Withered Wood and Dead Ashes—Making Sense of the Sacred Bodies of Kamatari at Tōnomine

Abstract

The portrait statue of Fujiwara Kamatari (614–669) enshrined at Tōnomine is well known for its agency and mantic powers. Known to crack whenever the stability of the clan was under threat, the icon was carefully observed and cared for. However, not one but two portrait statues of the Fujiwara ancestor existed at Tōnomine in the Heian period, until one was destroyed in the infamous 1208 attack by armed supporters of Kinpusen. This article proposes first to investigate the relationship between these two icons, to show how their dynamic interaction is at the source of the cracking episodes that came to define Kamatari’s cult in later centuries. Then, by looking at the ways in which members of the Fujiwara clan reflected on the nature of the remains of the statue lost in 1208 and on the role of the extant one, it draws attention to how the ritualization of Kamatari’s statue was also couched in Confucian ideas and practices.

1. Introduction

The wooden portrait statue (miei; also, goei 御影) of the first of the Fujiwaras, Nakatomi no Kamatari 中臣鎌足 (614–669), enshrined at Tōnomine 多武峰 was, throughout the medieval period, renowned and venerated for its mantic powers. From the late tenth century, the statue was known to rupture (haretsu 破裂) whenever members of the clan were facing threats that could potentially challenge their political authority, or even simply when Kamatari wanted to voice discontent to his descendants (Grapard 1984, pp. 253–56). The fissures, which altered the visual and material dimension of the statue, were also accompanied by resounding roars and unusual lightnings coming from the mountain, and specifically from the burial mound of Kamatari (DNBZ 118, p. 510a–b), further supporting the idea that any unusual phenomenon occurring on the mountain was an expression of the ancestor’s will. As ominous occurrences, cracks and rumblings were meticulously interrogated; the depth and width of each fissure was measured, reported to the court, interpreted through divination and, once the issue had been addressed or ritually handled, the statue was eventually repaired. Records of ritual petitions and divinations held in these occasions show a concern for both the welfare of individual Fujiwara members and for the prosperity of the clan, which, depending on the circumstances, influenced decision-making on clan-related and official matters, de facto coming to bear on the political life of the time. Thus, aside from materializing the ancestor’s agency, the outbursts of the statue were a tool to elicit or sanction Fujiwara decisions, as members of the lineage skillfully exploited its damages as a way of ensuring the continuity of their socio-political prominence.

Nonetheless, the meaning and function of Kamatari’s figure and icon changed over time. The sources show that cracking occurrences intensified in the Muromachi period (1392–1573), when he was deified as Tōnomine Daimyōjin 多武峯大明神 (Kuroda 2011, pp. 210–14). According to Kuroda Satoshi, the frequency of the bursts between the fourteenth and fifteenth centuries, paired with the emergence of specific rituals involving the portrait, also elicited a rise in the production of his painted examples and their dissemination beyond Fujiwara circles (Kuroda 2007, pp. 192–201). This marked a shift in the perception of Kamatari as patron of aristocratic and imperial power, to protector of society and the polity at large (Ibid., p. 196).

Fascination with Kamatari continued into the Tokugawa period, as suggested by the circulation of his portraits, as well as the proliferation of textual and visual representations of the Taishokkan 大織冠 (The Great Embroidered Cap, the honorific title of Kamatari) narrative, the renowned ballad drama (kōwakamai 幸若舞) focused on his daughter Kōhakunyo 紅白如 (Trede 2004). However, the cracking of the portrait statue came to an end in 1608. This date is somehow significant for yet another change. As emblematic of the statesman turned ancestral figure of his uji and then custodian of the land, elements of Kamatari’s legend seem to have been appropriated to justify the burial practices of both Toyotomi Hideyoshi 豊臣秀吉 (1537–1598) and Tokugawa Ieyasu 徳川家康 (1543–1616) in the early Tokugawa period (Gerhart 1999, fn. 34, p. 171; Bruschke-Johnson 2012, pp. 170–72). In the case of Ieyasu, Kamatari’s apotheosis as Tōnomine Daimyōjin may have also provided the blueprint of the successful deification of an ancestral figure (Boot 2000).1

All this suggests that, while the worship of Kamatari may have changed through the premodern period, the livelihood of his portrait remained the catalyst of different processes of signification. In this article, I would like to focus specifically on the ways in which the material form of the Fujiwara ancestor was construed in the medieval period by his descendants, through a close reading of relevant passages from sources compiled by Tōnomine clerics and diaries of members of the Fujiwara clan, written between the eleventh and thirteenth centuries. These documents are essential to gain a better understanding of the Heian-period worship of Kamatari and reassess some assumptions regarding his famous miei.

The first three sections of this article focus on the two portrait statues of Kamatari installed at Tōnomine in the Heian period (794–1185), up until the older one was destroyed in a fire provoked by the infamous 1208 (Jōgen 2; 承元二年) attack by Kinpusen 金峯山.2 While previous research has investigated the meaning of statue’s cracks, and investigated the painted portraits of Kamatari, their iconography, and ritual context in the Muromachi period, comparatively less exists in the English language on these two statues—to the point that, generally, scholarship talks about Kamatari’s ancestral statue in the singular. Part of this is due to the fact that official sources are ambiguous about it. The Brief Chronicle of Tōnomine (Tōnomine ryakki 多武峰略記; DNBZ 118; henceforth ryakki), a record of the site’s history compiled by the clerical supervisor (kengyō 検校) Seiin 靜胤 (fl. twelfth century) in 1197, first declares that one statue (ikku 一躯) of Taishokan is enshrined in the Shōryōin 聖霊院, but then introduces two (DNBZ 118, p. 504). Later sources, such as the fifteenth century Origins of Tōnomine (Tōnomine engi 多武峰縁起; DNBZ 118, pp. 477–84), only discuss one, the one that eventually was destroyed in the early Kamakura period, which is presented as still extant. Aside from contributing to the idea that only one miei existed, this has naturally also informed the notion that the oldest icon was the one cracking.

The first aim of this article is that of problematizing these views. I do so, first, by introducing the main narrative regarding the construction of the two portrait statues as presented in the ryakki, as this is the earliest extant institutional history of Tōnomine. Then, in the third section, I look at the way in which they were individually cared for during conflagrations that threatened their survival. Here, I rely primarily on the descriptions of major incidents found in the diaries of Fujiwara courtiers. This is because, interestingly, the ryakki omits all information regarding the separate handling of the icons that took place whenever the mountain was under attack. This examination shall reveal that the two icons indeed constituted a single body, but one that could “split” and duplicate if the circumstances necessitated. Furthermore, this duplication constituted a pattern followed by Tōnomine clerics whenever the site was under attack. Building on this point, I move to interrogate the extent to which the separation of the statues and the cracks were related, an issue that I tackle in the fourth section. Although leading to tentative conclusions only, this investigation is still functional to an understanding of the divinatory practices surrounding the statue of Kamatari, which emerged in the late Heian period as a response to unpropitious and uncanny events.

In the fifth and final section, I shift attention to the aftermath of the 1208 fire that resulted in the destruction of the older portrait statue, by examining a selection of passages from the diary of Konoe Iezane 近衛家実 (1179–1243), the Diary of the Inokuma Chancellor (Inokuma Kanpakuki 猪隈関白記; DNKR 15). I look specifically at exchanges between different Fujiwara members reflecting on the proper etiquette to follow when an ancestral statue is destroyed, and on the nature of the remains of the burned miei. This analysis has two aims. First, it wishes to bring into focus the nature of the medieval beliefs and practices dedicated to Kamatari at a time in which he was yet to receive the title of Daimyōjin. In the past, scholars have debated over the nature of the early cult of the Fujiwara statesman. Given the well-documented institutional links with Hieizan and the influence it had on the development of Tōnomine in the tenth century (Grapard 1984, 1992; Groner 2002), Grapard has interpreted the medieval period narratives surrounding Kamatari’s life, the funding of Tōnomine, and the development of specific rites at this site as the product of this Buddhist milieu. At the same time, he also suggests that elements of ancestral worship were also part of the Fujiwara statesman’s veneration (Grapard 1984, pp. 251–52). My contention here is that these ancestor worship elements were clearly couched in Confucian ideas and practices. Secondly, and more broadly, this section will also show the extent to which divinatory practices were essential not only in establishing whether a future course of action is correct but also in making sense of past ones. In both cases, however, it will be shown how, for any divination to occur, sufficient reliable precedents needed to be provided, making mantic operations dependent upon historical sources and record-keeping.

2. The Double Body of Kamatari

Very few sculpted images of Kamatari have survived. Kuroda, who has cataloged over one hundred of his portraits, could only identify seven sculptures in total, the majority of which seems to date to the late Tokugawa period (1603–1867).3 The only early example is the main miei of Tōnomine, which dates to the tenth century and is supposedly still enshrined in the honden 本殿 at Tanzan shrine 談山神社. However, this statue is hidden from view and inaccessible to the public (Kuroda 2007, pp. 47, 91), making its conditions, iconography, and dating hard to assess. As I shall discuss in this section, the textual sources only partly mitigate the lack of direct access to the actual object. Even in this case, the earliest extant institutional record offering insight on the construction of both this and the earliest icon, is the Tōnomine ryakki. While the actual details provided are scant, the text at least provides a narrative that helps to make sense of the icons’ significance and agency.

As far as the oldest miei is concerned, what the ryakki presents is likely the result of a pious legend, which took shape over the arch of the eleventh and twelfth centuries, a time in which, as we shall see, one of the two statues begins to crack. It is possible that Seiin, the compiler of the ryakki, knew little about this statue and, thus, pieces together information drawn from written records and oral transmissions. Among the different tenth-century sources he quotes, several are associated with Buddhist clerics who had acted as either kengyō and then heads (zasu 座主) of Tōnomine (or Myōrakuji 妙楽寺, as the site was originally called) before him, or by one of their close disciples. The most referenced when discussing the statues are three. The Ninishiki 荷西記, purportedly the record of the monk En’an 延安和尚 (fl. ninth century),4 which was committed to writing by a disciple in 917 (Abe 1983, p. 18);5 the Yōki 要記, the records of Jisshō 実性 (892–956), a cleric who had trained at Tōnomine from an early age, but also spent time at Hieizan, and went on to have a key role in the flourishing of the shrine-temple complex (Groner 2002, pp. 327–29); and the Kōki 後記, a “postscript” added by the fifth zasu Senman 千満 (d. 981). While these sources survive only in fragmentary form, it is clear from Seiin’s wording that his predecessors also relied on oral transmissions and local lore to address the establishment of the site and its sacred material culture.

The statues of Kamatari and their vicissitudes are briefly discussed in the second scroll of the ryakki, as part of the third section, focused on the Shōryōin, the hall dedicated to worship of Kamatari. However, the information here provided is best understood in the context of the foundation of the sanctuary, articulated across several sections of the first scroll. Here, Seiin foregrounds the way local geography is intimately linked to the physical body and the spirit of Kamatari, but also to his firstborn, the monk Jōe 定恵 (also 貞慧; ?–665), who is credited with building the first worship hall and installing the first portrait statue. According to Seiin’s account, old records (kyūki 旧記) maintain that Jōe, while training on Mount Wutai (Wutai shan 五台山) in China, dreamt about being suddenly whisked to Tōnomine.6 There, he encountered his father who, after announcing he had passed away, instructed him to build a temple and a pagoda on that very peak. By doing so, Kamatari declared, his spirit would descend on the mountain, protect their family’s descendants, and help disseminate the teachings of the Buddha. Following these instructions, Jōe returned from China, and created a mausoleum (reibyō 霊廟) for his father in the very location he was revealed in the dream (DNBZ 118, p. 490b). This entailed relocating and interring there Kamatari’s human remains (igai 遺骸), on top of which the site’s famous thirteen-story pagoda was erected (Ibid., p. 486a–b), and constructing the first worship hall (goden 御殿; Ibid., p. 503a). The latter was initially a simple small hut measuring three jō 丈 (approximately three square meters), built next to a large tree, said to emit an unusual glow (DNBZ 118, p. 503a). (As we shall see later in the article, this tree will play a crucial role in the survival of one of the two statues in 1208). In this hall, Jōe enshrined a reizō 霊像 (lit. a spirit icon), made by a sculptor of Omi province called Takao Maru 高男丸 (Ibid., p. 504b). Such is the extent to which the ryakki discusses this icon. However, thirteen-century records, such as the Inokuma Kanpakuki and the Shoji engishū, suggest that the icon was made of wood and measured three shaku (san shaku 三尺; DNKR 15:4, p. 194; DNBZ 118, p. 50a), a size I problematize later in the article. Unfortunately, however, none of these sources discuss the statue’s iconography, that is, whether it portrayed Kamatari in a standing or seated position or elaborate on any of its formal attributes.

Despite the paucity of information, annotations and discussions found in diaries of members of the Fujiwara family compiled between the eleventh and thirteenth centuries confirm that this icon indeed existed, although it is unlikely it had been commissioned by Jōe. Scholars generally agree that the foundation narrative outlined in the ryakki, albeit found across later sources on Tōnomine, is not historically accurate. For one, the dates of Jōe, an Hossō monk who spent time in Tang China, are far from being settled; different sources offer very different timelines for his life, ranging, as far as his death is concerned, from 665 to 714, with the former generally considered the most likely (Grapard 1992, p. 48; Bingenheimer 2001, p. 115; Bauer 2018, fn. 16, p. 209).7 This, however, means that Jōe passed away four years before his father, undermining the very premise of the ryakki. Unsurprisingly, in this regard, hagiographical scriptures such as the Chronicle of Kamatari (Kamatari den 鎌足傳) and Chronicle of Jōe (Jōe den 貞慧伝) make no mention of the dream of Jōe, the plea of Kamatari, or the funding of an ancestral site for the Fujiwaras on the mountain.8

Irrespective of its historical veracity, Grapard sees the attribution of the founding of Tōnomine, and by extension the creation of the first portrait of Kamatari, to a Buddhist monk as representative of the Buddhist origin of the cult (Grapard 1984, p. 253).9 This is further corroborated by the other elements of the account—the construction of a pagoda to house his remains, the institution and performance of Buddhist memorial services, which continued to be performed in the course of the Heian period, and the identification of the early site of worship as a temple, Myōrakuji (DNBZ 118, p. 485a). As hinted at in the introduction, the tenth century marked an important shift for the sanctuary, as it came under the influence of Enryakuji 延暦寺, whose monks, starting with the previously mentioned Jisshō and fellow Dharma brother Shinshō 真昇 (fl. tenth century), contributed to its development and expansion (Grapard 1984, pp. 252–53; Groner 2002, pp. 327–29). In this regard, the ryakki indicates that in 914 (Engi 14 延喜十四), the goden was reconstructed and officially named Shōryōin under the supervision of Shinshō and following a request by the then head of the Fujiwara clan, Tadahira 藤原忠平 (880–949; DNBZ 118, p. 503a), suggesting a move towards the officialization of Kamatari as shōryō 聖霊, that is, a protective ancestral figure.10 Fifty or so years later, a new icon of Kamatari was commissioned.

Although the sources do not reveal much about the conditions of the icon built by Takao Maru during these tenth-century reconstructions, Seiin writes that, under Senman, a new portrait statue was made “for the prosperity of the Fujiwara clan and the protection of the mountain” (DNBZ 118, p. 504b). This icon was realized by the busshi 仏師 Enso 延祚 (fl. tenth century), a sculptor from Tōdaiji 東大寺 responsible for several other statues at Tōnomine, including a Yaksuhi 薬師, Shaka 釈迦, and Amida 阿弥陀 triad, and a gilded Monju 文殊, which were installed in 969 (Anna 2 安和二年; DNS 1:19, pp. 4–5).11 Considering that Senman oversaw Tōnomine for a total of twenty-one years, between 960 and 981, it is, thus, likely that the new miei was indeed created in the second half of the tenth century.

Seiin’s brief remark reported above hints at the need for a new miei, but whether this was due to the poor conditions of the oldest is never explicitly stated. Instead, we are told that, eventually, the first statue made by Takao Maru was placed inside one commissioned by Senman (DNBZ 118, p. 504b), which may support the idea that the former had been damaged. When this deposit happened exactly is also not indicated, and Seiin merely states that this information comes from a “transmission of the elders” (korō sōden 古老相伝), an expression that recurs several times in the ryakki to shed further light on unusual events, such as the rumblings of the graves of Kamatari, which may not have been fully explained in the written records mentioned above.12 The only rough indication is a remark that, at the time of the Jōan era fire (1173, Jōan 3 承安三年), the icon had already been kept hidden for “quite some time” (DNBZ 118, pp. 504b–505a). This may suggest that the insertions of the original miei into the new one could have happened, if not when the second statue was installed, at some point between the late tenth and early eleventh centuries.

As anticipated in the introduction, in the ryakki, only one icon of the Taishokkan is said to be enshrined in the Shōryōin. This is unsurprising as, generally, statues with deposits are considered and accounted for as a single entity. Thus, for all intents and purposes, Kamatari had a single body, endowed with an inner and outer part, as many other statues at this time were. Statues with deposits were common in Japan at this time and are found more broadly across the East Asian context.13 Relics, scriptures, images, replica of the viscera, and other items were inserted in dedicated cavities for various purposes: to animate a newly made icon, to allow devotees to establish a bond with a deity, to safeguard important material culture, and even to dispose of sacred remains. However, what is interesting in this case is that Fujiwaras’ diaries compiled before and at the same time as the ryakki openly talk about two icons, which are differentiated as “visible portrait statue” (omote miei 表御影) and “principal” (hon miei; also, honmikage 本御影) or “hidden portrait statue” (ura eizō 裏影像; DNKR 15:4, p. 168, 187). To me, the fact that members of the clan were aware of the presence of two icons, which they distinguished based on their status and visibility, is the first indication that the emplacement of the inner icon was not permanent, an issue I address in the next section. Here, a further look at the information provided in the sections on conflagrations (enjō 炎上; DNBZ 118, pp. 487b–490a) and on the Shōryōin (Ibid., pp. 503a–505a) can help clarify both the timeframe and the movements of the two icons in the aftermath of destructive events. This will also reveal how, while it is generally assumed that, once placed inside a statue, deposits were not routinely taken out, if at all, the example of Kamatari’s statue was different.

3. Separation. Safeguarding the Body Kamatari

Between the eleventh and early thirteenth centuries, four major fires occurred at Tōnomine, all the consequence of skirmishes with other religious sites. The first three, involving Kōfukuji’s 興福寺 armed crowds, took place in 1081 (Eihō 1 永保元年), 1108 (Tennin 1 天仁元年), and 1173 (Shōan 3 承安三年) and are all recorded in the ryakki.14 The fourth, provoked instead by Kinpusen, occurred in 1208 (Jōgen 2, 承元二年) and is discussed in quite some detail in the diary of Konoe Iezane. In this section, I investigate the ways in which the two statues were handled during these incidents, by looking at a selection of diaries and accounts, and suggest that they may reveal a recurring pattern.

In the ryakki, Seiin records the events and damages suffered by the shrine-temple complex on the first three occasions in quite some depth. In fact, the chapter on conflagration is possibly one of the longest of the entire text (DNBZ 118, pp. 487b–490a). However, the cleric never explicitly mentions the fate of the icon(s), even when the Shōryōin is directly affected by the attacks; he simply notes how, to allow for refurbishments to the building, the icon(s) was either moved to another hall or to a temporary shrine. Based on this source, it would, therefore, seem that both statues were moved together as one, and always managed to escape the flames unscathed. However, Fujiwara’s diaries show another side of these events, and indicate that monks, scrambling to ensure the safety of ancestor’s bodies, often separated them and, in the process, also ended up taking questionable decisions.

A case in point is what happened during the incident of the third month of 1081,15 which is discussed in quite some detail in the Record of the Horikawa Minister of the Left (Suisaki 水左記), the personal diary of Minamoto no Toshifusa 源俊房 (1035–1121) and in Minamoto no Tsunenobu’s 源經信著 (1016–1097) Notes of Provisional Governor-General (Sochiki 帥記), both of which record the facts in real time. These sources reveal that, fearing the older miei would be damaged or destroyed, Tōnomine clerics decided to move it but, in the process, one of the two statues cracked, and it was unclear whether this was an incident or not (STR 11, p. 100). In fact, the crack happened not once but twice, with clerics becoming aware of the first fissure on the seventeenth day. Instead, the second appeared on the twenty-first day (Ibid., pp. 97–101). A dispatch sent from the mountain to Toshifusa, transcribed in his diary as part of the entry for the twenty-third day of the third month, presents the issue as follows:

Regarding the relocation of the sacred body, although it was decreed long ago that the small gotai would be stored inside the larger one, the former was taken out of the latter. During this time, cracks appeared on the face (or surface) of the bigger statue. Although the reasons why this happened were investigated, the monks at the temple said they did not know the causes. In terms of precedents, these are comparable to the crack of the Eishō era, which occurred because of the deadly fire at Yamashinadera (STR 11, p. 136b).16

Thus, while the crack could have appeared because of the move or, even, due to the separation of the two statues, it is also explicitly linked to a previous ominous event, a 1046 fire that resulted in the destruction of another ancestral temple of the Fujiwara family, Yamashinadera 山階寺 (another name for Kōfukuji). I shall go back to this issue in the next section, and instead focus here on other significant aspects of this short passage.

First, the text confirms that, already in the late eleventh century, that is, roughly one hundred years before Seiin’s redaction of the ryakki, the older icon was kept inside of the newer one, and this was presented as something that had been established “a long time ago” (maemae 前々). Second, counterintuitively, it was not the older miei, the one supposedly sponsored by Kamatari’s son, that cracked, but the newer one, commissioned by Senman in the tenth century, that did. Finally, and most importantly for the purpose of the current discussion, the emplacement of the older miei, albeit seemingly customary by his time, was not necessarily permanent. On this and other similar occasions, not only was the statue moved, but the hidden icon was taken out of the outer one, clearly in the hope of diminishing the chances of both items being damaged or destroyed.

For example, Kujō Kanezane 九条兼実 (1149–1207), preoccupied with the consequences of the 1173 fire, reflects on what had occurred a hundred or so year earlier and offers further context. In an entry of his diary, the Jeweled Leaves (Gyokuyō 玉葉), dated to the fifth day of the seventh lunar month of 1173, he reports conversations on how to deal with the statue(s) of Kamatari during this third major attack. At this time, some believed that moving them to a safe location, as had happened in the past, accompanied by timely divinations, would be the best solution. Others, however, recalled that some of the choices made in 1081 needed to be avoided at all costs. It is explained that, at that time, Tōnomine monks fled with the inner icon deep in the mountain and, in order to hide it, dug a hole in the ground, laid down a wooden plank, seemingly to protect it from the direct contact with filthy soil, placed the statue there, and covered it with earth. Kanezane remarks that this was a grave taboo (kinki 禁忌), which had hitherto not been recorded and that people preferred not to discuss (KKS 1, p. 303a). This incident is indirectly confirmed by Toshifusa. The kuge acknowledges how, in the aftermath of these events, divinations were held to establish whether the actions of Tōnomine’s cleric went against protocol or not. His concern was not so much the fissures on the face of the outer icon (as we shall see later, these were also divined), but rather the fact that the inner one ended up being placed unceremoniously on a simple slab of wood and covered with earth. This is clearly the taboo Kanezane eventually wrote about, for which offerings of cloth (heihaku 幣帛) were immediately sent to the ancestral shrine on the mountain (STR 11, p. 136a).17

Instead, during the second incident, which took place over several days in the ninth month of 1108, numerous buildings were set on fire (DNBZ 118, p. 488a–b). Although the ryakki indicates that the Shōryōin was only marginally affected, it still required minor reparations and so the statue of Kamatari was moved to the lecture hall (kōdō 講堂) for a period of time (DNBZ 118, p. 503a). While this is the extent of the ryakki’s explanation, Fujiwara no Munetada 藤原宗忠 (1062–1141) paints a dramatic picture of the events and reveals a palpable anxiety over the fate of the ancestral icon(s). In his diary, the Chronicle of the Minister of the Right (Chūyūki 中右記), he mentions being summoned by the kanpaku Fujiwara no Tadazane 藤原忠実 (1078–1162) at dawn, a few days into the attack (on the fourteenth day), and being immediately asked if he knew whether the “true body” (shōtai 正體) of Kamatari had been moved already or not.18 Munetada rushes to get more information, but it is only on the evening of the following day that a messenger from Tōnomine arrives with an official dispatch from the temple administrator. This confirms that numerous buildings had been set on fire and, fearing the flames would reach the Shōryōin, monks had taken the portrait statue of Kamatari out, and the whereabouts of the shōtai were being kept secret. A few days later, another missive identifies the shelter place as Higashiomae 東尾前 (STR 6:3, pp. 392–93).19 At this point, the situation looks dire. A messenger that Tadazane had sent to the mountain reports that “the ancestral shrine and the icon are still on top of the mountain, but all the halls have burned to ashes. The Buddha’s ground (ji 地) has no traces (shaku 跡). I look at the sky and my heart sinks!” (STR 6:3, p. 394a). Immediately, a divination is carried out to establish whether the course of actions taken by the monks at Tōnomine was propitious and, despite the devastation, it returns a positive answer. With the icon safe from the flames and only minor repairs to the Shōryōin—apparently only curtains, screens, and mats were destroyed, and the external structure had been hit by arrows—the image(s) is returned to its usual emplacement by the twenty-sixth day of the month (Ibid., p. 396b).20

While, in this instance, the sources do not explicitly mention that the smaller icon is removed from the larger one and moved independently, I believe we can infer that it was. For one, the dispatch received by Tadazane mentions that the statue was taken out or “extracted” (toridasu 取出), which is the same exact expression previously used by Toshifusa to explain the act of separating the two statues. Furthermore, just as he refers to the whole icon of Kamatari as “portrait statue” or mieizō 御影像 and differentiates the two using the adjectives small and big, Tadazane distinguishes between the mieizō from the “true body” or mishōtai 御正體. Although these two terms may well be used as synonyms, I would argue that they are here used to differentiate the visible, external shell from its inner form. Circumstantial evidence from a fourth incident supports this interpretation.

During the attack by Kinpusen of 1208, as a result of which the older icon would be destroyed, the two bodies were immediately separated and moved to two different locations. This time, we have more precise records of the two icons’ movements during the incident, because the then kanpaku Konoe Iezane ordered an investigation into the tragic outcome, and one of the official reports, penned and signed by the head ajari (kengyō ajari 撿挍阿闍梨) of Tōnomine, Dharma master Kangon 大法師觀嚴, is copied in his diary (DNKR 15:4, pp. 200–1). Here, we learn that the shrine custodian (goten azukari 御殿預) Gyōshin 堯心, with the help of four other clerics, took care of the visible icon (omote miei 表御影) and hid it under the large tree on the eastern side of the Shōryōin—the very same tree that emitted an unusual glow, prompting Jōe to establish the first ancestral hall next to it. The true icon (honmikage 本御影) was taken out of the omote, placed inside a chest (bako 箱) of official documents, and, after some peregrinations, transported to a hall called Dobutsudō 土仏堂. This decision is defined as being “in accordance with precedents” (nin senrei 任先例), which reinforces the idea that separating the two statues and bringing them to different locations when the cultic center was under threat was the customary thing to do.21 Once in the Dobutsudō, the chest of documents was hidden in a corner behind what are described as three large (jōroku 丈六) Buddhist statues—Shaka, Amida 彌陀 and Jizō 地蔵 (DNKR 15:4, p. 200).22 Unfortunately, however, this hall was entirely burned to the ground in the following days, and with it the Buddhist statues, the documents, and the true body of Kamatari.

When the fire subsided and they were finally able to access the premises, a group of Tōnomine clerics collected the ashes they believed could belong to the burned icons, although mixed and impossible to attribute, and stored them in a container, which was then placed next to the surviving statue at an undisclosed location in the mountain. Eventually, it was established that these ashes would be stored inside the omote, just like the original icon was, but this was decreed only after a months-long investigation of precedents and divinations (Lomi 2022). Once decisions were taken regarding the reconfiguration of the icons, offerings of cloth and prayers of dedication (kōmon 告文) were sent to three Fujiwara ancestral shrines, Kasuga 春日, Ōharano 小原野, and Yoshida 吉田, a pattern that recurs after each major event surrounding Kamatari’s statues(s). In this instance, these were delivered by three local Fujiwara clansmen—the Governor of Shimotsuke Province (Shimotsuke no kami 下野守) Fujiwara no Suemune 藤原季宗 (fl. twelfth-thirteenth centuries), the courtier (san’i 散位) Fujiwara no Sukekiyo 藤原佐清 (fl. twelfth-thirteenth centuries), and the Deputy Governor of Tōtōmi Province (Tōtōmi gon-no-kami 遠江權守), Fujiwara no Tamenaga 藤原為永 (fl. twelfth-thirteenth centuries) (DNKR 15:4, p. 204). Aside from these offerings, throughout the period of the attacks and subsequent reconfiguration of the icons, Iezane mentions repeated performances of the Benevolent-King Assembly (Ninnō-e 仁王會), but also of other rites to ward off calamities and malicious forces such as offering to Fudō (Fudō kuyō 不動供養), or the Five-Platform Ceremony (Godanhō 五壇法).23

From these different accounts capturing the actions of clerics on the mountain at times of unrest, we can infer that, since at least 1081, a procedure was in place to safeguard the integrity of the body of Kamatari. At these times, not only was the statue promptly moved, but duplicated, as the inner and outer bodies were separated. The inner and older icon, identified as the true or main body of the ancestor, was taken at a location which remained undisclosed until all menaces had subsided. The whereabouts of the outer are not always discussed, but it is clear it was not kept in the Shōryōin either. Separation and relocation not only ensured the safety of the older portrait statue of Kamatari, but also minimized the chances of both statues being destroyed, which would have threatened the continuity of ancestral worship. While the sources indicate this was a pattern, it was a flexible one. There was, for example, no single designated safehouse or fixed etiquette likely because the abrupt nature of the attacks and the unpredictability of fires required those in charge to decide on the spot where it was safe to take the two icons.

It is here that ex post facto divinations were necessary to mitigate any taboos generated by the clerics’ actions. As discussed, even well-meaning actions were not without negative consequences. On one occasion, the inner statue was placed in the filthy ground and covered in earth, on another, it was brought to a hall that ended up being burned to the ground. Divinations were thus held to reveal the extent of the taboo and what was required to avoid further calamities. Furthermore, these events compromised the wholeness of the material form of the statue in two ways which, as I argue in the next section, were interrelated if not originally concomitant: the splitting in two of the ancestral body and the manifestations of superficial cracks. These actions could be seen as both threatening the successful performance of ancestral rites and manifesting underlying ruptures and fractures in the fabric of the clan. Thus, divination was, here, not only a technology used to mollify the ancestor and solve internal tensions, but also what cemented the link between the integrity and well-being of the clan and that of the statue.

4. Kamatari Splits

The separation of the two statues at moments of crisis revealed by the sources discussed in the previous section raises questions of a practical and phenomenological nature. How, and how easy, was it to remove the inner statue without causing damage to the outer one? Was the 1081 crack a consequence of the removal of the inner body? Is there a relation between the uncoupling of the statues and the cracks? Lack of any description of or permission to examine the extant tenth-century external body of Kamatari makes it hard to know how the inner cavity of the statue was configured and accessed beyond any reasonable doubt. However, reflecting on how these icons and what the sources tell us about their dimensions fare in comparison with coeval examples with deposits may help tackle, although still tentatively, these questions.

As mentioned, neither the ryakki nor any other source sheds light on the origins and exact timing of the first insertion of the old icon into the newer. It is similarly unclear whether Senman commissioned the new portrait statue specifically to contain the one created by Takao Maru, maybe to mitigate its deterioration or damage, or whether the insertion happened at a later date. In fact, sources are vague on the matter, indicating only that it occurred “in the past” or “many years ago”. Whichever the rationale and timing, given the enlivening function that inserting old, powerful objects into newly made ones were thought to have in the Japanese context (Brinker 2011; Covaci 2016), the fact that this detail is only marginally mentioned in the ryakki, and then left out altogether in later temple myth-historical narratives, such as the Tōnomine engi, may seem rather unusual. It is possible that, as the engi was compiled in the fifteenth century, over two-hundred years after the destruction of the main miei, this earlier and infelicitous layer of the portrait’s history had been, if not entirely forgotten, at least downplayed, or put aside. It is also likely that the cracking episodes took precedence as distinctive features of Kamatari’s icon, which not only sanctioned its power but also set it apart from other ancestral and Buddhist statues. Still, this joint configuration is regularly mentioned in diaries compiled from the eleventh century onwards, suggesting that, at least at this time, it was acknowledged and considered customary.

Evidence from coeval and later examples suggests that sacred statues could be fitted with items at different stages and in different manners. Standing statues may have a removable panel or door at the back, through which items were inserted into their hollow bodies. This is the case as far as the famous Seiryōji 清凉寺 icon of Śākyamuni is concerned (Horton 2007, pp. 26–31). These small openings were generally sealed immediately after the deposit, but they could, theoretically, still allow for items to be taken out and put back in if one so wished. In the case of seated statues, ingress to the inner cavity could also be gained via its base, as these examples often had an open bottom (Washizuka 2007, pp. 92–96).24 In numerous instances, however, deposits happened in the phase of assemblage of wooden statues, without any door or immediately visible opening available, making it hard to gauge whether they have content or not. Thus, it is not uncommon for inner items to be revealed when statues are taken apart to be restored, as in the case of the Jizō statue of Denkōji 伝香寺 (Glassman 2002, pp. 386–90) and now even following unobtrusive investigations with X-ray technologies (Covaci 2016, pp. 11–14).

In the case of Kamatari’s icon, it would make sense if the tenth-century icon were to represent Kamatari in a seated posture, as this would be consistent with both coeval ancestral statuaries,25 as well as with extant painted or sculpted representations of Kamatari from Tōnomine (including the current Tokugawa-period madeachi), which all depict the ancestor seated (Ibid., p. 89; Kuroda 2007, pp. 41–68). Yet, this seems to be at odds with the sizing of the statues declared in the Inokuma Kampakuki. As previously mentioned, in this diary, the icon made by Takao Maru is referred to by Fujiwara no Nagakane as “three-shaku miei” (san shaku miei 三尺御影; DNKR 15:4, p. 194), an appellation also found in the coeval Shoji engishū (DNBZ 118, p. 50a). If accurate, this would mean that the inner icon measured around 90 cm. In the same sources, the external, so-called “bigger statue,” is instead labeled tōshin 等身, that is, “life-size”. According to Mimi Yiengpruksawan, “life-size” commonly denoted an icon meant to reproduce the normal physical proportions of a person, which meant around 160 cm for a standing statue and 90 cm for a seated one (Yiengpruksawan 1991, fn. 5, p. 330). Based on these data, if Senman’s tenth-century example were indeed seated, it would have the same height as Takao Maru’s. In this case, containing the older icon would only be possible if the latter were broken. Yet, even if it were standing, I doubt this statue could have a back opening tall enough to get a 90 cm object in and out whenever needed without a major disassemblage. In this sense, if these dimensions are correct, and the inner statue was intact, there is no scenario that would have allowed Tōnomine’s clerics to remove the shōtai without cracking open the omote (pun intended).

At the same time, there are reasons to doubt these measures. For one, 90 cm is a rather unusual size for a deposit. Sculpted images found inside statues’ cavities tend to be considerably smaller, measuring between 4 to 25 cm, with only remains of damaged icons suggesting a bigger original size.26 Furthermore, Heian-period sources, compiled when both icons were still extant, never explicitly say that the older statues measured three shaku, they simply call it the “small” icon—and I am not sure that 90 cm qualifies as “small”.27 The three-shaku appellation is first found in the Inokuma Kanpakuki, where it is used only once. Finally, when Tōnomine clerics attempt to protect the statue in the 1208 attack, they do so by placing it inside a chest of documents. This would have needed to be a rather big trunk to accommodate a statue of this length. It would have certainly been quite heavy too, making it an impractical container to move the statue swiftly and inconspicuously during an ongoing attack. Furthermore, we are told that only one person, Nengen Shōnin 念玄聖人, carried this crate to the Dobutsudō—an impossible task if the measures were correct. For these different reasons, I wonder whether Nagakane may not have been conflating or confusing the sizes of the two statues—the omote, if indeed a seated tōshin, would have measured 90 cm. Alternatively, three-shaku could have been used to indicate the size that the inner icon had before suffering some kind of damage, which drastically reduced its dimensions and maybe even justified its hidden status—although none of the sources ever hint at the inner icon being nothing but wholesome and, as mentioned, this length is only found in a handful of sources.

As things stand, it is impossible to ascertain exactly how the two statues fitted and were separated without investigating the extant icon. I am partial to the scenario in which a seated life-size statue of Kamatari of roughly 90 cm contained a smaller true icon, which could be removed via the base of the former, that is, by lifting and removing. Not only this seems to be more in line with coeval ancestral images, and standard representations of Kamatari produced at Tōnomine, but the act of “lifting” and “extracting” resonate with the way in which the sources talk about the separation process at the time of the attacks.

Even if the clerics in charge of the Shōryōin may not have had to drastically intervene on the material dimension of the outer body to bring the inner to safety, the fact remains that the earliest extant records we have of the cracking phenomenon, that is, the annotations found in the Suisaki and Sochiki, also happen to be the earliest record of the separation of the two statues. According to Kuroda, the account of the events provided in the Suisaki, which I have translated in the previous section, suggests that this was likely an accident or, as the sources phrase it, a natural occurrence, and that it is only in later sources that it was recast as an omen (Kuroda 2007, pp. 175–76). Furthermore, a closer look at the event as presented in the Suisaki shows deep concern for the fact that the statue cracks twice over the arch of a week, an event that was likely the first of its kind.

In the entry for the nineteenth day of the third month, Tsunenobu indicates that there had been daily reports of rumbling noises coming from the tomb of Kamatari, which had everyone concerned and, therefore, needed to be divined. He also adds that, a few days earlier, a crack was discovered on the face of the icon, and it was unclear whether “old precedents” (kyūji 旧事) of this type of phenomenon existed. The Onmyōji Kamo no Dōgon 賀茂道言 (d.u.) decreed that the divinations should be carried out only in light of existing precedents and, although there was no agreement, it was eventually ruled against it. However, when another crack appeared on the twenty-first day, the matter escalated. Tsunenobu writes:

The regent said: the repeated breaks of this icon are very suspicious; I wonder if they occurred naturally when the icon was being carried out. However, the matter was not thoroughly investigated, and there are frequent complaints. This being the case, how can we request a divination?

(STR 11, p. 100a)28

It is undeniable that the Fujiwaras wondered whether the statue broke because of the move, and, therefore, whether it was simply an accident requiring no further action. However, a concomitance of events—the attack, the fire, the removal, and placement of the inner icon in a ditch in the ground, accompanied by constant howling of the ancestral tomb—contributed to growing pressures to interrogate the meaning behind this occurrence. Against this context, the previously translated passage of the Suisaki, in which a dispatch from Tōnomine argues that the cracking event is comparable to something that had happened forty years prior and coinciding with the fire at Yamashinadera should be seen as an attempt to identify a precedent against which divinations could be held. Finally, the kanpaku Fujiwara no Morozane (藤原師実; 1042–1101) requested to carry out divinations aimed at establishing whether this crack was the result of tatari 祟り, which returned a negative outcome (STR 11, p. 100b). This is the only instance I encountered in which the cause of the cracks is explicitly articulated in this way.

Unsurprisingly, when we turn to the Tōnomine ryakki, the twenty-fourth day of the first month of Eishō 1 永承元 (1046) is identified as the first occurrence of this phenomenon, and it is the only instance in which detailed information is given. From the length of the crack (4 son 寸), to the rites held, and the officiants involved, much attention is given in providing evidence of this event. However, all mentions of Yamashinadera have carefully been removed (DNBZ 1118, pp. 510b–511a).29 So far, I have found no mention of this cracking event anywhere in sources coeval to this incident, or compiled before the Suisaki or the ryakki. I am thus inclined to think that this was an entry mixing real information drawn from sources outlining the ritual etiquettes followed in the case of the fire at Yamashinadera, with later measures of the crack. Although curated, the entry was not necessarily written maliciously. Rather, it could have been the product of the nature of the sources Seiin relied on, and driven by the previously mentioned interest in finding a precedent for the 1081 crack. Although further research to prove this is needed, I would consider this latter event as the one which marked the beginning of two “splits”: the separation of the body of Kamatari and the cracks of its exterior material form.

5. Ancestral Tablets, Withered Wood, and Dead Ashes: Reconfiguring the Statue Kamatari in the Early Kamakura Period

The disastrous outcome of the 1208 attack prompted an official investigation, which was aimed, on the one hand, at establishing culpability for the loss of the statue of Kamatari—was this partly to blame on due to negligence of Tōnomine’s clergy?—and, on the other hand, at deciding whether a new image should be commissioned and what to do with the ashes retrieved from the burned Dobutsudō. To tackle the latter two points, Kiyohara no Yoshinari 清原良業 (1164–1210), who held the position of the Senior Secretary (daigeki 大外記) at court, was tasked with collecting precedents from both Chinese sources and local records and compiling a report. Although I have not been able to locate the actual document, the salient points emerge from Iezane’s diary, as a good portion of the entries for the fourth month of Jōgen 2 (1208) is dedicated to the opinions of a group of notable members of the Fujiwara family,30 all explicitly referring to Yoshinari’s investigation. These passages, drawing heavily on Chinese sources, are clearly couched in Confucian ideas. Instead, never once is the figure of Vimalakīrti, of whom Kamatari was believed to be an incarnation (Grapard 1992, p. 238), mentioned or the Vimalakīrti Assembly (Yuima-e 維摩會), traditionally linked to the healing of the Fujiwara ancestor and often held in his memory,31 performed.

Regarding the possible commissioning of a new portrait statue of Kamatari, it was noted that previous cases offered a range of possibilities. However, Yoshinari’s research revealed that, while Buddhist icons were generally reconstructed and kami representations or objects were not, this was not a hard-and-fast rule. As Fujiwara no Nagakane 藤原長兼 (d. 1214) notes:

The geki’s report shows that, although there are precedents for both reconstructing and non-reconstructing, they hardly apply uniformly to shrines and temples. When it comes to kami, the “as if present” rite takes precedence, so even though nothing is rebuilt, ritual etiquette is enough. When it comes to Buddhas, making sculptures is a meritorious act, yet one should not haphazardly build anew

(DNKR 15:4, p. 193).32

Here, Nagakane points to the presence of a specific ritual etiquette allowing for the worship of spirits in absentia, the nyozai no gi 如在之儀, which would theoretically dispense from the reconstructions of lost sacred items. Identified as a practice that applied to spirits only, and not Buddhist icons, Fujiwara no Chikatsune 藤原親経 (1151–1210) also remarks that this is why numerous shrines lack an actual object or “seat” for the kami (DNKR 15:4, p. 191). The nyozai no gi derives from a famous Confucian exhortation to sacrifice to the ancestors and the spirits “as if [they were] present,” found in Analects 3.12 (Slingerland 2003, p. 21). Yet, this sentence, which has been the topic of much scholarly debate,33 is now generally taken to mean that one should carry out sacrifices with a sincere attitude of reverence, being “fully present” as if the spirits and ancestors were actually in front of you. Thus, it was originally meant as an admonishment to maintain a proper psychological disposition towards ritual actions, rather than an exemption from the creation of material forms for the ancestors or spirits.

However, Yoshinari and the Fujiwaras did not necessarily have only the Chinese context in mind. In Heian-period records, this term is used to indicate instances in which the Tennō could not be present to preside over prescribed rites due, for example, to illness or prohibitions, but these events went ahead as if he were actually there.34 Hori has further discussed how, by the eleventh century, the nyozai no gi had also come to indicate the protocol in place to ensure continuity and successful transfer of power, if the Tennō were to die before a successor had been selected and enthroned. In these circumstances, the death was not officially announced, and the corpse of the deceased ruler was treated as if alive until after his successor’s accession (Hori 1998, pp. 38–69).35 Against this context, mentioning the “as if present” rite as a precedent for dealing with the loss of the ancestral statue ends up equating Kamatari’s “true body” miei to the living body of the Tennō. As such, the statue is extended the same prerogatives of the ruler, meaning that just as imperial rites can sometimes take place without the Tennō, the ancestral veneration of Kamatari could theoretically continue without any material support. However, this should not be taken as a unilateral commitment to immateriality, which, given the circumstances, would be rather disingenuous. As we have seen, by this time, the Fujiwaras had been paying close attention to the welfare of this statue for well over a century, carrying out divinations if it needed to be moved, and demanding to know its exact whereabouts whenever there was turmoil on the mountain. Instead, the question as to whether a new main image was needed or not hinged on establishing a proper course of action based on the nature of the icon and on historical precedents.

This emerges quite clearly in the context of a discussion of the conventions surrounding spirit tablets in Chinese ancestral temples (kanka sōbyō 漢家宗廟). Hino Sukezane 日野資實 (1162−1223), always building on information provided by Yoshinari, claims that, although Chinese ancestral temples were affected by fire incidents, “instances of spirit tablets lost in fires are rare” and their reconstructions even more uncommon (DNKR 15:4, p. 190). In fact, he claims that, by the time of Emperor Suzong of the Tang (唐肅宗 r. 756–762), the imperial ancestral tablets had only been rebuilt twice (DNKR 15:4, p. 190).36 As Nagakane remarks, this is due to the fact that the Chinese equipped their ancestral shrines with a stone chamber for ancestral tablets (sōseki 宗祏), which guaranteed the safety of these wooden items even in the face of major fires.37 Thus, he wonders if Kamatari’s miei should not be treated like an ancestral tablet after all (DNKR 15:4, p. 194).

Certainly, likening an ancestral portrait statue made of wood, believed as it was to be the real body of Kamatari, to a wooden spirit tablet (mokushu 木主), which was also thought to materialize the presence of the ancestor, is not far-fetched. While these two objects are formally very different, and this was not certainly overlooked, they could still be seen to serve the same function.38 This association, however, reveals much about the sources Yoshinari relied on to redact his report on how to handle destroyed icons. The lack of any reference to ancestral portraits (whether painted or sculpted) is unsurprising, as in Early China, portrait statues of ancestors were not utilized for memorial rituals, and much of the precedents cited in Iezane’s diary seem to be drawn primarily from Han and, to a lesser extent, Tang dynastic records. Although Wu Hung suggested that the ritual use of ancestral portraits can be traced back to the Han (202 BCE—220 CE), Patricia Ebrey has instead contended that the presence of portraits in memorial contexts does not emerge until the fifth century at least (Ebrey 2004). Still, extant examples are scant, specifically when it comes to portrait statues. In this case, as she demonstrates elsewhere, written sources point to the existence of rituals centered on ancestral portrait statues in the Song (960–1267), although these are limited to Imperial ancestors (Ebrey 1997). From the exchanges reported by Iezane, there is no evidence that the Fujiwaras knew of the Song-period cult of ancestral statues or for example, were aware of the predicaments faced by the court to bring these icons to safety or to duplicate them as a result of the Jurchen invasion of the early eleventh century, which threatened the continuity of the practice (Ebrey 1997, pp. 71–77).

Still, ancestral portraiture was not unknown at this time in the Japanese context, and both Yoshinari and the Fujiwara kinsmen were necessarily aware of the existence of similar sculpted items, first and foremost the statue of another key political and religious figure, Shōtoku Taishi 聖徳太子 (574–622), installed at Hōryūji 法隆寺. Yet, they never once mention it. Built at the beginning of the twelfth century, this statue also has a deposit: a sutra box with scriptures, and an eighth century statue of the Bodhisattva Kannon standing on a rock formation, which in turn rests on a turtle (Carr 2012, pp. 36–40). At first, the absence of references to this statue is puzzling, as it shares several common traits with Kamatari’s. Both are representations of historical figures venerated as ancestors, both are considered shōryō 聖霊—protective sacred spirits—and enshrined in a homonymous hall, and both have double bodies. Yet, key differences set them apart. Even if Hōryūji’s Shōtoku has a double body, this ensemble functions as a three-dimensional rendering of honji-suijaku 本地垂迹: the visible statue, representing the prince in Confucian attire, is nothing other than the external manifestation of its inner original ground, the Bodhisattva of Compassion Kannon (Ibid., p. 38). As far as one can tell, this was not how the double body of Kamatari at Tōnomine was conceived, as both the inner and outer bodies are unequivocally “described” as portraying the Fujiwara ancestor. Furthermore, while numerous other statues of Shōtoku were produced, depicting the statesman in different forms, the same cannot be said of Kamatari.

Beyond the preoccupation with precedents, the issue of the reconstructions was discussed in relation to the presence of both the surviving omote and the ashes of mishōtai. In fact, the Fujiwaras’ concern related to both the reproducibility of the ancestral body, and the identification of the proper ritual etiquette to renegotiate the relationship between the destroyed statue and its remains, and the intact one. Many in fact believed that commissioning a new icon when one, revered for many years and perfectly efficacious (as mentioned above, this image was the one cracking), had survived the fire unscathed, was both practically impossible, and against another custom pertaining to ancestral tablets. Sukezane remarks:

Now, we need to carefully consider all the implications of placing [the remains] of the reizō inside a newly made icon. In terms of features, who is the contemporary craftsman who can reproduce its physical proportions? Furthermore, Yoshinari’s report says that a temple should not have two masters and the precedent of Duke Huan of Qi is unique and does not apply here

(DNKR 15:4, p. 190).39

With this statement, the kuge draws attention to the fact that remaking an icon that had been hardly ever visible but to a selected few since at least a century, and of which nothing remained, is a challenge that should not be underestimated. This, in turn, indicates that, as a portrait statue, the way the icon looked like mattered. Attempting to carve a new copy of what was considered as the true likeness of the ancestor, without having access to a blueprint of the original, was, thus, out of the question. In this regard, the issue of the creation of a new icon was not as straightforward because of the uniqueness of the icon.

Then, he draws from a renowned passage which Yoshinari lifted from the Book of Rites (Liji 禮記) to, again, argue that not rebuilding is the cautious thing to do. In this text, the quote “a temple cannot have two masters” is found in a chapter containing a series of exchanges between Confucius and his disciple Zengzi on the matter of ritual propriety, aptly called “The Questions of Zengzi” (Zengzi Wen 曾子問). In one instance, Zengzi asks whether it is possible for a shrine to have two ancestral tablets for the same ancestor, to which Confucius replies:

In heaven there are not two suns; in a country there are not two kings; in the seasonal sacrifices, and those to Heaven and Earth, there are not two who occupy the highest place of honor. I do not know that what you ask about is according to rule. Formerly Duke Huan of Qi, going frequently to war, made fictitious tablets and took them with him on his expeditions, depositing them on his return in the ancestral temple. The practice of having two tablets in a temple-shrine originated from Duke Huan

(Legge 1885, p. 323).

By implicitly maintaining the parallel between the statue of Kamatari and an ancestral tablet, Yoshinari’s reference to this quotation manages to kill two birds with one stone. On the one hand, it shows that, in principle, it is possible to have two icons of the same ancestor installed at a temple. Therefore, the fact that two statues had coexisted up until the recent incident was not against protocol. However, it also indicates that, in general, one should be the rule, and this serves the current circumstances well. With one statue left, the ancestral temple now only has one lord.

In a way, this comparison seems to cater to overlapping concerns. On the one hand, it acknowledges that the icon of Kamatari is somewhat different from a Buddhist statue and, at the same time, also unlike the material support of a kami. As the shadow body and portrait of the ancestor, it is closer to the symbolic body of the Tennō and the material one of the ancestral tablets. On the other hand, it draws from Chinese practices to address the lack of a consistent etiquette to deal with these types of events, something that seems to apply to all kinds of sacred material culture. Sukezane presses this point by reaching back to the damaging of the sacred mirror in the second year of the Kankō era (寛弘二年, 1005), and arguing that, even in the case of imperial regalia, although these can be remade, no official procedure exists (DNKR 15:4, p. 190).40 Yet, eventually, the injunctions fund in Confucian sources not to have two lords in the same temple, not to make multiple spirit tablets for the same ancestors, and not to disrespect the surviving life-size “outer” body of Kamatari by commissioning a new one, are enough to decide not to reconstruct without carrying out divinations.

This by no means prevented the Fujiwaras from looking at examples of destroyed kami and Buddhist statues or objects to reflect on possible solutions for the conundrum they were facing. This was especially the case when it came to the issue of the disposal or not of sacred waste. As mentioned earlier, the ashes retrieved from the charred rubble at the Dobutsudō were mixed and establishing what belonged to Kamatari’s icon and what to other items in the hall was deemed impossible. Buddhist clerics immediately decided to gather and store whatever they could, keep it safe together with the surviving icon—at an undisclosed location at the center of the mountain—and await further instructions. However, it is inferred that their preferred option was to inter them in purified ground (DNKR 15:4, p. 190). This solution was perceived by some of the kuge as being demeaning and unacceptable (Ibid., pp. 190, 197), by others something worthy of consideration given both the ashes’ unclear origins and the fact that, after all, they were the result of an unpropitious event (Ibid., p. 191). Even in this instance, it seemed that the precedents collected by Yoshinari, which pointed toward the universal custom of placing sacred ashes inside dedicated containers, existing or old statues, did not fully apply. This led some to speculate on the very nature of ashes. Sukezane writes:

Often, in the Sino-Japanese context, there is no standard when it comes to the ashes of a conflagration. It is said: “ashes are just dead fire,” and “a heart like dead ashes”. Even if there are rites to pay respect [to ashes], they are not necessarily a sacred thing. Yet, monks debate whether it is possible to bury them into purified ground!

(DNKR 15:4, p. 194).41

In this case, Sukezane adopts a matter-of-fact approach to the issue of the ashes. In the first instance, he is likely quoting from the Shuowen Jiezi 説文解字, a Chinese dictionary compiled by Xu Shen 許慎 (fl. first century CE) during the Later Han, which plainly explains, under the section dedicate to fire, that ashes are what remains once the ambers of a fire have died (hui: si huo yujin ye 灰: 死火餘燼也). At the same time, ashes can be much more than mere lifeless remains, and instead indicate a higher state of being. This is expressed through the second quotation, which is an abbreviation of the Daoist expression “body like withered wood and mind like dead ashes” (xing ruo gaomu, xin ruo sihui形若槁木心若死灰), found first in both the Zhuangzi 莊子 and the Huainanzi 淮南子, and then in later Daoist scriptures as well. In the former text, it is used to indicate the quality of a practitioner called Ziqi 子綦, who has reached such a deep state of meditative absorption that his external bodily form and self have been cast off and forgotten.42 In the Huainanzi, the sentence similarly denotes the body-mind of the perfected (zhenren 真人), who have lost their “physical frames” and “know without studying; see without looking” (Major 2012, p. 78). Thus, far from being simply the deadest of dead matter, across Daoist literature, a body like withered wood and a heart-mind like dead ashes mark the successful realization of self-so-ness as well as the achievement of a transcendental, immortal state.43 Unsurprisingly, Sukezane also acknowledges the importance that ashes hold within the Buddhist tradition, but it is Tokudaiji Kintsugu who, by remarking that “after the extinction of the Buddha, his ashes were placed in a flower stupa, thanks to which people can keep venerating his remains” (DNKR 15:4, p. 197), reminded his kinsmen that enshrining these ashes may be key to ensure the continuity of the ancestral cult.

Even in this instance, the Fujiwaras acknowledge that, not only establishing if the ashes belong to Kamatari or not is ultimately impossible, but even simply whether ashes are sacred or not, or potentially dangerous or not, is also hard to know based on textual sources, ritual practices, and devotional customs alone. However, unlike the matter of the reconstruction of the statue, divination is here introduced as a means of deciding which one, among the options identified through an analysis of the precedents, is to be implemented. The question divined by a group of seven onmyōji, thus, was “should the ashes of the miei be placed inside the life-size icon, yes or no,” and the result signaled this was a propitious solution (DNKR 15:4, pp. 202–3). Following the identification of an auspicious day, the ashes are then placed in a small container, which is carefully wrapped and inserted into the surviving statue. Finally, Kamatari was whole again, but this time, with a heart of actual dead ashes, aside from a body of withered wood.

6. Conclusions

The analysis of Heian-period diaries and Kamakura-period records of Tōnomine has revealed that, between the eleventh and early thirteenth centuries, the portrait statue of Kamatari at Tōnomine was made of an inner icon, older and considered as its essential, true (shō正) body (tai 體) and an outer one, a life-size shadow image (eizō 影像) that functioned as its container. While this emplacement was eventually written out of Tōnomine’s history and, with few exceptions, is never acknowledged in contemporary scholarship, it certainly does not represent a unique case. Numerous other statues, including the portrait statue of Shōtoku Taishi, had an inner and outer body. However, unlike Shōtoku’s miei, Kamatari’s is never construed in the documents examined in honji-suijaku terms, and, unlike other coeval examples, these two bodies did not always function like a single entity.

Sources detailing the actions of clerics at times of unrest on the mountain suggest that, by the Kamakura period, an unwritten protocol was in place requiring these two parts to be separated and taken to different secure locations. This strategy was clearly aimed at ensuring that at least one of the two would survive an attack and, as the events of 1208 show, this turned out to be an ultimately successful strategy. Although, at these times, emphasis was placed on the inner icon, because older and attributed to Kamatari’s firstborn, the second dedicated by the kengyō Senman was not overlooked either. On the contrary, it was carefully and closely scrutinized. By manifesting on its surface the traces of the ancestor’s mood and dispositions, this icon was also essential to sanction or condemn the decisions and actions of both the clerics at Tōnomine and the clan.

In this regard, although the phenomenon of the miei’s cracks is well-known and has already received scholarly attention, so far, the relevance of the inner–outer body dynamic in the production of the cracks had been overlooked. The present analysis has suggested that the two likely emerged in conjunction and were thus related. As discussed, in 1081, in the midst of an arson attack, Tōnomine clerics decided to separate the otherwise joined bodies of Kamatari and remove the original shōtai from its external visible form. Immediately or shortly thereafter, the latter cracked not only once, but twice. Whether the result of a direct intervention on the material dimension of the outer icon in order to get the inner out, a mishap due to improper care during the move, or a natural response to environmental changes such as the relocation of the statue and the fact that it was suddenly empty, is impossible to establish, and further research is needed. However, it is clear from the sources that not only, up until that date, the older statue had remained hidden inside the newer one, but also that the cracks had probably never occurred before. Attempts at identifying earlier occurrences, which eventually resulted in the creation of new cracking events a posteriori, should be first understood against a context in which precedents seemed to have been necessary to carry out divinations. This is something that the meticulous scrutiny of prior cases of destruction of sacred items, which informed the report of the daigeki Yoshinari in 1208, also supports.

While, in the four cases investigated here, the ousting of the inner body coincided with the externalization of Kamatari’s will in the forms of cracks, these events did not come to an end with the incineration of the shōtai. Eventually, with its heart turned to ashes, the wooden body of Kamatari remained communicative, as indicated by sources such as the Taishokkan haretsuki (DNBZ 118, pp. 518–30). In this sense, it is only by linking together the circumstances that led to the removal of the inner statue, the act of separating the inner and outer body, and the manifestation of the crack that we can appreciate how the language of the cracks became a way of talking about past unpropitious events as well as construing future threats to the clan.

The sources also reinforce the degree to which this icon was, in the Heian and Kamakura periods, at the forefront of the clan’s concerns. Every time the cultic site was under threat, leading members of the Fujiwara clan persistently and apprehensively sought reassurance of the safety of the miei, sponsored protective rituals, and held divinations. Both the ryakki, and other coeval accounts indicate that Tōnomine was, by the end of the twelfth century, a large cultic center with a rich material culture. We have a list of the many Buddhist sculptures which populated its halls, many of which were lost to the different fires. Yet, nowhere in the diaries is the safety of these Buddhist icons ever addressed, their whereabouts traced, or the causes of their destruction as thoroughly investigated. This indicates that the statue of Kamatari was, unsurprisingly, considered as qualitatively different from the other sacred objects on the mountain.

Nowhere does the otherness of this icon become more evident than in the aftermath of the shōtai’s incineration. The elucubrations of the Fujiwara nobles recorded by Konoe Iezane in his diary on what to do with the ashes and the remaining statue derive from the absence of an established etiquette to deal with this incident. However, this is not due to the particularity of the event as such; sacred items, as Yoshinari’s report indicates, got destroyed all the time. Rather, it is linked to the fact that the object in question is an ancestral statue. Yet, in none of the instances introduced in the Inokuma Kanpakuki is the destruction of another ancestral icon mentioned, to the point that what turns out to be the closest comparable example is an ancestral tablet. So, how are we to interpret this alleged lack of precedents? Were there really no previous examples of destroyed portrait statues? While I do not doubt the accuracy of Yoshinari’s report or the sincerity of the exchanges, I also think that further research into ancestral portraiture and its destruction may help assess whether Kamatari’s case was truly one of a kind.

Furthermore, the opinions of the Fujiwara nobles recorded by Iezane, and more generally Heian period diaries outlining the vicissitudes of the ancestral bodies, foregrounds an understanding of Kamatari as an ancestral figure without ever once explicitly or implicitly mentioning his association with Vimalakīrti or any other Buddhist figure. Instead, when reflecting on which ritual etiquette one should follow in order to pay respect to an ancestral figure whose portrait has been lost, the discussion first revolves around an originally Confucian custom, which was, at the time, relevant to the Tennō rather than just any ancestor. However, while Iezane and his interlocutors use Chinese ancestral worship customs as a compass to establish proper ritual etiquette, they weigh their considerations against a broad range of practices and precedents, which involve Buddhist material culture, imperial regalia, kami worship, and even reach back to seemingly unrelated Daoist sources. Inevitably, this led to the coexistence of multiple, at times conflicting, interpretations of what the icon of Kamatari was—the body of the ancestor, an ancestral portrait, a withered piece of wood—and what its ashes were—the reconfigured body of the ancestor, sacred remains, problematic trash, just dead fire.

More broadly, it is here that one comes to appreciate just how much agreement over the nature of specific phenomena and objects depended on the mediation of divinatory practices and how much divination required the identification of precedents. On the one hand, in the course of the Heian period, members of the Fujiwara clan came to reflect and make sense of their own disquiet and decision-making through divinatory practices firmly rooted in the material dimension of their ancestral portrait statue. On the other hand, the very essence of the ancestral body, in whichever form or configuration it found itself, was also interrogated, construed, and recast through divinatory practices. Further research into the ways in which divination was used to assess the ontology of material culture is needed, as this may lead to a reframing of sacred objects (and what was left of them) as contextual and fundamentally fluid because produced by the dialectical interaction between theoretical and mantic knowledge.