Environment as Palimpsest: Layers of Buddhist Imagery on Kyŏngju Namsan during the Unified Silla (668–935 CE) Period

Abstract

This study unpacks the Buddhist assimilation of Namsan (South Mountain) in Kyŏngju through the creation and aggregation of Buddhist sculptures and structures on its slopes during the seventh to tenth centuries. Though steeped in native lore regarding nature deities and efficacious rocks, auspicious geological features such as Namsan were recast as part of a Buddhist landscape filled with manifestations of the Buddha and his attendants. These images served to demarcate claims of Buddhism’s place in the peninsula and were situated within sites that were previously marked and claimed by indigenous systems of belief. Employing an approach that draws parallels with David Harvey’s concept of urban environments as palimpsests, this paper reveals that Namsan was a multifaceted site, with military fortifications, temples, and rock-carved sculptures augmenting its spiritual and political significance. The repeated installation of Buddhist imagery ‘recovered’ the mountain, subsuming indigenous beliefs under Buddhist practices. This research finds that Namsan’s landscape was purposefully layered, reflecting the dialectical relationship between various communities and their religious and social practices over time. Analyzing Namsan as a palimpsest underscores the strategic appropriation of the mountain’s materiality and sacrality to establish a Buddhist territory deeply intertwined with the Silla elite’s politics and ideologies.

1. Introduction

Few places on the Korean peninsula are as abundant in medieval material history as the slopes of Namsan (南山 South Mountain) in present day North Kyŏngsang Province.1 Located in the middle of the Kyŏngju basin, just south of the ancient Silla capital, Namsan is a set of two smaller mountains that contain hundreds of artifacts that date primarily from the seventh to ninth centuries. Roughly forty valleys flow from Namsan’s two primary peaks, Kŭmo 金鰲 (elevation 468 m) and Kowi 高位 (elevation 494.6 m), within a geographical area measuring 8 km from north to south and 4 km from east to west (Figure 1). Their organically connected paths, which follow the run-off streams along the valleys, and myriad bluffs contain the most Buddhist images, structures, and temple remains of any one place on the Silla-Unified Silla period (traditional dates, 57 BCE–935 CE) site.2 As such, this mountain and its surrounding area have been the center of many scholarly studies ever since colonial Japanese archeologists surveyed and excavated Korea’s oldest monuments in the first two decades of the twentieth century.

The literature on the Buddhist relics of Namsan has primarily focused on individual sites or sculptures, offering sophisticated and nuanced interpretations of singular pieces, especially those with textual references (Mun 2003a, 2003b; Kim 2011, 2016; Ha 2018; Seo 2018). Artworks such as the Buddha triad discovered on Samhwa Ridge and the enigmatic sculptural program of Puch’ŏbawi (Buddha Boulder) at T’apkok 塔谷 (Pagoda Valley) have been extensively discussed by previous scholars, who have suggested possible ritual uses and highlighted the significant resonances that these images have with the theological currents of the time (Mun 1977; Ch’oe 2000; Kim 2016). Foundational as they are, such works of research necessarily forgo a holistic interpretation of the mountain and prioritize a static subset of the artworks that constituted Namsan’s visual and ideological repertoire. While benefiting greatly from such critical scholarship, this paper incorporates seldom studied stone sculptures into an analysis of Namsan that centers the mountain’s aggregated character and interprets the landscape as both being constitutive of and constituted by a process of ideological change.

That change is the adoption of Buddhism by the Silla state in the sixth century. Few shifts have affected the social and political dynamics of pre-modern Korea as much as the appropriation of Buddhist concepts, institutions, and culture by peninsular elites.3 This process affected all levels of society and fundamentally altered how people lived in and understood the world around them (Ahn 1989). Richard D. McBride II’s discussion of the doctrinal and cultic developments in Silla during this pivotal time provides an exceptionally thorough and nuanced investigation of the “domestication of Buddhism” on the Korean peninsula (McBride 2008). He outlines various ways in which Silla elites adapted the cults of Maitreya, Avalokiteśvara, and Hwaŏm Buddhism (Ch. Huayan) to suit their particular spiritual and political needs, including that of localizing the religion. Though McBride acknowledges that a complete picture of “‘Buddhism on the ground’ would also include a detailed discussion of art historical materials”, as a historian of Korean Buddhism, he limits his use of visual and material culture, relying primarily on a philological methodology that supports his exposition of the theological and scriptural adaptation of Buddhism in Silla (McBride 2008, p. 7). This study builds upon and complements McBride’s work by positing that one of the most visible ways in which this ideological change was reified was through the installation of Buddhist images throughout the landscape, especially at or around sites that were significant to long-held indigenous beliefs. By analyzing the Buddhist re-inscription of Namsan, a nexus of military, political, and spiritual power, this study leverages multiple avenues of entry into this very significant process.

To conceptualize the dialectical relationship between this process (becoming Buddhist) and the form that process took in the landscape (the palimpsest environment of Namsan), it is useful to consider the geologist David Harvey’s theorization of cities and urbanization. He differentiates between the thing (‘the city’) and the process (‘urbanization’) by positing that the former precipitates out of the latter, writing “we should think of things as products of processes” (Harvey 1997, p. 22). This concept is valuable to this study because the form of Namsan illuminates parts of the process that do not survive in the textual corpus. Yet Harvey complicates the simple binary by articulating that “things, once constituted, have the habit of affecting the very processes which constituted them” (Harvey 1997, p. 22). The city and its many forms influence the way that urbanization operates within its borders, creating a loop of influence where the past impresses upon the present and into the future. Indeed, Harvey observes that the various built environments, networks of social relationship, and administrative regions, etc. that become the various layers of a city appear and function in ways very specific to their particular social contexts. He states,

“What has gone before is important precisely because it is the locus of collective memory, of political identity, and of powerful symbolic meanings at the same time as it constitutes a bundle of resources constituting possibilities as well as barriers in the built environment for creative social change”.

(Harvey 1996, p. 50)

Seen in this light, the pre-Buddhist meanings of sacred rocks, mountainous landscapes, and Namsan are important to understand as they influenced the process of converting both the land and the people who lived there.

By adapting this model to interpret the palimpsest of Namsan and the historically specific processes that precipitated it, this paper takes a theoretically unique approach that differs considerably from other studies that look at Namsan as whole. Instead of attempting to review all the significant Buddhist structures and images on the mountain, the paper examines specific sites that illuminate certain aspects of the social processes that are at the core of this study. First, I provide an example of the layering process on Namsan by looking at the development of the Namsan New Fortress and the accretion of Buddhist images in the northern part of the mountain near Yunŭl Valley. This highlights the variegated symbolism and use of the mountain even before the Unified Silla period. Then, I examine the ways in which Silla kings adopted Buddhist tropes, forgoing native ideologies, to subsume a pre-Buddhist sacred landscape into a Buddhist one. In this section, I discuss the assimilation of pre-Buddhist sites into the Buddhist fold to suggest ways in which important locales in Silla were purposefully layered. I then turn to a sculptural group at a site called Ch’ilburam to illuminate the ways in which specific Buddhist doctrines were aggregated on Namsan over time. This leads to the major contribution of this paper, which is an in-depth visual and spatial analysis of the repeated creation of rock-carved images, stone sculptures, and temple sites throughout the seventh to ninth centuries in Samnŭng Valley. This section analyzes how these manifestations of Buddhist presence redefined and ‘re-covered’ the highly charged space of this particular landscape. Here, I use the term ‘re-cover’ in the sense of covering again, as well as of reclaiming something (in this case, the territory of the mountain). Lastly, I turn to a brief explanation of the continued mortuary use of this mountain before and after the adoption of Buddhism and how we may see these indigenous practices despite, and, in part, thanks to, the environmental palimpsest created by the process of conversion.

2. Multiple Uses of Namsan: Fortresses and Buddhist Developments around Yunŭl Valley

An example of the kind of multifaceted layering that occurred during the Silla period on Namsan can be seen in the northern third of the mountain around the remaining walls of the Namsan New Fortress (Namsan sinsŏng 南山新城), which was erected in 591 CE under the orders of King Chinp’yŏng 眞平 (r. 579–632).4 Gradually, over the course of three centuries, the sphere of the Silla court stretched out from the central fortified enclave of Wŏlsŏng 月城 (location in Figure 1), built during the fourth century, into the upper slopes of the mountain (Kang 2006). As an important measure of defense against attacks from the south by the Kaya confederacy (42–562 CE) and later rebels, it needed to be serviceable and well maintained. Extant inscriptions on a dedicatory stele dated to 591 speak to the great investment of human labor that went into constructing the new fortress by indicating that workers were brought in from all over the kingdom.5 Interestingly Namsan sinsŏng stele fragment no. 1 presents what is ostensibly a vow made to the king that the construction would not topple within three years, suggesting that the structural integrity of the fortress had been lacking in the past, requiring rebuilding within three years of its production.6 This fortification and its previous iterations attest to the military use of Namsan from the early stages of Silla history.

Currently, the entire perimeter of the Namsan New Fortress, identified through various surveys, is estimated to be around 4 km, and it snakes along the ridges of the northern tip of the Namsan area (Cho n.d.). Being so close to the capital, and on top of one of Silla’s most sacred mountains, it is surrounded by numerous temples and images. Directly to its northeast are some of the earliest Buddhist sculptures on Namsan: the seated figure in a niche on Pulgok (Buddha Valley) and the rock-carved images at T’apkok (Stupa Valley) are thought to date to the late seventh century (locations in Figure 1).7 Close to its northern wall, the Maitreya triad discovered on Samhwa Ridge (Figure 2) is widely considered to be the sculpture commissioned by the monk Saengŭi in 644.8 This triad, currently in the Kyŏngju National Museum, continued to be an important devotional image well into the eighth century.9 The close proximity of these images to the walls of the fortress point to the densely populated and multifarious usage of the mountain’s northern spaces, especially during the sixth, seventh, and eighth centuries.

During the early ninth century, another image was carved onto a rocky outcrop near the western side of the fortress walls. In Yunŭl Valley, the only dated rock carving on Namsan is of three seated Buddhas on two adjacent sides of an irregularly shaped boulder: a seated pair that faces south and a single seated Buddha facing west (Figure 3, location in Figure 1). An inscription found next to the central Buddha’s proper left shoulder reads “T’aehwa kunyŏn ŭlmyo” (太和九年乙卯), a date correlating to 835 CE.10 By placing these images so close to the wall of the fortress, Silla Buddhists of the ninth century were creating new layers of the landscape by responding to what was already there. They were marking and developing the areas of the mountain that were logistically and ideologically significant to them.11

In addition to the rock-carved imagery on the mountain’s slopes, the base of Yunŭl Valley was also being developed. Excavations at the mouth of the valley have unearthed a large temple site with stone Buddhist images and incised roof tiles that identify the structure as Ch’angnim-sa 昌林寺 (Kyerim munhwajae yŏn’guwŏn 2000, p. 60). Based on historical records, Ch’angnim-sa is thought to have been built in the eighth century, and then expanded in 855 under the auspices of King Munsŏng 文聖 (r. 839–857) with the addition of a three-story stone pagoda which still survives today (Ha 2021, p. 409). Recent scholarship by Ha Jungmin connects fragments of stone plates inscribed with the Lotus Sutra (妙法蓮華經) discovered in the environs around Kyŏngju as originally belonging to Ch’angnim-sa and having been created around the same time as King Munsŏng’s renovation of the temple (Ha 2021). Ha estimates that the entire sutra was carved on seven large plates of granite that were about 180 × 180 cm in size and that a single stone contained about 85 lines of text. Such a difficult undertaking speaks not only to the importance of the Lotus Sutra at the time but also to the belief in the merit-making power of copying, carving, and reciting sutras.

Similar practices have been well studied in the context of cave temples and outdoor rock carvings in China during the sixth and seventh centuries. Notable examples include the Leiyin Cave at Fangshan near Beijing, where some ten thousand stone slabs were inscribed with sutras (Lee 2010); the Xiangtangshan caves in Hebei Province, where sutras were inscribed on the walls of caves commissioned by Northern Qi rulers and elites (Tsiang 1996); and the sutra engravings made on cliffs and boulders on various mountains in Western Shandong Province produced during the late sixth and early seventh centuries (Wang and Ledderose 2014). These projects would have been known in the Silla capital through Korean intermediaries such as the Silla merchant general Chang Pogo 張保皐 (fl. late eighth to nineth century), whose activities centered around the autonomous residential sectors of Shandong called Silla pang 新羅坊, where many Silla immigrants lived (Xu 2016), and monks such as Chajang 慈藏 (d. between 650 and 655) and Ŭisang 義湘 (625–702), who returned to Silla after studying in China. Though the dearth of contemporaneous texts makes it difficult to prove that Silla Buddhists had seen the Chinese stone sutra projects, they certainly had a strong awareness of continental practices and Buddhist imagery. It is therefore likely that the inscribed Lotus Sutra from Ch’angnim-sa reflected, at least in part, similar devotional objects and rituals in China. In this way, the layering of images, spaces, structures, and meanings onto Namsan over the centuries was not simply a continuation of the types of activities that had animated the mountain before but the creation of a site imprinted with annotations that referenced various religious and social realities.

3. Adapting and Assimilating Native Mountains and Deities

Pre-modern sources indicate that the Silla elite had worshiped Namsan as a sacred place since at least the Three Kingdoms period (traditional dates, 57 BCE–668 CE) and that it continued to hold religious and political significance long after the acceptance of Buddhism in the sixth century.12 Traces of indigenous mountain cult practices can be seen in texts and through monuments on and around the topography of the mountain. According to the thirteenth-century Samguk yusa (三國遺事 Memorabilia of the Three Kingdoms), Namsan was one of the “Four Numinous Sites” sayŏngji 四靈地 that developed alongside the system of “Three Mountains and Five Sacred Peaks” samsan oak 三山五嶽, which was used in Silla governance.13 The believed efficacy of these mountains lent mystical symbolism to the Buddhist imagery that was then later layered onto their peaks and valleys. Unfortunately, this pre-Buddhist system of worship at mountains is not well understood, and even the identity of the central “Three Mountains” is still contested (Ch’oe 2013, pp. 249–50).14 Iryŏn 一然 (1206–1298), the monk compiler of the Samguk yusa, identifies the “Four Numinous Sites” as Ch’ŏngsongsan (east), Ojisan (south), P’ijŏngsan (west), and Kŭmgangsan (north).15 Ojisan 亐知山 is another appellation for Ojiam 亐知巖 which is a location on Namsan. It was there that heads of state gathered “at a giant rock” to make important decisions during the reign of Queen Chindŏk (r. 647–654).16 It was believed that the power of the four numinous sites ensured successful outcomes which were of the utmost importance, especially during the period right before the unification wars that led to Silla’s domination over the entire peninsula in 668.

Though the debate is still open about when specific mountain cults began to take on a Buddhist flavor, it is generally agreed that, by the late eighth century, monasteries and images had settled into many of Silla’s sacred locales (McBride 2008, p. 22). The Silla court incorporated indigenous deities, such as the dragon and the mountain spirit, into the Buddhist fold, and the places they were associated with were transformed into Buddhist places. Among the clearest episodes that communicates the Buddhist ‘recovery’ of local deities is the tale of the “Holy Mother of Mount Fairy Peach” in the Samguk yusa.17 The story revolves around the goddess of West Peak, a place that is also called Mt. Fairy Peach, possibly due to associations with earlier Daoist conceptions of the Chinese Queen Mother of the West and the peach of immortality (Lee 1993, p. 94). She is said to have appeared to a nun in a dream and told her that she could find a string of coins underneath the seat of her shrine. She directed the nun to use the money to commission images of Buddhas, deities, the gods of Heaven, and the gods of the five sacred mountains of Silla that were then to be enshrined in a monastery on her mountain. Furthermore, the nun was to hold a divination ceremony for all living beings on the tenth day of each month in the spring and autumn. The goddess’s beneficence was meant for the creation of merit for all beings so that they might enter paradise, a concern more closely associated with bodhisattvas. By superimposing the qualities of a bodhisattva onto an indigenous mountain spirit, Silla Buddhists appropriated this local deity and her mountain into the Buddhist realm.

Furthermore, many important Buddhist locales of the Unified Silla were already significant cultic sites that had supernatural beings inhabiting them. The legends surrounding the construction of Hwangnyong-sa 皇龍寺 (August Dragon Temple) and its two famous features—the nine-story wooden pagoda and the sixteen-foot Buddha statue—are productive to investigate because they touch upon many tropes of Buddhist place-making, including the creation of narratives about “traces” of past Buddhas (Robson 2009, p. 1359). The Samguk yusa relays that, in 553, King Chinhŭng 眞興 (r. 540–576) was planning to build a ‘Purple Palace’ on an auspicious site in Kyŏngju. But, after an “august dragon” appeared there, he erected a Buddhist temple instead and called it Hwangnyong-sa.18 In a separate tale we are told that, during his pilgrimage to Mt. Wutai in Shanxi Province, the aforementioned monk Chajang learned from a deity near Lake Tai that the dharma guardian dragon of Hwangnyong-sa was protecting the kingdom of Silla under the orders of Brahma (Whitfield 2012, p. 221; Lee 1993, p. 88). Here, the dragon is presented as a deity amenable to Buddhist belief. In a different section of the Samguk yusa, this spot is identified as “South of the Dragon Palace” and is one of the seven locations in Silla where vestiges of past Buddhas could be found.19 Digging deeper into the site’s pre-Buddhist associations, McBride suggests that this may have been the center of an indigenous royal dragon cult where rituals for rain were enacted (McBride 2008, pp. 23–24).

The Hwangnyong-sa foundation legend clearly indicates that Korean Buddhists were drawing upon strategies of localizing Buddhism from their Chinese counterparts. On the same trip to Mt. Wutai, Chajang was also told by Mañjuśrī, the bodhisattva of wisdom, that Hwangnyong-sa was a divine place where the former Buddha, Kaśyapa, had meditated.20 The bodhisattva explained that it was a vestige of a monastery dating from the time of the past Buddha and that his meditation stone was still there. In the “discovery” of this trace of the past Buddha, Silla elites were able to traverse the physical and temporal distance between Indian Buddhism and their own lived reality. This rock, on which Kaśyapa became enlightened many eons before, was called the “Meditation seat stone” yŏnjwa-sŏk 宴坐石, and Iryŏn states that he was able to see it in its conserved state when he visited the temple. By verifying the existence of this artifact through his own experience of it, Iryŏn offers his readers mediated physical proof of a karmic connection between the Korean peninsula and the past Buddhas. Seen from a different perspective, Silla and Koryŏ Buddhists’ rhetorical mapping functioned to retroactively ‘reclaim’ the peninsula as Buddhist territory.

Another aspect of the Hwangnyong-sa legend that points to a preoccupation with recovering Silla’s Buddhist past is the story of the temple’s sixteen-foot Buddha image. In this anecdote, an unmanned boat is discovered floating towards the eastern coast of Silla. It contained a message from King Aśoka, who, having failed three times to cast a Buddha triad, loaded the materials and statue models onto a ship, praying that it would reach a country with a karmic link (yuyŏn 有緣) where a “sixteen-foot image of the honored visage” (changyuk chonyong 丈六尊容) could be made.21 The boat had drifted for 1300 years before reaching Silla. In 574, the metals were transferred to Muning Forest, where Silla artisans were able to successfully cast a sixteen-foot Buddha on the first try. The triad was enshrined in a Buddha Hall of Hwangnyong-sa under the mandate of King Chinhŭng. One year later it began shedding tears—an omen of the king’s impending death.22

This account highlights the understanding that the Silla kingdom was karmically linked to the empire of Aśoka, the archetypal Buddhist land led by the paradigmatic cakravartin king. Given his fame, Aśoka and his acts of piety were widely emulated by Buddhist rulers throughout East Asia. Large sets of pagodas were created to mimic the 84,000 stupas to which Aśoka was said to have distributed relics throughout Jambudvīpa in one day (Shen 2012).23 Images found in Sichuan that identify themselves as “(A)yuwang xiang” (阿)育王像 (Aśoka image) also evince the efficacy afforded to the Aśokan type.24 Unlike Chinese narratives, where images and relics made and dispersed by Aśoka or his daughters are discovered by the faithful, in the Hwangnyong-sa story, emphasis is placed on the skill of the recipients and the predestined connection between the places of Buddhist pasts and presents.25 The ancient materials and instructions crossed time and space to find receptive and skilled practitioners who could accomplish a tremendous task—one that not even Aśoka was able to achieve.

The sculpture’s association with the legendary Mauryan emperor worked to emphasize its efficacious nature and link the patronage of Silla rulers to the projects of ancient sages.26 These ideological strategies, developed by members of the court since King Chinhŭng’s reign but effected in earnest around the late seventh and early eighth century, were used to promote the idea that Silla was a land ruled by cakravartins and karmically connected to the Buddhas of past eons (McBride 2008). A particularly influential reading of the Silla adoption of Buddhism was first articulated by Rhi Ki-yŏng in an article that identified Silla as a “Buddha Land” (pulgukt’o 佛國土) (Rhi 1974). This concept has been discussed by many scholars and, though fraught with revisionist and anti-colonial language, is an expedient framework within which we can analyze the rapprochement between indigenous beliefs and Buddhism and better understand the rationale behind the environmental palimpsest of Namsan (Kim 1974; Hwang 1975; McBride 2008).27

The textual foundation for Rhi’s “Buddha Land” ideology is the story of the Koguryŏ monk Ado, who brought Buddhism to Silla in the third century at great personal peril. Codified in a stele inscription (Ado ponbi 阿道本碑) preserved only in the Samguk yusa, Iryŏn relays that Ado was told by his mother that there were seven places in the land of a sage king born in Kyerim (another name for Silla) where remnants of past Buddhas could be found.28 He was encouraged to go there and set up temples, as they were predestined to be monastery sites from the previous Buddha’s time.29 Included among these sites is Ch’ŏn’gyŏng (Heavenly Mirror) Forest, the site of Hŭngnyŭn Temple; Sinyu (Spirit Wandering) Forest, the site of Sach’ŏnwang Temple; and, as discussed above, south of Dragon Palace, the site of Hwangnynong Temple. This excerpt demonstrates the territorial conversion of indigenous sites into Buddhist places through narrative means by Buddhists invested in reclaiming sacred landscapes.

This textual account is buttressed by archeological evidence that suggests that Hwangnyong-sa was constructed using repurposed Bronze Age (ca. 1st millennium BCE) dolmens.30 Scholar Han Joung-ho posits that, based on their size and shape, the stone foundations of the main Buddha icon and the right bodhisattva were actually two halves of the same stone (Han 2014). A reconstruction of the stone shows that it is of the same size and shape as other dolmens used to mark Bronze Age tombs found in the vicinity of Hwangnyong-sa and the nearby Bunhwang Temple. Furthermore, Han proposes that, given the dimensions of Kasyapa’s meditation stone, as recorded in the Samguk yusa, it is highly likely that a type of standing dolmen (sŏndol; ipsŏk 立石) was similarly converted to become that contact relic of the past Buddha (Han 2014, p. 55). Excavations around Nangsan, another sacred mountain one mile northeast of Namsan, have discovered many megaliths that are thought to have been used to demarcate the boundaries of ritual spaces, settlements, and agricultural fields of the prehistoric era (Gim et al. 2019). It can be concluded that the area where Hwangnyong-sa was constructed had been a significant ritual site since before the introduction of Buddhism to the peninsula. The ancient dolmens that were used to identify native sacred spaces were appropriated as the meditation stone of Kasyapa and the foundation stones upon which the sixteen-foot Buddha and the wooden pagoda of Hwangyong-sa were constructed.

Building on an already strong native belief in the efficacy of rocks, Buddhist practitioners incorporated the medium into their territorial conversion. The absorption of rock worship into Buddhism can be seen in an episode from the Samguk yusa where a samyŏn sŏkpul 四面石佛 (four-sided stone with Buddha images) was discovered beneath the ground on the side of a mountain. While on his way to Paengnyul Temple 佰栗寺, King Kyŏngdŏk (r. 742–65) heard a voice chanting the name of the Buddha from the earth and had his men dig at that site, from whence they found a “great rock with Buddha images carved on it facing in the four directions”.31 Kyŏngdŏk ordered a temple to be constructed on the auspicious site and named it Kulbul Temple 掘佛寺 (lit. Buddha-digging Temple), but Iryŏn records that, by his time, it had already fallen into ruins, leaving only the stone monument by the roadside.32 In the same passage, Iryŏn also notes that, during the reign of King Chinp’yŏng, a huge rock with the Buddhas of the Four Directions carved on all four sides suddenly appeared after falling from the top of a mountain. The king had a temple constructed next to the rock and charged a monk to keep the miraculous boulder clean and attend to it by burning incense continually.33 According to the text, the mountain was called Sabulsan 四佛山 (Four Buddha Mountain). The fact that the Buddhist images were already in the earth, awaiting a sage king, functioned not only to reinforce the karmic merit and symbolic power of the crown but also to legitimize the creation of Buddhist institutions on Silla’s sacred mountains.

The layered conversion of the Silla landscape into a Buddhist one, whether it is specifically called pulgukt’o ideology or not, is apparent in the textual and archeological evidence. Activated by the karmic connections that link beings across kalpas, strata of Buddhism were rhetorically constructed and then physically reified on the peninsula through a process of intentional reuse and aggregation. Seen in this way, mountains were imprinted with layers of time, covered again and again with human activity, like a palimpsest of previous ages of the dharma. It is for this reason that the discovery of rocks with images of the Buddha—the materiality of which was already thought to be efficacious—was possible. Instead of simply adopting a foreign religion, Silla was recast as originally having been a Buddhist land, with evidence of its past karmic links being buried deep within the ground. The mighty boulders and stones that comprised the rocky landscape of the mountains surrounding Kyŏngju were thus invested with the power of past Buddhas. This practice of ‘re-covering’ a landscape is nowhere more apparent than in the case of Namsan.

4. The ‘Re-Covery’ of Namsan: Accretions of Stone Buddhas

Both the Samguk yusa and the Samguk sagi (三國史記 History of the Three Kingdoms) speak of Namsan as a numinous and strategically important site. It was where major fortresses and military storehouses were constructed through government-enforced organized labor, and it was also the believed abode of a mountain deity, Śākyamuni Buddha, and miraculous images of Maitreya. Throughout the Silla and Unified Silla periods, kings, members of the royal family, and high-ranking officials would visit Namsan and its many auspicious sites and temples. Some were buried on or near the mountain, and many royal tombs remain along its borders today. Monks and devotees would go to Namsan to venerate images of Buddhas and bodhisattvas, performing ceremonies that included offerings of tea and native song.34 Namsan was a conspicuous site for indigenous and Buddhist devotion and ritual and a composite space of layers of different times.

Large or strangely shaped rocks were local points of power, first on a supramundane level, where miraculous occurrences happened; then on a social level, where groups of authority figures met; and then on a religious level, where spirits were said to reside. Belief in the power of rocks to effect positive outcomes or house deities were appropriated in combination with the numinous spatial context of mountain worship to create a unique type of Buddhist space that incorporated indigenous practices into Buddhist space-making ideologies. More specifically, Buddhas that appear in the guise of rocks on the liminal space of the mountain gave credence to the idea that Buddhism was native to Silla and that the animistic local beliefs in rocks and mountains were a precursor for the discovery of this understanding by the proper (i.e., Buddhist) rulers. This process of conversion precipitated the repeated inscription of Namsan with layers of symbolism and ritual.

4.1. Building on the Past: Ch’ilburam Buddha Group and the Sinsŏnam Bodhisattva

Of the many rock-carved images on Namsan, one of the best known is the rock-carved Buddha group at Ch’ilburam 七佛庵 (Figure 4 and Figure 5, location in Figure 1) (Kim 1974, p. 277).35 This is a unique set of seven Buddhist deities carved in bas-relief on two boulders. The larger boulder is a flattened semicircle featuring a Buddha triad. The other is a rectangular pilar with four seated Buddhas, one on each side. Scholars have dated this statue group to the first half of the eighth century based on stylistic comparisons with other dated sculptures from the same time period.36 Though this figural group has received much attention in the past, it is especially germane to this paper because it exhibits the layered features of different social and historical contexts and the practices that were significant at those times at one site (Park 2012; Ha 2018; Seo 2018).

At Ch’ilburam, the Buddha triad carved on the flattened side of the larger boulder features a Śākyamuni Buddha flanked by two smaller bodhisattvas. The bodhisattva to the Buddha’s proper right holds a bottle, an indicator that it is Avalokiteśvara. The attendant to his left holds the stem of a flower and may be Maitreya (Pak and Whitfield 2003, p. 176). The main figure identifies itself as Śākyamuni through his bhumisparsha, or earth-touching, mudra. This particular gesture references the moment in Śākyamuni’s life where he is challenged by the demon Mara and calls upon the earth to bear witness to the merit that he has earned throughout his countless lives in order to attain enlightenment.

The earth-touching mudra appears in Tang dynasty sculptures in the late seventh century and appears in Korean imagery during the early eighth century. The main Buddha wears an Indian style robe with one shoulder bare. The way the drapery pools in front of the Buddha’s crossed legs relate it stylistically to contemporaneous and slightly earlier Tang sculptures and Indian images of this bhumisparsha Buddha type. Relief sculptures from Baoqingsi 宝慶寺 in Xi’an, some of which have inscriptions dating between 703 and 724, show notable similarities with the main Buddha triad at Ch’ilburam.37 One such example, now in the Nara National Museum, depicts a bhumisparsha Buddha figure dressed in a one-shoulder Indian style robe whose folds are articulated in much the same way. The cascading fan of fabric that pools in front of their legs further points to their shared late Gupta–early Pala period referent. The two bodhisattvas flanking the Buddha in both triads stand on lotus pedestals and wear flowing robes with a sash draped from their right shoulders across their chest and wrapped around the left side of their waists. In both Chinese and Korean cases, the bodhisattvas hold their attributes with one hand and stand in a slight tribangha pose. Both sets of attendant figures have the same undecorated, bas-relief halo, and the overall composition of the three figures together, carved within the half-dome frame of the large rock, evinces an affiliation with early Tang dynasty steles found in Longmen.38 This similarity underscores the textual records that attest to the active communication between Tang and Silla during the seventh and eight centuries and the reception of certain Indian styles through Chinese and Korean intermediaries (Lopez 2017).39

One of the most famous instances of monks bringing Indian texts and images into China is the Chinese monk Xuanzang 玄奘 (602–664), best known for having traveled to India and recording his journey in his Record of the Western Regions compiled in the Great Tang (Da Tang xiyouji 大唐西域記). He writes here of an auspicious image of a seated Buddha enshrined in the famed Mahabodhi Temple in Bodhgaya, first built by Aśoka in the third century BCE to mark the place where the historical Buddha achieved enlightenment, making the same earth-touching gesture with his right hand. Another Tang monk Yijing 義淨 (635–713) was said to have made the pilgrimage to Bodhgaya and brought back a statue of a seated Buddha displaying the earth-touching mudra. Given that, along with Xuanzang and Yijing, other monks and envoys made the journey to India in the seventh century, it is probable that the seated bhumisparsha Buddha at Ch’ilburam was of a new type that was brought into the Chinese iconographic repertoire at least by the early Tang and then transported over to Korea by Silla monks shortly thereafter. As mentioned above, Korean monastics such as Ŭisang traveled between the two countries, transporting texts and images with them as they moved.

Another link to Buddhist trends and practices on the continent can be found in the fragments of stone40 sutras excavated near the Ch’ilburam site (Figure 6).41 In total, five pieces of sutra texts have been found and identified as being excerpts from the Diamond Perfection of Wisdom Sutra (Jin’gang bore boluomi jing 金剛般若波羅蜜經) translated by Kumarajiva 鳩摩羅什 (344–413) and the Sutra of the Vows of the Medicine Buddha (Yaoshi liuliguang rulai benyuan gongde jing 藥師琉璃光如來本願功德經) translated by Xuanzang.42 Mention should be made here of the material and textual overlaps between these carved sutras at Namsan and the slightly earlier carvings of sutras on stone found in the mountains of western Shandong Province created during the Northern Qi (550–577) or Sui (581–618) dynasties (Wang and Ledderose 2014). Of note is a giant 2064 square meter carving of the first fifteen sections of the Diamond Sutra that runs across a granite streambed, now called Sutra Stone Valley on Mount Tai. This lithic text is not only the largest of all the carved sutras in Shandong but also the least understood, as no accompanying dedication or historical text records its date or the identity of its donors.43 Though the exact patronage of Sutra Stone Valley awaits further study, Claudia Wenzel (2016, p. 62) has proposed to see this project as an exercise in “establishing a Buddhist site amidst the sacred geography of Mount Tai that was already layered with Daoist gods and imperial rites”. Referencing Mt. Tai’s long history as the site of imperial feng and shan sacrifices, Wenzel reads the monumental project as a response to previous practices of inscribing the landscape of this sacred mountain. Though the scale of the Namsan epigraphy is nowhere close to that of the characters in Sutra Stone Valley,44 they are analogous in their stone materiality, their choice of scriptural focus, and their motivation to forge a Buddhist space within an already spiritually charged landscape.

Directly in front of the east-facing triad is a four-sided stone that features bas-relief images of individual Buddhas on each of its four sides (Figure 5). Commonly described as a sculpture of the Buddhas of the Four Directions, this image is perplexing in its iconography as well as its positioning. The four figures are each seated cross-legged on a lotus flower and are dressed in similar double shouldered dhotis which ripple down and across their bodies in steady parallel lines. Their mudras and occasional accouterments are the only differentiating features about them. They all have undecorated pointed halos and measure between 83 and 118 cm in height. Despite the larger scale of the Śākyamuni triad, scholars have suggested that the four-sided Buddha stone was in fact the main icon of this site (Seo 2018, p. 70). A Koryŏ period rooftile with the inscription “四□寺” found near the site bolsters this theory of the centrality of the four-sided Buddha stone carving at Ch’ilburam.

The basic composition of a four-sided stone or column with images of Buddhas on all four sides has its most obvious precedents in the pagoda-shaped shrines and four-sided steles from China during the Northern Qi to Tang dynasties (Wong 2004, pp. 65–70). The idea of the Buddhist stele and its associated designs were familiar to the Koreans by the seventh century, as exemplified by inscribed steles excavated from temple sites around Sejong in South Ch’ungch’ŏng Province (Jang et al. 2014, pp. 40–46). Scholars are mostly in agreement about the identities of two of the four Buddhas: Amitabha is carved on the west side and Bhaiṣajyaguru (Medicine Buddha) is on the east (Figure 5) (Kim 2017, pp. 28–29). Unfortunately, identification of the other two figures is less clear. The iconographic program of the Buddhas of the Four Directions is wildly inconsistent, even within the same time period and place. Although lists of Buddhas associated with the four cardinal directions appear in multiple texts dating from the early fifth to the tenth centuries, depictions of these Buddhas are often at odds with the textual prescriptions (Ho 1968–1969, figures 10, 15, 37).

Nevertheless, given that the icon directly facing devotees is Bhaiṣajyaguru and the fact that the entire site is oriented towards the east, the direction of the pure emerald world where this deity is thought to reside, Ha Jungmin has posited that the Medicine Buddha should be understood as the most important of the seven-deity group. The discovery of fragments of a stone carving of the Sutra of the Vows of the Medicine Buddha supports this reading (Ha 2018, pp. 318–19). Ha further points to the numerous holes and divots found on the rear of the large triad boulder and along the top of the four-Buddha stone to suggest that there was most likely a wooden temple structure at Ch’ilburam that would have enclosed the two stone sculptures as well as the stone sutras (Ha 2018, pp. 338–43). Based on prescriptions on how to worship the titular deity in the Sutra of the Vows of the Medicine Buddha, Ha proposes that this sutra carving, along with the aforementioned carving of the Diamond Sutra, were installed on the north and south walls of the wooden structure and that practitioners would circumambulate the four-Buddha stone while chanting the sutra to have their wishes fulfilled (Ha 2018, pp. 336–37). Correspondingly, Seo Ji-min has proposed that this site be understood in the context of the cult of Bhaiṣajyaguru based on the fact that, during the first quarter of the eighth century, during the reign of King Sŏngdŏk, the Samguk sagi records no less than 21 natural disasters that caused sickness and death all over the kingdom (Seo 2018, pp. 71–73). Thus, a temple on an efficacious mountain dedicated to the worship of the Medicine Buddha during the first half of the eighth century correlates with the social and political needs of the time.

Not more than 50 m further up the mountain path towards Kowi peak is a bas-relief carving of a seated bodhisattva on the rock cliff of Sinsŏn-am 新仙庵 (Figure 7, location in Figure 1) The figure’s left leg is bent towards his right knee in an almost half pendant posture, and his right foot is resting on a lotus flower. The deity is seated on a high dais, partially covered by its robes, that is supported by a billow of clouds. His full and smooth face, as well as the curvilinear delineation of the eyes, eyebrows and mouth, recall the skilled technique of the main Buddha image at Sŏkkuram grotto in Mt. Toham, Kyŏngju (National Treasure 24). However, its squat proportions, the absence of negative space between the limbs and torso, and the elaborate pedestal date it to the late eighth or early ninth century. Stylistically, this image seems to have been carved later than the double stone monument only a few minutes’ walk away.

Rhi Juhyung identifies this bodhisattva as Maitreya based on the nagapuṣpa (dragon flower) twig held in his right hand (Rhi 2013, p. 111). Iconographic features, such as the tall crown with a flat band around the bottom, the scarf around the shoulders, ribbons that crisscross over the body and the knees, and the elegant jewelry, are echoed in other images of Maitreya, such as the standing Maitreya bodhisattva stone statue from Kamsan-sa (National Treasure 81), dated to 719, and the pensive bodhisattva (National Treasure 78) at the National Museum of Korea, from the late sixth to early seventh century. Though there are significant differences between the rock-carved image and the seated bodhisattva—the opposite legs are lifted, the hands are engaged in different activities, the introspection of the gilt bronze statue is gone in the stone sculpture—the Sinsŏn-am bodhisattva’s pensive position, the draping of the seat with excess cloth, and the lotus flower supporting the lowered foot suggest that its artisans were at least aware of the pensive bodhisattva type when they created the rock carving. Interestingly, this sculpture has attributes that were important for Buddhists of the late eighth to early ninth centuries. For example, the clouds upon which he sits and the nagapuṣpa he holds indicate that this depiction is of Maitreya in Tushita heaven. Going beyond the image of the deity as a youthful princely figure, this sculpture evinces an understanding of the importance of Maitreya’s pure land that gained popularity after the exegetical work performed by the monk Wonhyo 元曉 (617–686) in the second half of the seventh century (Choe 2015).

Drawing upon the cult of a different deity, the Sinsŏn-am bodhisattva evinces the layering of the worship of Maitreya literally on top of the valley where a Bhaiṣajyaguru temple had been built. In order to reach Sinsŏn-am, one must pass by Ch’ilburam and climb a few minutes up to the rocky ledge upon which the Maitreya image has been carved. The bodhisattva is hewn from a coarse granite cliff that opens up to a south-facing view of lesser mountain peaks within the Namsan area. To the left of the carving are the remains of a small stone lantern, indicating that this image had at least a small temple nearby. In front of the carving is a narrow 152 cm ledge with a precipitous drop. Though the ledge continues further past the carving, this image cannot be circumambulated like the four Buddha stone at Ch’ilburam. Thus, in addition to a different devotional focus, this later accretion of Buddhist imagery also animates the landscape in a different way. Given the restricted layout of the worship space in front of the image, it is possible that a limited set of practitioners supplicated the bodhisattva with offerings of tea and song.45

The four Buddha stone pillar and Buddha triad at Ch’ilburam were established first within the slopes of Bonghwa Valley. Yet, within the century, another Buddhist site was placed on top of the valley, leading the devotee up towards one of the central peaks of Namsan. The two sites demonstrate the ongoing augmentation of the heights and vales of the sacred mountain. That the later image is further up the mountain path is paralleled in the installation practices of ritual images that generated the Buddhist space of Samnŭng Valley.

4.2. Samnŭng Valley: A Layered Path to the Top of the Mountain

Samnŭng Valley 三陵溪 (“Three Tomb Valley”), located on the western side of Namsan, contains more Buddha images than any other, with nine sites of Buddhist imagery and sixteen figures in total (inset of Figure 1).46 The aggregated landscape of this valley evinces the continuous infusion of Buddhist presence on a particular part of the mountain over time by different hands. The valley’s name is derived from the three royal tombs, lined up in a neat row, at the base of the mountain where the valley starts. These three tombs are said to be the posthumous abodes of the eighth monarch of Silla, Adalla Isagŭm (154–184 CE), the fifty-third monarch of Unified Silla, King Sindŏk (912–917), and the fifty-fourth monarch, King Kyŏngmyŏng (917–924). These identifications have been applied to the tombs because of where the Samguk sagi states these three rulers were interred.47 Scholars have cast doubt on the plausibility of these identifications as the gap between the first burial of Adalla and the later burials of Sindŏk and Kyŏngmyŏng are more than 700 years apart (Kang 2015, p. 181). It should be noted that the tumuli should be considered one of the last additions to the overall program of the valley, and thus influenced the experience of Samnŭng for those from the tenth century onwards exclusively.48

The remains of ten temple sites have been identified in Samnŭng Valley, six of which have stone statues or rock-carved imagery within their vicinities.49 They have been identified as Samnŭng Valley temple site number 1, 2, 3 and so on by the Kyŏngju National Research Institute of Cultural Properties (hereafter the KNRICP) and are labeled in consecutive order, with temple site 1 being the closest to the base of the mountain and temple site 9 being closest to the peak of the mountain along the valley (Kyŏngju munhwajae yŏn’guso 2000). A total of twelve Buddhist images have been discovered near sites 1, 2, 3, 5, 6 and 9; five of them are free-standing sculptures and seven are rock-carved images. Nine of these images are stylistically dated to the eighth and ninth centuries. Of the remaining three, two have been identified as Koryŏ period images from the tenth century and one is an image from the modern period. Along the valley, at sites 1, 3, 6 and 7, the remains of stone pagodas have also been found. The best-preserved specimen, from site 6 and now in the Kyŏngju National Museum, has been stylistically dated to the ninth century.

Table 1. Samnŭng Valley stone sculptures in order.

Valley and Temple Site	Image Name	Date Range	Figure #
Samnŭng, site 1	Stone Seated Medicine Buddha	Modern
	Stone Standing Buddha	8th cent
Samnŭng, site 2	Stone Seated Buddha	Late 8th cent/early 9th cent	17
	Rock-carved Standing Avalokiteśvara	Late 8th cent/early 9th cent	10
Samnŭng, site 3	Stone-incised Standing Buddha Triad and Seated Buddha Triad	Late 8th cent/early 9th cent	11
Samnŭng, site 5	Rock-carved Seated Buddha	10th/13th cent
Samnŭng, site 6	Stone Seated Buddha	Late 8th cent	9
	Stone-Incised Buddha	9th–10th cent	14
	Stone Seated Medicine Buddha	Early 9th cent	8

Data organized from Kyŏngju National Research Institute of Cultural Properties (Kyŏngju munhwajae yŏn’guso 2000, Kyŏngju Namsan).

lists the images along with their temple site, current location, and proposed dating as recorded by the KNRICP. The range of dates conveys the varying scholarly opinions about specific sculptures. However, we can be fairly certain that, by the tenth century, most of the images and temples listed were installed in Samnŭng Valley, or had left their mark on the valley, and that visitors making their way up to Kŭmo Peak would have seen and stopped at these places. Tracing the accumulation of these Buddhist markings on the landscape, we can see that practitioners at different points in time were responding uniquely to the built environment in a process that was colored by the religious disposition of their historical context. By choosing to repeatedly carve on this valley, the symbolic meaning of Samnŭng was reaffirmed and re-covered. Taken as a whole, this valley presents the great variety of approaches to stone sculpture, devotional imagery, and ritual practice, as well as alternative methods of imprinting Buddhist imagery onto the landscape that characterized the process of Buddhist conversion over this long period.

Notable for its unique combination of iconography is the seated Medicine Buddha discovered at site 6 and currently housed in the National Museum of Korea (Figure 8). This statue is often touted as a representative example of Unified Silla stone sculptures from the early ninth century because it evinces the same earth-touching mudra and sculptural sophistication of the mid-eighth century Sŏkkuram Buddha but lacks its elegant body proportions and is joined to an emphatically elaborate halo and pedestal. The shoulders and knees of the Samnŭng Valley Bhaiṣajyaguru are much more compact than the regal Sŏkkuram icon and, overall, the sculpture corresponds more closely with the stone Vairocana Buddha at Piro-am of Tonghwa Temple in Taegu dated to 863. Though identified as a Bhaiṣajyaguru because it holds a round object in its proper left hand, it also presents the bhumisparsha mudra of the historical Buddha. This hybrid formulation was not uncommon in the late eighth and into the ninth century, and at least eight other images on Namsan alone present this same amalgam of mudras. Lena Kim posits that this “iconographic confusion” could be attributed to sculptors’ misunderstanding and to a tendency to disregard strict iconographic rules (Kim 1998, p. 273). However, it is also plausible that the bhumisparsha composition became popular after its importation in the seventh century and was adopted for other Buddha images as well. This is revealed in a number of stone and iron medicine Buddhas exhibiting the earth-touching mudra from the Unified Silla and Koryŏ (918–1392) periods.50 Seen as an emblem of enlightenment, the intrinsic state of Buddhahood, this synthetic tendency suggests a far more inclusive understanding of Buddhist doctrine than normally afforded to artisans.51

Further augmenting the layered quality of this site, the bhumisparsha mudra can be found on another free-standing stone sculpture discovered only 40 m away downhill. The seated Buddha of Samnŭng Valley temple site 6 (Figure 9), which is still in situ on Namsan, is less intricately carved and has suffered more damage than the Bhaiṣajyaguru sculpture but still evinces a sculptural skill that harkens to the Sŏkkuram Buddha. Indeed, this figure, with its attenuated waist, broad back, and thin one-shouldered dhoti, suggests that its makers were closer in time and technique to those of the Sŏkkuram grotto. Though the two Samnŭng Valley sculptures were discovered on the same temple site, the KNRICP’s studies have shown that the seated Buddha was most likely placed outdoors whereas the Bhaiṣajyaguru was placed within a structure built into a small space next to a cliff (Kyŏngju munhwajae yŏn’guso 2000, pp. 248–51). This corresponds to the differing amount of corrosion found on both images and at the very least suggests a difference in the social processes that went into their creation. That is to say, the two groups that produced the images had differing understandings of where Buddhist images should be placed. Remains of a stone lantern and stone lotus flower podium, as well as the aforementioned ninth-century stone pagoda, were also found at temple site 6, indicating that there was a well-developed monastic establishment at this location and that they were integral parts of a temple’s visual program (Kyŏngju munhwajae yŏn’guso 2000, p. 248). If we accept that the Medicine Buddha and its enclosure were later additions to the temple site, we can deduce that the upper reaches of the valley were built up later than the lower and that Samnŭng temple site 6 enjoyed continuous patronage during the eighth and ninth centuries.

This pattern of expansion aligns with the general sequence of images embedded into the valley. The Koryŏ period rock carvings are both further up or higher into the slopes of the mountain than the two seated stone Buddhas. This is not to overstate the temporal range that is visible in this valley. Most of the sculptures and rock carvings date to the century between 750 and 850 CE. What is notable in Samnŭng, however, is the heterogenous sculptural methods implicating the various hands that densely layered the landscape here.

Unlike the previously mentioned free-standing sculptures, the rock-carved standing bodhisattva at temple site 2 is a high-relief carving of a 152 cm tall deity crafted from a rocky crag that juts out from the earth a few feet away from the valley path (Figure 10). Hewn from the surface of a boulder, the image protrudes halfway out from the rock at around 15 cm. The effect of this type of carving technique is that the deity seems to be emerging out of the granite matrix from which it is made. Though highly eroded, one can still make out a bottle in the deity’s left hand, the traces of a small Buddha depicted in his crown, and some of the incised scarves that cover his body. Thus, this figure has been identified as an Avalokiteśvara. Besides the small lotus flower pedestal that is roughly carved under the bodhisattva’s feet, there is no evidence that an effort was made to situate the figure within a structure (Kyŏngju munhwajae yŏn’guso 2000, p. 251). However, the image is backed by a large columnar boulder that seems to announce the presence of this deity to travelers making their way up the mountain along the path. This suggests that artisans took advantage of the natural context to ensconce the sculpture within the landscape. Combined with the emanating visual effect of the bas-relief treatment, this image illustrates a way of making and presenting Buddhist deities that is different from what has been discussed thus far.

This representation of Avalokiteśvara is, in fact, one of the first images travelers would see along the Samnŭng Valley. A few minutes’ walk up the mountain, they would then encounter the second layer of Buddhist imagery embedded into the landscape on a large, black, living rock. This site, labeled Samnŭng temple site 3, features a set of two triads—one of a standing Buddha flanked by two kneeling attendants, and the other of a seated Buddha flanked by two standing attendants (Figure 11 and Figure 12). Incised as shallow line drawings on a natural expanse of rock, this set of figures evinces yet another method of Buddhist image-making in the valley. The six deities are not given any kind of three-dimensionality and are inscribed onto the mountain in the same manner that a textual inscription would be made on stone. No effort has been made to create volume through shadowing or the natural undulation of the rock. The representational effect here is much the same as that of printed images or line drawings. Thus, the carving techniques used to create these two triads were very different from those used to create the free-standing Buddha sculptures or even the high-relief bodhisattva.

In fabrication, form, and scale, the double triad carving in site 3 recalls the rock-carved Medicine Buddha Triad on Pangŏsan in Haman, South Kyŏngsang Province about 100 km southwest of Namsan (Figure 13). Craftsmen at both sites employed the same shallow incised line technique on an unworked, very flat, very dark living rock. Both places feature a standing Buddha and standing bodhisattvas who share visual similarities in the way their ears and ushnisha are depicted, the way their feet are splayed out in an unnatural manner on top of their lotus pedestals, and the way thin strips representing scarves flow over and around the bodhisattva’s laps. The Buddha figures at both sites measure around 3 m in height and are placed close to the ground. Luckily, the Pangŏsan triad has an inscription that dates it to 801. Given the parallels, the Samnŭng Valley double triad rock carving can be reasonably dated to the eighth or early nineth century as well.

The shallowness and simplicity of the Samnŭng carving makes the figures difficult to see, especially if they are shielded underneath a wooden enclosure. A long rivet that runs along the top of the living rock and the presence of peg holes suggests that there was an edifice installed here at some point or that a rain drainage system was put in place to prevent erosion.52 Since no evidence of pigments or gilding remain on these exposed images, it is difficult to prove, but Youngsook Pak maintains that stone sculptures of the Silla period were undoubtedly painted and decorated with gold leaf, citing extant examples from Chinese cave temples at Yun’gang and Qingzhou (Pak and Whitfield 2003, p. 39).53 If these images were painted, then they would be much more visible underneath a structure, and the strangely absent eyes of the seated Buddha to the right may have been painted in. The unmodulated, straightforward lines that remain on the rock would then also make sense, as they form the basic outline of a much grander visual program.

That these images may have been enshrined, and, more so, that they may have been painted to stand out from the landscape helps explain the remoteness of another rock-carved image of the Buddha, this time in temple site 6. Not far from where the seated Bhaiṣajyaguru was discovered, a very rudimentary and probably incomplete image of a Buddha can be found high up on a cliff, facing south and overlooking the valley path (Figure 14). The carving was carried out with the same incised technique as the double triad set, but only the head, neck, and shoulders have been carved.54 It is remote because, unlike the other images layered onto Samnŭng Valley that have been discussed, this one was carved onto the upper reaches of a cliff where worshipers would not be able to pray directly in front of it. This necessitates a different type of interaction between the deity, the devotee, and the landscape. Instead of the direct and localized surroundings of a boulder or a platform, the vast surrounding landscape is invariably implicated in the context of the icon being venerated from afar. The image also underscores the ubiquity of the Buddhist presence in the valley: manifestations of Buddhas are incised, carved, and shaped onto all levels of mountainous space.

However, the most arresting example of a Buddhist image that anticipates its distant visibility is the giant rock-carved Buddha from temple site 9, near Sangsŏn Hermitage 上禪庵 (Figure 15). This sculpture is 5.21 m tall, carved on a rocky bluff, and overlooks a steep valley. As a visitor makes their way along the ridge of Samnŭng towards Kŭmo Peak, they would come to a grand vista of the deity situated within his mountainous environment (Figure 16). This initial view of the icon prepares the visitor as she continues up the mountain and reaches a side path which leads to the large Buddha and a small ledge on which it can be worshiped up close. Not only does this layer of Buddhist activity take advantage of the ways in which people move through the landscape, it also makes use of the natural features of the rocky mountain to present a manifestation of the Buddha that accentuates one part of the figure. His head is three-dimensional and refined, albeit a bit flattened. The wide face, thick lips, swollen eyes, and the mannered portrayal of the chin diminish his overall liveliness. This, however, is contrasted with the rest of his body, which is only summarily depicted as incised lines on the unmodeled surface of the cliff. This specific combination of depicting the head and shoulders in greater three-dimensionality and with a higher level of polish than the rest of the body is a feature more commonly seen in rock-carved images of the Buddha from the late ninth to early eleventh centuries.55 While it is difficult to say why this style of sculptural practice developed around this time, the style of the giant Buddha near Sangsŏn Hermitage suggests its later date of production and renders the temporal range of the Buddhist layering of the valley visible to the visitors of the mountain.

Thus, the imagery of this valley offers a way to consider the variety of times and groups that produced the environmental palimpsest of Namsan. In the absence of textual or epigraphic sources, it is difficult to say with certainty what the process and context for each sculpture was. A particularly enigmatic example is the headless seated Buddha statue from temple site 2 (Figure 17). This image is sumptuously dressed in several layers of finely detailed outer garments, robes, and inner garments replete with beautifully tied belts and knots. In its present condition it is 160 cm in height; therefore, in its complete state it would have been a sizable and awe-inspiring icon. Yet its original location, beyond the fact that it is from Samnŭng Valley, is unknown. Therefore, it can only serve to reinforce the notion that the mountain and this valley were sites of repeated, substantial, and deliberate patronage.

In addition to its densely aggregated character, this valley was also very likely a place of concealment. It is unfortunate that the indigenous practices of the time can only be seen obliquely in the process of their sublimation by the Buddhist institution. Very little, outside of occasional allusions to royal- and state-level rituals, was recorded about the worship of boulders and large expanses of rock in Silla before the acceptance of Buddhism. Yet looking at the location of natural features that are known to have held significance to the Silla people, we can approach an understanding that the Buddhist sculptures and temple sites in Samnŭng Valley were positioned very purposefully. For example, 50 m away from the incised double triad rock carving of site 3 is a place currently known as Kulbawi (Cave Boulder) where a small cavity contains vestiges of fire and offering practices (Kyŏngju munhwajae yŏn’guso 2000, p. 160). A bit further north is a place called Yongwangdang (Dragon King Hall), where a natural 4 m-wide pool is filled with water that trickles down from the mountain (Kyŏngju munhwajae yŏn’guso 2000, p. 161). Yongwangdang also has an enclosed space where people, even today, leave offerings for the dragon king. Though more research is needed on these caves and pools to clarify their dates of use, it is likely not a coincidence that so many spiritually charged landforms can be found so close to the temple sites.

4.3. Graves on Namsan

Given the sacred character of Namsan and its proximity to the capital of Silla, it is not surprising that the mountain was a common site for funerary rites. Like most of the valleys on Namsan, Samnŭng has many graves. Several well-preserved burial urns excavated from Samnŭng have been published in the KNRICP’s catalog of Namsan (Kyŏngju munhwajae yŏn’guso 2000, pp. 136–37). Based on the material and the shape of the stone and ceramic urns, as well as the more general acceptance of the Buddhist process of cremation, they are thought to date to a point after the 7th century. Many of these urns are made of earthenware and often feature an inner and outer container. One such burial was composed of a Chinese Yue ware ceramic inner urn and an earthenware outer urn, indicating that this type of burial method and location was employed by well-to-do members of society (Kyŏngju munhwajae yŏn’guso 2000, p. 136, images on pp. 368–69).

As previously mentioned, Samnŭng Valley derives its name from the three tumuli that line its entrance. Beyond these mounds, there are five other tumuli that can be found along the western and eastern bases of Namsan, and eleven in surrounding areas. Clearly this space was considered a favorable burial site. During the Unified Silla period, and most likely even before the unification, geomantic principles were especially important in the selection of tombs (Vermeersch 2001, p. 188).56 Due to a paucity of contemporaneous sources, it is unclear what Silla people understood Namsan’s geomantic characteristics to be; however, given the presence of many royal tumuli in the vicinity, it is safe to say that the mountain was considered auspicious.57

Furthermore, the presence of megalithic burials on Namsan suggests that this function was not a unique development of the Silla kingdom. Bronze Age period dolmen tombs have been found in Kuksa Valley and within the valley near the Yongjang temple 茸長寺 (locations in Figure 1) (Kyŏngju munhwajae yŏn’guso 2000, p. 126). Though a thorough excavation has yet to be conducted on these tombs, the discovery of stone ax heads, rope-pressed pottery sherds, and stone blades confirm the use of this mountain during the Bronze Age (Kyŏngju munhwajae yŏn’guso 2000, p. 345). Considering the spatial context of the Yongjang dolmen at the mouth of the valley, it would have been clear to later Buddhists visiting the temple that this had once been a sacred burial ground. As recounted above, repurposing ancient rocks as foundation stones in Buddhist monasteries was already underway at nearby Hwangnyong-sa. While this particular megalith was not reused as such, the space of the valley and its numinous symbolism was ostensibly ‘recovered’ by Buddhist monastics when they constructed the Yongjang temple in such close proximity.

The Yongjang temple was renowned both in the eight century and well into the subsequent dynasty as the residence of the Silla Yogācāra master Taehyŏn 大賢 (fl. 742–765). He is recorded in the Samguk yusa as frequently circumambulating a miraculous sixteen-foot seated image of Maitreya which would turn its face to follow him along his path. Presently, the Yongjang temple site contains a headless statue carved in the round, seated on a tall, very unique three-tiered lotus pedestal.58 The presence of Taehyŏn and the extraordinary icon must have brought followers to this valley in some number.59 With the installation of such a Buddhist nexus, and the establishment of a separate temple site with a stone seated Medicine Buddha icon 600 m away, in Pŏptang ravine temple site 1 (location in Figure 1), Silla Buddhists augmented the symbolic and formal layers of the mountain to further their ‘re-covery’ of Namsan.

5. Conclusions

The process of re-covering Namsan as a Buddhist place was given form largely during the Unified Silla period by re-inscribing and repurposing already spiritually charged and strategic locales. As an example, when the military needs of the capital changed, with the subjugation of the Kaya statelets in the sixth century and the consolidation of power by Silla in the seventh century, the fortifications on Namsan began to play a more symbolic role, playing host to Buddhist imagery. In the wake of the adoption of Buddhism by the Silla court, their efforts to convert the people and the landscape were predicated on the incorporation of pre-Buddhist beliefs into a program of ideological and physical transformation. Places like Hwangnyong-sa were built on top of ancient dolmens and auspicious boulders were recast as meditation stones or manifestations of Buddhist presence. Through the dissemination and subsequent reification of the idea that Buddhism was ingrained in the landscape of Silla, places that had once been the dwellings of indigenous spirits came to be understood as intrinsically Buddhist.

On Namsan, its strategic, social, and religious meaning lent power and mystical symbolism to the Buddhist imagery that later augmented its peaks and valleys. The mountain became a spatial palimpsest, where layers of sculptures, architecture, and images of various times were constructed in relation to pre-existing forms and beliefs. Through an analysis of the inscription and re-inscription of Buddhas and bodhisattvas in Samnŭng Valley, this study has proposed the continued importance of this form of visual conversion during the Unified Silla period while also highlighting the differences in their installation and ritual use. By investigating the layered character of the environment and examining the dialectical relationship between the superimposed strata, this paper has presented a spatially and temporally holistic assessment of Namsan that differs from previous scholarship, which is primarily focused on specific sites.

By the time of Silla’s unification of the peninsula in 668, the belief in the numinous power of cultic rocks had been categorically subsumed into the Buddhist program. For the people of Silla, the mountains that surrounded their capital in the Kyŏngju basin, and the large granite boulders that composed them, were hallowed. Thus, appropriating the materiality and sacrality of stone was a strategic choice. Namsan’s images can be understood as manifestations of the idea of a Buddhist territory not only because they evince the belief that Silla was a kingdom layered with the past eons of the dharma but also because they created an environment-as-palimpsest of religious practice and artistic creation. Though the indigenous religious beliefs in mountain spirits, efficacious rocks, and auspicious sites were subsumed into the Buddhist fold, they can still be seen through and underneath the layers of temples and stone sculptures that dot the Korean peninsula today.