident through an economic or military lens: the monastic institutions have been recognised as complex entities with political, economic, and kin-based ties, but the Sakanoue essentially remain uncomplicated warriors. The present re-examination of the dispute indicates that, in order to comprehend the over-layering of contested webs of authority at Amano, we need to see the Sakanoue faction in the same light as the recognised complexity of the temples, moving further away from the monks versus warriors and implicit religious versus secular modes of authority. The position of Sakanoue Morizumi in the Amano dispute cannot be understood by considering local warriors as leaders of armed forces and overseers of cultivators without also understanding them as ritual practitioners and religious figures, whose authority over both their extended family and land depended heavily on ceremony and prestige derived from their connection to temples and shrines. Wada Shūjō came closest to this in the 1960s, when he observed that the sphere of influence of the Amano Shrine in terms of belief considerably overlapped with the region of the Sakanoue family’s territorial influence (Wada 1965, p. 17), though regrettably no evidence for this was put forward. It should, therefore, be no surprise that, after losing Amano, the main line of the Sakanoue family focused their efforts and investment on developing the family’s ritual centre in Daidenbōin’s Ōga estate. In 1294, Sakanoue Morizumi’s sons constructed a five-storey pagoda in his memory at Iōji (Noritake 2002, p. 11). Iōji was the administrative temple (bettōji) for the Ōga Great Shrine, the centre of the estate community, and the construction of this memorial to Morizumi was evidently intended to position it as a bodaiji, a family memorial temple, while also securing the connection to another local land-related shrine.