The Journal of Korean Art and Archaeology

“The scene depicted in the painting represents the court of the Vairocana Buddha. Flanked by the bodhisattvas Manjusri and Samantabhadra, the central deity Vairocana sits in the lotus position with his hands assuming an unusual mudra that helps to distinguish him iconographically. In front sit the Ten Bodhisattvas, facing a large assembly of assorted members of the Buddhist pantheon: these figures can be identified through cartouches inscribed in gold ink, and match perfectly the description of the Buddha’s assembly recorded in the sutra text… Representational habits and comparison with dated works suggest that this painting was produced in the first half of the 14th century, and the auspicious srivastas on the palms of the central deities reflect a common Goryeo-period attribute. Although its function is unclear, it is possible to speculate that it was hung as part of a ritual observance involving the Sutra of Perfect Enlightenment such as the ‘Perfect Enlightenment Retreat’ formulated in the 8th century by the Chinese monk Zongmi.” Fig.1. A Transformation Tableaux of the Sutra of Perfect Enlightenment. Late Goryeo Dynasty (918-1392), c. Thirteenth or fourteenth century. Ink and colors on silk, 165.5 x 85.5 cm. Photograph ©2013 Museum of Fine Arts, Boston. To understand the iconography and reference of the Boston scroll, this paper focuses on the Rocana triad, or images of the Rocana assembly, and traces the lineage of the Goryeo painting back to China. Scholars have previously identified a number of similar Rocana triad images in Sichuan, Hangzhou, and Japan, and have asserted that icons of this type derive from Chinese images produced for the Huayan (華嚴, Avatamsaka) Sect of Buddhism in the Five Dynasties (907-960) and Song (960-1279) periods. Since most of the related images are found in Sichuan, some scholars have speculated that Sichuan might be the likely origin of the iconography. However, the earliest extant similar image is located in Hangzhou, which has traditionally enjoyed a much higher status than the Southwest region of China, in terms of both religion and culture. In fact, it is almost inconceivable that, during the Song and Yuan (1279-1368) periods, the far Southwest region could have exerted any influence over Hangzhou, which was then the nation’ s primary political and Buddhist center. Therefore, the images found in Hangzhou should be the key to understanding the iconography and references of the Boston Goryeo scroll. The goal of this paper is to analyze the transmission of Huayan Buddhist art from China to Korea by focusing on the iconography of the Boston Goryeo scroll and the Rocana assembly found in niche 5 of Feilaifeng in Hangzhou. The paper examines how the artists inherited the tradition and created unique features for the Rocana assembly, as well as how Hangzhou played an important role in the transmission of Huayan Buddhist images. Textual Reference Actually, the Boston Goryeo scroll does not correspond exactly to the sutra. The full name of the Sutra of Perfect Enlightenment is Dafangguang yuanjue xiuduoluo liaoyijing (大方廣圓覺修多羅了義經), and it was translated by the Kashmir monk Buddhatrata (佛陀多羅) in 655 at the White Horse monastery (白馬寺) of the Eastern Capital Luoyang. A Chinese Mahayana sutra divided into 12 sections, the sutra details a series of discussions on meditation practice, with the historical Buddha Sakyamuni answering questions asked by Manjusri, Samantabhadra, and ten other bodhisattvas. The text deals with such issues as the origin and meaning of ignorance, sudden and gradual enlightenment, and original Buddhahood. Although most figures in the sutra correspond to those depicted in the Goryeo scroll, they do not match precisely. In both the Goryeo scroll and the sutra, those gathered to listen to the Buddha’s preaching include the 12 bodhisattvas; 80,000 strong men holding vajras (八萬金剛); 28 Heavenly Kings (二十八天王, including Brahma, 大梵天王); and 100,000 Ghost Kings (十萬鬼王) and Guardian Kings (諸天王). However, some attendees mentioned in the sutra do not appear in the scroll; the eight groups of protectors (天龍八部) are absent, for example, as are the King of Sumeru (須彌山王), and several others. The most notable difference, however, is the identity of the main Buddha; in the sutra, the main Buddha is Sakyamuni, but the scroll features Rocana, the Supreme Buddha and the main Buddha of Huayan teaching (see the discussion below), whose name does not appear in the sutra. Meanwhile, many of the iconographic and formal details of the Goryeo scroll are not mentioned in the sutra. In the painting, for example, the main Buddha is seated in the lotus posture, wearing the typical robe and a crown with a strip of fabric that hangs down from either side over his shoulders. His elbows are bent, and his exposed forearms are raised and flexed, permitting the hands to form a mudra with the palms facing upward and the thumbs touching the ring fingers. This iconography—especially the unusual mudra—distinguishes the main figure as Vairocana, rather than as Sakyamuni. Moreover, the sutra states that Bodhisattvas Manjusri and Samantabhadra were the first of the 12 great bodhisattvas who asked the Buddha to preach the Buddhist dharma, but it does not otherwise afford them any special position. In the scroll, however, those two bodhisattvas are larger than the other figures and very prominently placed alongside the main Buddha, dominating the upper half of the representational space. The three figures form a triangle, with Manjusri seated on a lion to the left, and Samantabhadra seated on an elephant to the right. Each bodhisattvas is flanked by a groom and a small boy who kneels in front of each animal. Though not described in the sutra, this composition and iconography claim a long religious and artistic lineage. In addition, the artist painted ten demons on the left and eight strong male figures on the right, representing the 100,000 Ghost Kings and the 80,000 strong men who hold vajras. Beneath these two groups are four guardian kings: two on either side, with an attendant behind each. On the right side, one guardian king holds a pagoda, portraying the Northern Guardian King Vaisravaṇa. In the lower section are 15 bodhisattva-like figures, representing the 28 Heavenly Kings (二十八天王); the figures include the Brahma, who is mentioned in the sutra, and also Sakra Devanam-indra, who is not mentioned in the sutra. Two apsarasas fly above the main Buddha (one on either side), representing the eight groups of protectors mentioned in the sutra. This iconography raises obvious questions. Since most figures in the scroll correspond directly to the sutra, why has the main Buddha been switched from Sakyamuni to Rocana? Why did the artist highlight the three figures of Rocana, Manjusri, and Samantabhadra? In fact, the Boston Goryeo scroll’s Rocana triad is not based exclusively on the Sutra of Perfect Enlightenment, but draws from a long tradition of Huayan belief and art. In Chinese Buddhist texts, the Rocana triad is called Huayan sansheng, or the “Three Saints of Huayan” (華嚴三聖). The worship of Rocana Buddha is associated mainly with the three Huayan sutras.3 In these sutras, Rocana is the superior Buddha of dharma bodies and the dharmarealm. Rocana, also named Vairocana (毘盧遮那)4, lives in a pure land called the “Lotus World,” which attracted numerous devout Buddhists. The Huayan Sect was founded during the Tang Dynasty (618-907) by Monk Dushun (杜順, d. 640). According to Chapter 44 of the Huayan sutra, which was translated by Buddhabhadra (佛馱跋陀羅, 359-429), Sakyamuni Buddha preached the dharma at Jetavana, a park near Sravasti. At that gathering, the Buddha was assisted by two exceptional bodhisattvas: Manjusri (文殊), an idealization of wisdom, and Samantabhadra (普賢), famed for keeping his vows. Although the sutra highlights these two great bodhisattvas, it does not mention Rocana Buddha, and is thus does not relate directly to the Rocana triad. By contrast, Rocana triads are widely discussed in books and treatises by Chinese monks, especially those of the Huayan Sect. For example, the monk Fazang (法藏, 643-712), another founder of the Huayan Sect, often mentioned Rocana and the two great bodhisattvas Manjusri and Samantabhadra in his works, as in Chapter 1 of Huayanjing tanxuan ji (華嚴經探玄記, Record on Exploring the Profound Theory of Huanyan Sutra). The Rocana triad depicted in the Boston Goryeo scroll is not recorded in the Sutra of Perfect Enlightenment, but does appear in some Chinese monks’ commentaries related to the rituals of the sutra. After translating the sutra, the Tang monk Zongmi (宗密, 780-841) wrote a treatise called Yuanjue jing daochang xiuzheng yi (圓覺經道場修證儀, Rite in the Ritual Place for the Practicing and Testimony of Yuanjue jing), in which he described how to conduct the ritual of the sutra. According to Monk Zongmi, to conduct the ritual, one should install three images—Rocana, Manjusri, and Samantabhadra—to serve as the focus of meditation and worship. Zongmi then mentions the other ten bodhisattvas and various deities from the Sutra of Perfect Enlightenment, to support and protect the ritual place (daochang, 道場). A Northern Song (960-1127) monk named Jingyuan (淨源, 1011-1088) also wrote commentaries on this ritual, which were compiled in 1069 under the title Yuanjue jing daochang lueben xiuzheng yi (圓覺經道場略本修證儀). Jingyuan states that one should install the Rocana triad in the hall for the ritual, so that they appear there as the Three Saints of Huayan. A case containing the Sutra of Perfect Enlightenment should be placed in front of the Rocana Buddha, together with a lotus lantern, incense, and various dharma implements. Buddhist devotees entering the ritual place should chant the name of Rocana while making one circumambulation around the Buddha. Then the monks would invite other deities, including the ten other great bodhisattvas, deities, and protectors mentioned in the sutra, and make offerings to them. These records clearly demonstrate the importance of Rocana Buddha and his principal attendants in the ritual of the Sutra of Perfect Enlightenment. Rather than faithfully depicting the scenes as described in the sutra, however, the Boston scroll actually represents the model for conducting the ritual of the sutra. The fact that neither the sutra nor the commentaries provide any textual references for the iconography of the Rocana triad raises several questions: Why are the two bodhisattvas represented on animals and assisted by a groom and a child? What is the reference for this unconventional iconography? The answers to these questions may be found by tracing the tradition of the Boston scroll back to China, particularly the earliest surviving examples of a similar Rocana triad, found in the Hangzhou area, in eastern coastal China. Artistic Tradition Feilaifeng (“Peak that Flew in from India”) is the largest extant rock-cut cave temple site in the Hangzhou area; it is notable for its caves and cliff sculptures produced from the tenth through the fourteenth centuries. The earliest surviving images from this site date to the tenth century, while more than 200 images survive from the Song Dynasty, and more than 100 remain from the Yuan Dynasty. Among these images is a Rocana triad similar to the one depicted in the Boston Goryeo scroll (Fig. 2). The triad, found in niche 5 at Feilaifeng, is dated 1022, and bears an inscription that states that the image depicts 17 figures of Rocana (盧舍那, Lushena) Buddha’s assembly.5 The main Buddha in niche 5 and the one in the Boston scroll are similar in terms of costumes and gestures, and the attendant bodhisattvas in both images are sitting on the same animals, indicating that main figure in each work is Rocana Buddha. Fig.2. Rocana Buddha assembly in niche 5 of Feilaifeng. Northern Song Dynasty (960-1127), dated to 1022. Images of Feilaifeng Peak (飛來峰造像) edited by Gao Nianhua (高念華). (Beijing: Wenwu chubanshe, 2002, fig. 17). Like the Boston scroll, the three main figures in the Feilaifeng triad form a triangle and are larger than the other figures, demonstrating their importance in the group. Behind both Samantabhadra and Manjusri are two guardian kings, followed by two bodhisattvas. Samantabhadra is seated atop an elephant while Manjusri rides a lion, and the grooms for the animals can be seen in the lower left and right corners of the niche, respectively. At the bottom of the composition, two small boys are crouch on either side of the pedestal in front of the animals. Two apsarasas, one of eight groups of Buddhist guardians, appear on the left and right, above the lobed outer edge of the niche; the apsaras were carved in lower relief than the other figures. All these figures symbolize the members of the gathering to hear Rocana’s sermon. The niche and the scroll feature similar compositions and related iconography for the major figures. In fact, the scroll seems to be a more complex version of the Feilaifeng triad. By the second half of the eighth century, images of the Rocana assembly had become popular, and often featured Samantabhadra riding an elephant and Manjusri riding a lion, along with grooms, small boys, and other attendants. A good example of such a depiction is found in cave 155, which dates to the ninth century, at the Chonglong shan (重龍山, “Double Dragon Mountain”) grottoes in Zizhong, Sichuan province. In this cave, Rocana wears a robe similar to that of the Rocana at Feilaifeng. Manjusri is shown atop a lion and Samantabhadra on an elephant, and both are accompanied by a groom and a small boy. Beside each stands one lesser bodhisattva and one guardian king, so the total number of figures is fewer than the Feilaifeng triad. Ishida Hisatoyo has reasonably suggested that the two small boys in the Rocana assembly at Chonglong shan represent Sudhana, a boy from Fucheng (“Blessed City,” 福城) who is mentioned in Chapter 62 of the Huayan sutra translated by Siksananda. According to the sutra, when Sudhana was born, numerous treasures filled his chamber. In his past lives, he had made offerings to Buddhas and established his good karma by committing no wrongdoings, so people called him “Shancai” (善財, “Kindness and Assets”). According the Huayan sutras, Sudhana once visited and paid homage to 53 individuals, including Manjusri and Samantabhadra. Thus, the two small boys in the image from cave 155 of Chonglong shan represent Sudhana’s visits to each of the two bodhisattvas. The depiction of the small boys in cave 155 is so similar to that of the Goryeo scroll and the Feilaifeng niche that we can only conclude that this tradition continued into the fourteenth century. Although the Rocana figures from both the Boston scroll and Feilaifeng are thought to derive from the Huayan Sect, the style of their crowns are of a type more commonly associated with Esoteric Buddhist art. In fact, Rocana is the main Buddha in both Esoteric and Huayan Buddhism, and eminent monks from the two schools have used the same concept to explain the meaning of Rocana/Vairocana. For example, in Chapter 3 of Huayanjing tanxuanji (華嚴經探玄記, Record on Exploring the Profound Theory of Huanyan Sutra), Fazang (法藏, 643-712) states: Today [I] checked the Indian version [of the Huayan sutras] again; all of them call [the superior Buddha] ‘Vairocana’ (毗盧遮那). ‘Rocana’ (盧舍那) could be translated as ‘Light Shining’ (Guanmingzhao, 光明照), and ‘vai’ (毗) means ‘universally.’ [Thus, the name of this Buddha] means ‘Light Universally Shining’ (光明遍照). In Commentary on the Mahavairocana Tantra, an important work of Esoteric Buddhism, the Tang monk Yixing (一行, c. 673-727) says that: “The Indian pronunciation of ‘Vairocana’ is another name for the sun, meaning to remove darkness and shine everywhere.” This is why Rocana/Vairocana also is called the “Great Sun Buddha” (大日如来). Therefore, the image of Rocana/Vairocana wearing a crown can be found in both Huayan and Esoteric Buddhism. The origin of the iconography of crowned Rocana can be traced to the Esoteric Mahavairocana Tantra, translated by the Indian monk Subhakarasimha (善無畏, 637-735) and Yixing in 724. This scripture says that the arrangement of Rocana’s hair (i.e. his chignon) acts as his crown. Chapter 4 of Commentary on the Mahavairocana Tantra by Yixing says that because his chignon resembles a crown, Vairocana looks like a bodhisattva. There are currently no known references to a crowned Rocana Buddha associated with the Huayan Sect of the Tang Dynasty, so these Esoteric sutras may also have served as textual references for the production of Rocana images for Huayan Buddhism. This textual influence may also be seen in a woodblock-printed edition of Huayan sutra from the Northern Song period. This printed Huayan sutra, with 80 chapters, was translated by the Khotanese monk Siksananda (實叉難陀, 652-710) in 990-1000, under the sponsorship of monks from the Longxing monastery (龍興寺) in Hangzhou. The frontispiece of the sutra features a Rocana triad, with Rocana wearing a crown, flanked by Manjusri on a lion and Samantabhadra on an elephant, and Sudhana in front of the two bodhisattvas. However, this Rocana Buddha makes a different mudra from the Rocana from Feilaileng and the Boston scroll, with his right forearm raised and his left arm extended down and to the right. As mentioned, the Rocana figures from both the Boston scroll and the Feilaifeng triad hold their hands in an unusual mudra that cannot be found among any other extant images from the Tang Dynasty. Kamata Shigeo has theorized that the Rocana triad images were transmitted eastward from the Sichuan area during the Five Dynasties (907-960) and Song periods and that the triad at Feilaifeng may reflect this influence. The fact remains, however, that no other image of Rocana with similar crown, costume, and mudra can be found among extant Tang and Five Dynasties images in Sichuan. In addition, the Feilaifeng triad is the earliest known work with this type of Rocana image. Also, as Hangzhou, the location of Feilaifeng, was the capital of the Southern Song (1127-1279), it was much more important than Sichuan for the development of Huayan Buddhism (see discussion below). Therefore, it seems highly unlikely that Sichuan could have been the source of the Feilaifeng triad. Niche 5 at Feilaifeng was carved in 1022, after which time many Buddha figures and triads appeared with a similar iconography, i.e. with a Buddha wearing a crown or performing the special mudra and seated between Manjusri and Samantabhadra seated on their respective animals. For instance, cave 3 of the Miaogao shan grottoes in Dazu county, Sichuan, which dates from the Song period, has a seated Rocana wearing a crown and flanked by Manjusri on a lion and a Samantabhadra on an elephant (Fig. 3). A groom and a Sudhana figure appear beside each bodhisattva, and two apsarasas fly above Rocana. Clearly, some features of the Rocana assembly at Feilaifeng, including the composition and the crown, were also in use in Sichuan by the end of the eleventh century. A few Buddha images outside of Hangzhou include a mudra similar to that of the Rocana at Feilaifeng. For example, the Dharma Wheel Pagoda (法輪塔) at Dafowan (大佛灣) in Dazu, which dates to the Southern Song, includes a seated Buddha (with the features of a typical Buddha) making the same mudra (Fig. 4). Although this figure may not represent Rocana, the similarity indicates a connection between Hangzhou and Sichuan. Another example is the Vairocana image painted in The Long Roll of Buddhist Images (Fig. 5), attributed to Zhang Shengwen (張勝溫), a painter in the Dali kingdom (937-1253). The inscription on the side of this Vairocana translates to “Hail to the Buddha of the Great Sun shining everywhere” (南無大日遍照佛). As mentioned, “Buddha of the Great Sun” is another name for Vairocana in Chinese texts. Instead of a crown, this figure has an ushnisha, a protuberance on the top of Buddha’s head symbolizing the expanded wisdom that he gained at his enlightenment. His elbows are flexed, with the hands raised up to shoulder level, making the same mudra as the Northern Song Rocana at Feilaifeng. This scroll is dated between 1173 and 1176, contemporaneous with the Southern Song, and thus is later than the Feilaifeng Rocana. Fig.3. Rocana assembly in cave 3 of the Miaogao shan grottoes, Dazu, Sichuan. Song period (960-1279). Collection of the Sculptures from Chinese Grottoes, vol. 7, Dazu (中國石窟雕塑全集-7-大足), edited by Zhongguo shiku diaosu quanji bianji weiyunhui (中國石窟雕塑全集編輯委員會). (Chongqing: Chongqing chubanshe, 2000, fig. 107). Fig. 4. Seated Buddha on the Dharma Wheel Pagoda at Dafowan in Dazu county, China. Southern Song Dynasty (1127-1279). Summit of Treasures: Buddhist Cave Art of Dazu, China by Angela Falco Howard (Trumbull, CT: Weatherhill, Inc. 2001, fig. 117). Fig. 5 Hail to the Buddha of the Great Sun Shining Everywhere, part of the The Long Roll of Buddhist Images attributed to Zhang Shengwen (張勝溫, c. twelfth century), in the collection of National Palace Museum in Taibei. A Study of the Nan-Chao and Ta-Li Kingdoms in Light of Art Materials Found in Various Museums (南詔大理國資料綜合研究), by Li Lin-ts’an (李霖燦). (Taibei: The Institute of Ethnology Academia Sinica, 1967, plate XXV). Huiyin Monastery and the Transmission of the Rocana Assembly Image The preceding discussion does not suggest that the Rocana Buddha images from Sichuan and Yunnan were directly inf luenced by the image carved in niche 5 of Feilaifeng. To elucidate the artistic source of the iconography of the Boston scroll, we must examine the historical importance of a Huayan monastery and teaching center called Huiyin-si, or “Intelligent Karma Monastery” (慧因寺), located in Hangzhou. During the Song period, Huayan teaching underwent a shift. From the fifth into the early ninth centuries, belief in Rocana prospered in Buddhist practice, research, and art. However, in the fifth year of the Huichang reign (845) of the Tang period, the Huayan Sect began to decline; this occurred during the reign of Emperor Wuzong (r. 841-846), an ambitious ruler who did not believe in Buddhism and who openly persecuted Buddhists. Toward the end of the tenth century, however, the Huayan Sect began to recover, thanks in large part to support from Chinese monks and patronage from the Goryeo royal family. In the early period of the Yuanyou reign (1086-1094), the Goryeo Buddhist Controller Uicheon (義天, 1055-1101) traveled to Bianjing (present-day Kaifeng, in Henan province), the capital of Northern Song, in order to ask the Song court to teach him about Huayan and to assist him in transmitting it to Korea. In 1088, the Song court recommended that Uicheon go to the Huiyin monastery in Hangzhou to study the teachings of Huayan Buddhism. In conjunction with the monk’ s request, the court changed the Huiyin monastery from a Chan institution into a Shifang jiaoyuan, or “Ten-direction Teaching Cloister” (十方教院), belonging to the Huayan Sect. Therefore, the Huiyin monastery was nicknamed the “Goryeo monastery” (高麗寺). From that time on, the Huiyin monastery was the most important place in China for the study of Huayan teachings. In 1101, King Sukjong of Goryeo (r. 1095-1105) sent an envoy to commission a hall for Huayan sutras at the Huiyin monastery, and also ordered that an image of Rocana flanked by Samantabhadra and Manjusri be installed in the same hall. The Huiyin monastery continued to play a key role in Huayan education during the Southern Song period, and shared a particularly close relationship with the Southern Song imperial family. Emperor Ningzong (r. 1195-1224), for example, wrote a title plaque reading “Pavilion of the Huayan Sutras” (華嚴經閣) for the monks of Huiyin. During the Baoqing (寶慶) reign (1225-1227) of Emperor Lizong (理宗, r. 1225-1264), the court established the “Hall of Ruling Deities” (神御殿) for the benefit of Emperor Ningzong (寧宗, r. 1195-1224) and the “Confession Hall for the Period of Surangama Blessing the Sages” (愣嚴祝聖期懺堂) in the monastery. Meanwhile, the court commanded that the coffin of Princess Chengguo (成國公主), one of the emperor’s aunts, be placed at Huiyin to make offerings. In the Chunyou (淳祐) reign (1241-1252), Emperor Lizong commanded that Guangguo yuan, or the “Extending Fruit Cloister” (廣果院), should be assigned to Huiyin monastery. All of these records indicate the strong patronage of the Southern Song imperial family, and Huiyin’s status as the leading monastery of Huayan teaching. Given the prestige of Huiyin monastery, the Rocana triad of Huiyin monastery could have served as a reference for artists during the twelfth and thirteenth centuries. The Huiyin triad no longer exists, but the triad in niche 5 at Feilaifeng likely gives some idea of its appearance, since it represents some of the typical iconographic characteristics and style of that period. The previously discussed similarities between the images at Feilaifeng and those from Sichuan and Yunnan suggest a transmission between these three areas. Although it seems highly unlikely that the Feilaifeng triad could have been regarded as a key work that influenced other regions, that image may serve as a representative example of a Rocana triad as depicted in the Hangzhou area in the Song period. Moreover, the Huiyin and Feilaifeng Rocana triads probably shared similar iconographic characteristics. The similarities of Rocana images from Feilaifeng, Sichuan, and Yunnan can likely be attributed to the influence of the Huiyin monastery. The Feilaifeng triad can also help us to examine the transmission of Huayan teaching between Hangzhou and Japan. Some Japanese collections include paintings of Rocana making the same mudra as the Rocana in the Feilaifeng triad. Around 1185—late in the Heian period (平安, 794-1192) and early in the Kamakura period (鐮倉, 1185-1333)—Vairocana and Sudhana images related to the Huayan sutras became a popular subject in Japanese painting. Ishida Hisatoyo has suggested that this phenomenon probably was related to the transmission of Huayan belief from Song China. For example, Painting of Good Teachers in the Sea Congregation of Huayan (華嚴海會善知識圖), from the collection of Todaiji Temple (東大寺) in Nara, shows a seated Vairocana with a pose, mudra, and costume that resemble those of the Feilaifeng Rocana, and surrounded by virtuous visitors (Fig. 6). Ishida reasonably asserts that the Todaiji painting reflects the Song style and that the artist probably had references from the Song China. Fig.6. Kegon kaie zenchishiki zu (華嚴海會善知識圖, Painting on the Good Teachers in the Sea Congregation of Huayan), from the collection of Todaiji in Nara. Ca. Kamakura period (1185-1333). “Paintings of the Huayan Sutras” (華嚴經繪) by Ishida Hisatoyo (石田尚豊) in Arts of Japan (日本の美術) 270, fig. 12. In seeking the source of Vairocana images in Japanese painting, Ishida provides two possible Chinese works. One is a Vairocana triad now in the collection of Kenchoji Temple (建長寺) in Kamakura, in which the main figure assumes a posture and mudra similar to the Feilaifeng Rocana, but with his hands in front of his chest (Fig. 7). Manjusri appears at the lower left, riding a lion, and Samantabhadra appears at the lower right, riding an elephant, an arrangement that closely resembles the three main figures of the Feilaifeng triad. This painting was brought to Japan from the Southern Song in 1246 by a Chinese monk named Lanxi Daolong (蘭渓道隆, 1213-1278). Ishida’s second possible source is the Feilaifeng triad, which he states represents a style of Rocana assembly that became popular in the Zhejiang area during the Song period. Hence, Ishida believes that the iconography of Huayan painting was probably imported into Japan via Zhejiang early in the Kamakura period. Fig.7. Vairocana Triad from the collection of Kenchoji (建長寺). c. first half of thirteenth century. “Paintings of the Huayan Sutras” (華嚴經繪) by Ishida Hisatoyo (石田尚豊) in Arts of Japan (日本の美術) 270, fig. 14. Ishida’s argument is convincing, but the specific relationship between the Feilaifeng triad and Huayan painting in Japan requires further research and clarification. As discussed, the Huiyin monastery held a place of great influence in Huayan teaching during the Song period, and the Vairocana triad at the Huiyin monastery, no longer extant, was probably similar to the one at Feilaifeng in terms of iconography. When Japanese monks visited Southern Song and sought to acquire Huayan paintings as models for their own works, their first choice would have been the Huiyin monastery. The Feilaifeng triad, therefore, can be considered as a bridge connecting Huayan belief and the art of Hangzhou—particularly Huiyin monastery—and Japan. Hangzhou and the Boston Rocana Scroll The relationship between Hangzhou and Japan is also significant for exploring the relationship between Hangzhou and Goryeo in the area of Huayan teaching and its art. While the Boston scroll was painted by a Goryeo artist, its iconography certainly derived from China. During the Song period, the Huiyin monastery would have been the destination of choice for Goryeo monks who wished to study Huayan Buddhism and bring works of Buddhist art associated with the Huayan Sect back to Goryeo Korea. The Goryeo scroll is believed to have been painted in the thirteenth or fourteenth century, which would likely make it contemporaneous with China’s Yuan Dynasty (1279-1368), meaning that the artist could have followed old traditions or possibly used a new iconography from Yuan. In the same tradition, an extant Goryeo sutra frontispiece dated to 1357, which depicts the same sutra as the Boston scroll, also includes a Rocana assembly, showing the Rocana Buddha with similar costume and mudra. The Goryeo tradition of depicting Rocana with the special mudra also can be seen in a Rocana Buddha painting from Sinwonsa Temple in South Chungcheong Province, painted in 1664 (during the Joseon Dynasty) (National Treasure 299). Under Mongol rule, imperial patronage and elite participation in Huayan Buddhism remained strong in the Hangzhou area. According to Ming (1368-1644) texts, when Kublai Khan (1215-1294) heard that the Huiyin monastery in Hangzhou was the most important temple of the Huayan Sect, he sent gold and commissioned a stele for the monastery. A Goryeo prince who was the son-in-law of Kublai Khan invited a Huayan monk named Pangu (盤谷, active first half of thirteenth century) to preach Huayan Buddhism in the Huiyin monastery. In addition, some well-known Yuan artists enthusiastically participated in the artistic activities of the monastery. For instance, Zhao Mengfu (趙孟頫, 1254-1322) composed several stele inscriptions, and Huang Gongwang (黄公望, 1269-1354) painted a mural in the Seven Patriarchs Hall. Therefore, the Huiyin monastery, which was still patronized by Goryeo royalty during the Yuan period, would have been the first choice for Goryeo monks and Buddhist artists looking to bring Huayan icons back to Korea. As a result, the Boston Goryeo scroll can be associated with the iconography that flourished in the Hangzhou area.

The main halls of Buddhist temples of the Joseon Dynasty are often crowded with numerous paintings. They typically have paintings enshrined on three walls, each of which carries its own meaning and significance. On the central wall, behind the main statues, there is usually a large painting that is meant to evoke the world of Buddha, which the main hall is meant to embody. However, no statues are placed in front of the right or left walls of the hall, and the paintings on those sides are intended to be objects of religious worship. Many scholars have categorized Joseon Buddhist paintings according to their placement in the main hall (i.e. central, left, or right wall), but there has been some disagreement about the respective meaning of the three locations and the arrangement of the paintings (Hong Yunsik 1977, 143-152; Mun Myeongdae 1986; Jeong Myounghee 2013). Previous research has demonstrated that the various methods of organizing and arranging the main hall are not the result of Buddhist doctrine, but instead likely reflect an attempt to incorporate folk religious beliefs, which often involve praying to various deities for specific earthly needs or for good fortune, rather than focusing on the afterlife. However, the differences and changes in arrangement are too broad and numerous to be explained with just a few simple principles. The role and significance of Buddhist paintings in the organization of temple space must be re-examined, particularly in the context of religious rituals. During the Joseon Dynasty (1392-1910), one type of Buddhist painting sometimes enshrined in the main hall of Buddhist temples was a Hyeonwangdo (現王圖), depicting King Yama (閻羅大王), the fifth king of the “Ten Kings of the Underworld,” who is known as Hyeonwang in Korean. Until now, Hyeonwangdo (literally, “paintings of Hyeonwang”) have been neglected by scholars of Buddhist art, who typically have considered them to be merely a subgenre of paintings depicting the underworld. Notably, however, while other paintings of the underworld were usually enshrined in the Judgment Hall (冥府殿), Hyeonwangdo were enshrined in the main hall. This article focuses on Hyeonwangdo in order to examine how Buddhist paintings with the same theme acquired differing religious meanings, depending on where they were enshrined. A close examination of the scroll painting Hyeonwangdo from Seongbulsa Temple (1798), now housed in the National Museum of Korea (Fig. 1), provides more details about the background, production, and function of Hyeonwangdo during the Joseon Dynasty. In March 2005, a discovery was made that significantly enriched the understanding and interpretation of this particular painting. While the painting was undergoing conservation treatment, several objects and a document entitled Hyeonwangtaeng wonmun (現王幀願文) were found installed inside the upper rod of the scroll. Through a document called Hyeonwangcheong (現王請), which is a liturgy of the formal procedures for performing the ritual for Hyeonwang, and through a ritual called “Hyeonwangjae” (現王齋, final ritual for Hyeonwang), I shall investigate how the practices of worshipping, making offerings, and praying to Hyeonwang became formalized and institutionalized. In this article, I shall also explore how Hyeonwangdo developed separately from Siwangdo (十王圖), which are paintings of the Ten Kings of the Underworld, through the process of formalization (Jeong Myounghee 2005). Fig. 1. Hyeonwangdo from Seongbulsa Temple. 1798. Color on silk, 103.7 x 85.0 cm. (National Museum of Korea) Background of Hyeonwangdo Origin of Hyeonwang and the Hyeonwang Faith Hyeonwangdo are a special type of Buddhist paintings that were enshrined in the main hall of Buddhist temples during the Joseon Dynasty. The name “Hyeonwang” refers to King Yama (閻羅大王), the fifth king of the underworld. In Buddhism, there are Ten Kings of the Underworld—from the first king Jingwangdaewang (秦廣大王) to the tenth king Odojeollyundaewang (五道轉輪大王)—and paintings depicting all ten kings are known as “Siwangdo.” Hyeonwangdo, on the other hand, are paintings that depict only King Yama, the fifth king, typically accompanied by a group of deities and guardians in the lower part of the painting. While Siwangdo typically feature vivid scenes from hell, no such scenes appear in Hyeonwangdo. The most important difference between Siwangdo and Hyeonwangdo, however, is that the former are enshrined in the Judgment Hall, while the latter are enshrined in the main hall. Previous studies have speculated that Hyeonwangdo were produced at temples that were unable to construct a Judgment Hall, or that they were simply substitutes for Siwangdo (Hong Yunsik 1980, 58; Kim Jeonghui 1996, 172). However, in some cases, both Hyeonwangdo and Siwangdo were produced at the same temple. Furthermore, although both types of painting concern the afterlife, they clearly differ in terms of the subject matter and the site of enshrinement. Therefore, we can assume that there are also differences between the rituals and religious practices associated with the two types. These differences can be explained by examining texts on Buddhist rituals to see how the Hyeonwang faith had become differentiated by the late Joseon period. In China, the first record of King Yama as one of the Ten Kings of the Underworld comes from Fo shuo yuxiu shiwang shengqi jing (佛說預修十王生七經, Sutra Spoken by Buddha on Rituals to Be Performed While Living, in Preparation for Judgment by the Ten Kings of Hell, hereinafter Sutra on the Ten Kings), which was translated into Chinese around the ninth century, late in the Tang Dynasty (618-907) (Kim Jeonghui 1996, 101-103). The sutra describes the ritual of chiljaeui (七齋儀), which involves making an offering to the ten kings and repenting of sins in order to be reborn in the Pure Land, or Western Paradise, after death. Importantly, in Sutra on the Ten Kings, Shakyamuni predicts that King Yama would eventually achieve enlightenment and be reborn as Hyeonwang, thus marking the origin of the name Hyeonwang. Whereas the other nine kings are thought to preside in judgment over the souls in hell, King Yama is identified with Ksitigarbha Bodhisattva, who promises to save souls that have already entered hell. Thus, people believed that they should pray to King Yama, i.e. Hyeonwang, in order to solicit mercy for deceased loved ones who they feared might be in hell. The basic evidence associating King Yama with Ksitigarbha Bodhisattva comes from two Buddhist sutras: Dasheng daji dizang shilun jing (大乘大集地藏十輪經, The Scripture on Ksitigarbha and the Ten Wheels in the Great Mahayana Compendium, translated by Xuanzang [玄奘, c. 602-664]) and its translated version, Dafangguang shilun jing (大方廣十輪經). In 1246, during the Goryeo Dynasty (918-1392), Sutra on the Ten Kings was engraved on woodblocks for printing, and the King Yama section features an illustration of King Yama together with Ksitigarbha Bodhisattva (Fig. 2). This illustration is based on part of the sutra, the translation of which reads: “Through his strong, mysterious will to realize his vow and lead people to enlightenment, Ksitigarbha Bodhisattva manifests himself in various forms: Brahma, the Four Heavenly Kings, a beast, a guardian of hell, and King Yama.” From the time when the Ksitigarbha faith was established, King Yama has been regarded as an incarnation or embodiment of Ksitigarbha Bodhisattva, a belief that is evinced by various sutras. Notably, however, in terms of the general iconography and entourage, Hyeonwang was initially depicted in the same manner as the other kings of the underworld. The only exception is that Hyeonwang is not shown holding a mirror at the entrance of the underworld, unlike King Yama, who clearly holds a mirror when depicted as one of the Ten Kings of the Underworld. In other words, even though sutras proclaimed that King Yama was Hyeonwang, the Hyeonwang belief had not yet become differentiated from the general belief in the ten kings. Fig. 2. Woodblock of Sutra on the Ten Kings. 1246. Treasure #734-3. (Haeinsa Temple). Belief in Hyeonwang led to the rise of a ritual called “yesujae” (預修齋), based on Sutra on the Ten Kings, which was one of the representative Buddhist rituals during the Joseon Dynasty. A description of the formalities and procedures of yesujae was published in Yesu siwang saengchiljaeui channyo (預修十王生七齋儀簒要, Book of Formalities of Rituals to Be Performed While Living, in Preparation for Judgment by the Ten Kings of Hell, hereinafter, Book of Yesujae Formalities). The general belief was that, by conducting the yesujae ritual, a living person could begin to fulfill his or her obligations for the afterlife. The fifth section of Book of Yesujae Formalities offers more details about King Yama, stating that King Yama will eventually become Hyeonwang and will achieve Buddha’s limitless virtue and sanctity. The section goes on to describe the land over which Hyeonwang will rule, as well as Hyeonwang’s excellence and superiority among the Ten Kings of the Underworld (Fig. 3) (Bak Semin 1993, 2-106). Fig. 3. A Book of Yesujae Formalities. 1576. Compilation of Buddhist Rituals and Relevant Records of Korea vol. 2 (한국불교의례자료총서 2권). (Seoul: Samseongam, 1993, p. 106). Despite the strong emphasis on Hyeonwang in Book of Yesujae Formalities, however, the belief in King Yama/Hyeonwang had not yet differentiated itself from the overall belief in the Ten Kings of the Underworld. Therefore, with no worship dedicated solely and specifically to Hyeonwang, there was not yet any demand for Hyeonwangdo. The production of Hyeonwangdo presupposes King Yama’s status as an independent figure of worship, apart from the other nine kings. But when did this differentiation happen? Was there any doctrinal basis behind it? The answers to these questions cannot be found in the orthodox sutras regarding the underworld, but they can be explicated by considering a JoseonDynasty publication entitled Hyeonwangcheong (現王請), which was a liturgy of the formal procedures for performing the ritual for Hyeonwang. The ritual for Hyeonwang underwent changes over many centuries, eventually sparking the demand for a new type of painting and transforming the arrangement of space in Korean Buddhist temples. Hyeonwangcheong and the Production of Hyeonwangdo During the late Joseon period, Buddhist rituals of prayer and offering before sacred images, including Buddhas and bodhisattvas, were codified into the cheongmun (請文) ritual. Literally meaning “written request,” cheongmun invites a particular deity to appear, after which offerings are made to that deity. Hyeonwangcheong is a document that describes how to perform the cheongmun ritual for Hyeonwang. Furthermore, the document itself was used in a ritual at the altar for Hyeonwang that was located inside Daeungjeon Hall. Hyeonwangcheong focuses explicitly on Hyeonwang (along with a few other deities that aid Hyeonwang), so it is reasonable to assume that it provided the basis for the practice of Hyeonwangdo. The earliest known record of the publication of Hyeonwangcheong dates to 1574, in Gwongong jebanmun (勸供諸般文, Compilation of Buddhist Rituals, Fig. 4). Fig. 4. Hyeonwangcheong from Gwongong jebanmun. 1574. Compilation of Buddhist Rituals and Relevant Records of Korea vol. 1 (한국불교의례자료총서 1권). (Seoul: Samseongam, 1993, p. 663). Buddhists believe that after death, a person’s soul goes through ten successive judgments to determine the status of rebirth in our next life. Traditionally, the first judgment is by the first king, Jingwangdaewang (秦廣大王), and it takes place on the seventh day after death; the final judgment is by Odojeollyundaewang (五道轉輪大王) and occurs on the third anniversary after death (Jeong Gak 2000, 317). However, Hyeonwangcheong introduces another ritual, called Hyeonwangjae, to be performed on the third day after death. Thus, the Hyeonwangjae ritual offered a way to pray for the liberation of the deceased’ s soul after only three days, rather than after three years. Such a ritual occurring on the third day after death is otherwise unknown in the Buddhist canon. Hyeonwangcheong begins by explaining the purpose of the ritual for Hyeonwang, before describing how to invite the deities to be present for the ritual and make the offerings. According to the text, on the third day after a person has died, the loved ones of the deceased should perform the ritual of Hyeonwangjae, in order to pray for the liberation of the soul of the deceased and for rebirth in the Pure Land, through offerings and dedication to Buddhism (Bak Semin, 1993, 1:663-664). Unchanged over the years, the ritual described in Hyeonwangcheong is still practiced today. According to Hyeonwangcheong, the ritual should not be dedicated to King Yama, the fifth judge of souls in the afterlife, but rather to Hyeonwang, King Yama’s enlightened form, who is believed to guide the soul of the deceased into the underworld. The procedure described in Hyeonwangcheong was later refined in Gwongong jebanmun (1574) and Jebanmun (1624), and eventually finalized as Hyeonwangjae uimun (現王齋儀文, liturgy for Hyeonwangjae) (Fig. 5), in the 1719 edition of Jebanmun published by Haeinsa Temple. Interestingly, some of the ritual books refer to Hyeonwang as “Seongwang” (聖王). Despite this discrepancy, all of the descriptions are basically the same, with some minor variance among the details of the procedures and the invited deities of Hyeonwangjae. Fig. 5. Hyeonwangjae uimun from Jebanmun. 1715. Compilation of Buddhist Rituals and Relevant Records of Korea vol. 1 (한국불교의례자료총서 1권). (Seoul: Samseongam, 1993, p. 663). According to Gwongong jebanmun (1574), on the third day after a person dies, Hyeonwangjae should be performed to petition Hyeonwang to save the soul of the deceased. The 1694 edition of Jebanmun made by Geumsansa Temple states that Hyeonwangjae should be performed on the third day after death in order to save the soul of the deceased from getting lost in the darkness of the underworld, so that they may eventually be led to the Pure Land. This may explain why there are no scenes of hell featured in Hyeonwangdo (Bak Semin 1993, 2-680 and 681). According to the text, the first deities to be invited are Hyeonwang in the center, along with Indra and Brahma. The next two deities to be invited are Daeryunseongwang (大輪聖王) and Jeollyunseongwang (轉輪聖王). These two kings are not mentioned in Sutra on the Ten Kings, but they regularly appear in the descriptions of Hyeonwangcheong. They also appear in various other books of Buddhist rituals, including Yojimmun (要集文), Yojip (要集), and Cheongmunnyojip (請文要集). As such, Daeryunseongwang and Jeollyunseongwang were gradually defined as Hyeonwang’s left and right attendant bodhisattvas, respectively (Fig. 1a). Hyeonwangdo from the late eighteenth century clearly differentiate Daeryunseongwang and Jeollyunseongwang in order to emphasize the composition of the triad. Fig. 1a. Diagram of Hyeonwangdo from Seongbulsa Temple. ❶ Hyeonwang, ❷ Daeryunseongwang, ❸ Jeollyunseongwang, ❹ judge, ❺ messenger of death, ❻ clerk. After Daeryunseongwang and Jeollyunseongwang, a judge and a clerk are to be summoned, in order to record the good and evil deeds from the underworld. In some of the descriptions of the ritual, two messengers of death named Gamjaesaja (監齋使者) and Jikbusaja (直符使者) are also summoned. These secondary figures (i.e., clerk, judge, messengers of death) also appeared in Siwangdo, but they were gradually omitted from Hyeonwangdo. However, the composition of the triad, with Hyeonwang and his two attendant bodhisattvas, was maintained and can be seen in various types of Hyeonwangdo. According to the books of Buddhist rituals, performance of Hyeonwangjae on the third day could exempt the deceased from the ten judgments that occur over three years, as well as from punishments for deeds in a former life. In other words, Hyeon wangjae was performed with the explicit purpose of allowing the deceased to be reborn in the Pure Land. This capacity for prompt salvation differentiated Hyeonwangjae from the belief in the ten kings, and likely explains why it became popularized as an independent ritual. The records show that the ritual eventually came to be performed in the main hall of a Buddhist temple. The area for performing the ritual included an altar for Hyeonwang, with Hyeonwangdo hung over the altar. Thus, Hyeonwangdo serve a unique function among Buddhist paintings, and such function required a space outside of the Judgment Hall. Hyeonwangdo from Seongbulsa Temple Objects and Hyeonwangtaeng wonmun from Hyeonwangdo from Seongbulsa Temple One of the most important extant Hyeonwangdo is Hyeonwangdo from Seongbulsa Temple, now in the collection of the National Museum of Korea. Measuring 103.7 cm in height and 85.7 cm in width (Fig. 1), it comprises three pieces of silk that have been joined together (width 27.5 cm, 33.0 cm, and 24.5 cm, respectively). The painting is not mounted, and the areas above and below the formal composition are painted turquoise. A red rectangular area at the bottom of the painting (Fig. 1b) bears an inscription that indicates that the painting was made in the twelfth month of 1798 at Buljiam Hermitage at Mt. Baegun, and then enshrined in Daeungjeon Hall of the “main temple.” The inscription does not name this “main temple,” but a later discovery (described below) revealed that it was Seongbulsa Temple. The inscription also gives the names of two patrons who were responsible for the production of the painting, Beomcheol (軓哲) and Choui (楚儀). Beomcheol is recorded as “Former Seungtong” (僧統, monk superintendent), and Choui as “Tongjeong” (通政, honorary title). Finally, the inscription reveals that the monks who produced the painting—Bohun (普訓) and Yeonggyeom (暎鎌)—were categorized as geumeo (金魚, lead monk painter) and dopyeonsu (都片手, assistant monk painter) respectively. In addition, Monks Daro (達悟) and Deongmin (德敏) assisted in the creation of the painting. Fig. 1b. Inscription from bottom of the painting, describing the production. 6.1 x 26.0 cm. During the process of restoring the painting, four quadrilateral holes were found in the scroll’s upper rod, leading to inner cavities. Inside these cavities were various objects, along with another written description of how the painting was produced. The objects included grains, seeds, metal, glass pieces, colored thread, and a Sanskrit dharani, written in red (Fig. 1c). The objects were wrapped in colored paper, and written on the paper were the Chinese characters for the five directions—east (東), west (西), south (南), north (北) and center (中) (Fig. 1d). Fig. 1c. Objects found enshrined within the rod of Hyeonwangdo from Seongbulsa Temple. (National Museum of Korea). Fig. 1d. Object and wrapping paper marked with Chinese character for “east” (東). (National Museum of Korea). Buddhist paintings and sculptures often had various objects enshrined within them. The procedure and details for enshrining Buddhist artifacts can be found in the Josanggyeong Sutra, which describes how to make the objects, how to install them, and how to perform the associated rituals. According to the sutra, a cylindrical container should be used to install objects inside a sculpture, while a quadrilateral container should be used for a painting (Josanggyeong Sutra 2006, 190-191).1 The sutra states that the objects to be installed include five jewels (五寶), five grains (五穀), five scents (五香), and five medicines (五藥). For a sculpture, these objects would be placed in a cylindrical container and then inserted into the statue, usually through a hole in the bottom or back. For a painting, the quadrilateral container was usually hung over the front or back of the painting. The containers were often lost in transport, so few such quadrilateral containers for installing objects in Buddhist paintings survive today (Fig. 6). Fig. 6. Rectangular container for enshrinement and objects. Jisimgwimyeongrye (至心歸命禮). (Yesan: Sudeoksa Temple Museum, 2004, p. 275). Notably, no quadrilateral containers were used to install the objects in the shaft of Hyeonwangdo from Seongbulsa Temple; instead, quadrilateral cavities were carved into the roller, and the objects were installed inside those openings. Thus far, this is the only known instance of such an installation. Thus, in addition to being an important example of the tradition of Hyeonwangdo, this painting is also an invaluable artifact for studying the installation of objects inside Buddhist paintings. The installation is also unusual because it lacks several of the objects that are listed in the Josanggyeong Sutra. The omitted items include various paintings of Sanskrit letters, such as Oryunjongjado (五輪種子圖), Jinsimjongjado (眞心種子圖), and Junjegujado (准提九字圖) (Fig. 7). In addition, the installation included simplified versions of the flag (幡), cover (蓋), and vajra (杵), each made from colored thread. Fig. 7. Oryunjongjado (五輪種子圖), Jinsimjongjado (眞心種子圖), and Junjegujado (准提九字圖). Josanggyeong Sutra translated by Taegyeong (태경). (Seoul: Unjusa Temple, 2006, pp. 190-191). Along with the objects, the shaft also contained a short document entitled Hyeonwangtaeng wonmun (現王幀願文), which tells how the painting was made and where it was enshrined (Fig. 1e): Early in the ninth month of 1798, in Daeungjeon Hall of Seongbulsa Temple at Mt. Baegun in Hamheung, Hamgyeong Province, a Buddhist sculpture was gilded. The golden color was resplendent, but unfortunately, it seemed that it would not be possible to produce a Hyeonwangdo to accompany the sculpture. Many people who worked on the gilding expressed their desire to produce a Hyeonwangdo, so they gathered the resources left over from the gilding, and then added more. They requested skilled artisans to produce the painting, and thus were able to produce this Hyeonwangdo, which was then hung in the hall. The primary patron was Monk Beomcheol (軓哲), and another patron was Monk Choui (楚儀). The painting was made by two other monks, Bohun (普訓) and Yeonggyeom (暎鎌). Hyangamdang Deokjang (香岩堂 德藏) supervised [the work] to ensure that the painting was properly made according to Buddhist formalities. After the painting was finished, the final ceremony was held to paint in the pupils of the eyes, and Seongyeo (性洳) recited the scripture at this ceremony. Gwanseol (寬說) and Yongundang Neunghwal (龍雲堂 能闊) worked to find more donors to contribute to the painting. Jonghan (宗閑) was in charge of offering food to Buddha, and Munyeol (文悅) cooked for the project. The entire project was supervised by Geugwon (克元). I pray that this good deed to which we have all contributed will allow our parents to sit on the nine-tiered lotus pedestal and eventually be reborn in the Pure Land. This prayer was installed inside the painting and enclosed on the 17th day of the 12th month of 1798. Fig. 1e. Hyeonwangtaeng wonmun document from Hyeonwangdo from Seongbulsa Temple. 36.0 x 22.2 cm. (National Museum of Korea). This document provides concrete information about the date and location of the painting, as well as important details about the duration of the project, the process of raising funds, and the sponsors. It clearly states that “skilled artisans” were requested, indicating that Monks Bohun and Yeonggyeom were trained painters from other temples. Interestingly, the document does not mention Daro and Deongmin, the two assistants who are mentioned in the inscription on the painting. They likely were monks from Seongbulsa Temple who assisted the two primary artists. The description also notes that both Beomcheol (the primary patron) and Geugwon (the general supervisor) were former Seungtong, while Jonghan (who offered food to Buddha) was a former Juji (住持, abbot). Based on this, the production of the painting was led by former Jongsa (宗師, eminent monk). Finally, the text includes a short prayer expressing the hope that the good deed of producing the Hyeonwangdo might enable the participants’ deceased parents to be reborn in the Pure Land. This wish illustrates how Hyeonwangdo were received among Buddhists at the time. Composition and Characteristics of Hyeonwangdo from Seongbulsa Temple In the painting, Hyeonwang is seated in the center, facing to the right, with a folding screen behind him (Fig. 1). Surrounded by his retainers, he is depicted larger and more prominently than the others. He wears a red outer robe over a blue inner robe, as well as a wonyugwan (遠遊冠), the traditional royal crown worn by Joseon kings on special occasions, over an inner headpiece of woven horsehair. Balanced atop the crown is a book of scriptures that has been tied into place to prevent it from falling. The depiction of Hyeonwang (Fig. 1f) resembles that of King Yama from Siwangdo. On the upper left side is a celestial child holding a silk fan and a maiden holding a rectangular fan. The other figures include a messenger of death holding a sword and various retainers holding a document, a scroll, a tablet, and a gold case for a seal. The judge, the clerk, and the child are known figures of the underworld, and their depiction resembles that seen in Siwangdo. However, in Siwangdo, these figures are typically shown in their proper roles of assisting in the judgment of the dead; in this Hyeonwangdo, however, they are arranged so as to direct the viewers’ gaze back to the central figure of Hyeonwang. Fig. 1f. Face of Hyeonwang. One notable figure is the celestial maiden at the upper right, who is actually given more emphasis than Daeryunseongwang and Jeollyunseongwang. She holds a tray on which appears a blue animal with a red jewel atop its head. This strange creature has the face of a fish, but also a tortoiseshell pattern of hexagons on its back (Fig. 1g). Such features are reminiscent of a creature called a bangyu (蚌魚), listed in the “Birds and Beasts” section of Sancai Tuhui (三才圖會, Collected Illustrations of the Three Realms, Fig. 8). Even so, the red jewel on the animal’s head suggests that it might in fact be a Makara (Sanskrit, 摩竭魚 in Chinese), a mythical creature that is often thought to have a cintamani, or Buddhist jewel, on its head. According to Hindu folklore, Makara are marine animals that combine aspects of fish, dolphin, crocodile, and elephant (Mochizuki Shinko 1954; Kim Lena 2001, 24-25; Han Jaewon 2006, 197-236). In Hindu art, Makara often appear as vehicles for Varuna (god of water and the oceans) and Ganga (river goddess). In addition, in the mural of Mogao Cave 158 in Dunhuang, Gansu province, China, a Heavenly Deity listening to Buddha’s teaching wears a hat shaped like a Makara. A similar Makara-shaped hat can be seen in the relief representing a Heavenly Deity carved into pagoda from the Unified Silla Period (668-935). According to an early Buddhist text entitled the Ekottara Agama, a person who falls in front of King Yama cannot be saved from hell, just as a “person caught in the mouth of a Makara cannot be saved.” This sentence equates falling into hell with being devoured by a Makara, which may explain why a Makara became a motif in Hyeonwangdo. Furthermore, the Makara is shown in a relatively submissive position, reinforcing the great power of Hyeonwang (Kim Wolun 1995, 342-346). Fig. 1g. Celestial maiden and makara. Fig. 8. Bangyu (蚌魚) in the “Birds and Beasts” section of Sancai Tuhui. Ming Dynasty, 1607. Today, there about 100 extant Hyeonwangdo, including those from Girimsa Temple (1718), Seonamsa Temple (1730), Samjangsa Temple (1739 or 1799), Sudasa Temple (1781), and Baekheungam, Eunhaesa Temple (1782) (Kim Yunhui 2011, 105-109). In the late eighteenth century, there seem to have been several styles of Hyeonwangdo. In particular, Hyeonwangdo from Tongdosa Temple (1775, Fig. 9) shares many key iconographic similarities with the one from Seongbulsa Temple. Hyeonwangdo from Tongdosa Temple is wider, but features the same composition, with Hyeonwang seated in the center with a folding screen behind him and retainers all around. These similarities are especially notable given the spatial and temporal separation between Hyeonwangdo from Tongdosa Temple (1775, Gyeongsang Province) and Hyeonwangdo from Seongbulsa Temple (1798, Hamgyeong Province). They show some differences in terms of design, pigment, and outlines, and the composition of the Seongbulsa painting is more simplified than that from Tongdosa. Despite these differences, they may have been produced from originals that were visually similar. Fig. 9. Hyeonwangdo from Tongdosa Temple. 1775. Color on silk, 108.0 x 122.0 cm. Buddhist Painting of Korea vol. 2 (한국의 불화 2). (Seoul: Research Institute of Sungbo Cultural Heritage, 1995, fig. 44). Notably, in Hyeonwangdo from Seongbulsa Temple, the judge and clerk are wearing headscarves, a motif borrowed from Taoist art. This style of depiction of Taoist figures was prevalent in the Buddhist art of the late Joseon period, as seen in books of prints, such as Xianfo qizong (仙佛奇蹤, Marvelous Traces of Transcendents and Buddhas) and Samjangbosaldo (三藏菩薩圖, Painting of Bodhisattva of the Tripiṭaka Teaching). The iconography and composition of Hyeonwangdo from Seongbulsa Temple were still being employed in the late nineteenth century, as demonstrated by the Hyeonwangdo from Yonghwasa Temple (Fig. 10) in Tongyeong (1875). The basic elements of the composition are remarkably similar, although differences can be seen. For example, Hyeonwang wears a gilt bronze crown, rather than a wonyugwan. In addition, the celestial maiden is now a celestial child, and the Makara has been replaced by an animal resembling a toad (Fig. 10a). A toad also appears in an earlier Hyeonwangdo (1750), now housed at Dongguk University Museum (Figs. 11 and 11a). In Taoism, toads are typically associated with a Taoist master named Liu Haichan (劉海蟾); “haichan” refers to a toad that lives in the ocean. For instance, Sancai Tuhui includes an illustration of Liu Haichan next to a three-legged toad (Cho Insoo 2000, 127-148). Thus, the Hyeonwangdo exhibit different appropriations of both the Makara, a Hindu symbol associated with water deities, and the toad, a symbol from Taoism. As yet, the precise reasons and meaning for the shifting use of the Makara and toad are not clear, but it demonstrates how unfamiliar motifs were adopted and simplified during the Joseon Dynasty. Fig. 10. Hyeonwangdo from Yonghwasa Temple. 1875. Color on silk, 96.0 x 78.5 cm. Buddhist Painting of Korea vol. 26 (한국의 불화 26). (Seoul: Research Institute of Sungbo Cultural Heritage, 2002, fig. 36). Fig. 10a. Celestial child and toad. Fig. 11. Hyeonwangdo from Dongguk University Museum. 1750. Color on silk, 114.0 x 103.0 cm. Buddhist Painting of Korea vol. 18 (한국의 불화 18). (Seoul: Research Institute of Sungbo Cultural Heritage, 1999, fig. 35). Fig. 11a. Celestial child and toad. After Hyeonwangdo from Seongbulsa Temple, more Hyeonwangdo began following the composition of Siwangdo. Although some Hyeonwangdo, such as the one now in the collection of the Horim Museum in Seoul (1854), maintained the original composition with the retainers arrayed around Hyeonwang, the more popular style appropriated the composition of Siwangdo but without the scenes of hell. Originally, unlike the Siwang (Ten Kings of the Underworld), who judged the souls of the deceased, Hyeonwang was believed to guide souls to the Pure Land. However, Hyeonwangdo from Gimnyongsa Temple (1803) (Fig. 12) and Hyeonwangdo from Silleuksa Temple (1809) show Hyeonwang holding a sword, which typically symbolizes a judge of hell. This new representation of Hyeonwang as the judge of hell indicates that the originality and independent meaning of the Hyeonwang belief had gradually weakened. Seongmun uibeom (釋門儀範, Guidebook for Buddhist Rituals), published in 1935, shows that a ritual for Hyeonwang had been formalized as one of the ritual procedures in Daeungjeon Hall, but apparently the unique iconography of Hyeonwangdo was not strong enough to endure. As a result, Hyeonwangdo were eventually re-incorporated into the other Buddhist paintings of the underworld. Fig. 12. Hyeonwangdo from Gimnyongsa Temple. 1803. Color on hemp, 128.0 x 98.0 cm. Buddhist Painting of Korea vol. 9 (한국의 불화 9). (Seoul: Research Institute of Sungbo Cultural Heritage, 1999, fig. 100). Reconstruction of Seongbulsa Temple In the course of my research, I discovered that historical records sometimes referred to Seongbulsa Temple as “Yongheungsa Temple.” Once thought to be separate temples, Seongbulsa Temple and Yongheungsa Temple are, in fact, different names for the same temple. This realization emerged from my study on Painting of Indra and Guardian Deities from Yongheungsa Temple (龍興寺帝釋神衆圖, 1798, Fig. 13), in Hamheung, Hamgyeong Province. I eventually learned that this painting was enshrined in the same hall as Hyeonwangdo from Seongbulsa Temple, which indicated that the two temples were in fact one and the same (National Museum of Korea 2000, 184-185). The confusion can be traced back to the “Taegosa Temple Law” (太古寺寺法), which was enforced in April 1941 by the Japanese Government-General of Korea, which attempted to unify the naming system of Korean temples. According to the Taegosa Temple Law, the temple was to be referred to as “Yongheungsa Temple,” even though it was commonly called Seongbulsa Temple in earlier historical records. Fig. 13. Painting of Indra and Guardian Deities from Yongheungsa Temple. 1798. Color on silk, 202.0 x 172.0 cm. (National Museum of Korea). Several books about artifacts that are now located in North Korea referred to Seongbulsa Temple as “Yongheungsa Temple,” in accordance with the adopted system. However, other sources, including Map of Hamheung (咸興府地圖, 1872), confirm that this temple was known as Seongbulsa Temple until the nineteenth century (National Research Institute of Cultural Heritage 1997, 242-251; Jogye Order of Korean Buddhism 2006, 208). In addition, the “Buddhist Temples” (佛宇) section from Records of Hamgyeong Province (咸鏡道邑誌, Fig. 14), also produced in 1872, states that Seongbulsa and Buljiam were located on Mt. Baegun, confirming that the name “Seongbulsa” was still being used in the nineteenth century. Fig. 14. Records of Hamgyeong Province. 1872 This confusion may be linked to the fact that Seongbulsa Temple was rebuilt several times over the centuries. These reconstructions are documented in an 1852 text entitled “Record of Multiple Reconstructions of Seongbulsa Temple at Mt. Baegun” (白雲山成佛寺還建舊基序, hereinafter, “Record of Reconstruction”), which can be found in Volume 2 of Historical Records of Temples in Korea (朝鮮寺刹史料), published by the Japanese Government-General of Korea in 1911. According to “Record of Reconstruction,” Seongbulsa Temple originated from a temple called Buljiam, which was designated as the temple of King Taejo (r. 1392-1398) before he ascended the throne. The record states that the temple was damaged by a major fire in 1578; it was relocated to the northwest of the original site and then rebuilt, but not until 1626. Then in 1792, the temple was destroyed by another fire, and was relocated and rebuilt a second time, about four kilometers below the fortress. Hyeonwangdo from Seongbulsa Temple was produced in 1798, while the temple was being relocated and rebuilt for the second time. Notably, Hyeonwangtaeng wonmun (the document installed in the painting) states that the production took place in Buljiam Hermitage, rather than in the main temple, likely because the main temple was still under construction. The gilded statue was not completed until six years after the 1792 fire. On the other hand, the two paintings of Hyeonwangdo from Seongbulsa Temple and Painting of Indra and Guardian Deities from Yongheungsa Temple were completed relatively quickly. According to the document installed in the rod of the scroll, the plan for producing the painting came about in September, when the gilding for the statue was finished, and the painting was finished in December. Thus, the entire process—including planning, raising funds, identifying artists, and creating the painting—took only about three months. As such, the combination of Hyeonwangtaeng wonmun and “Record of Reconstruction” enable us to create a timeline for the painting’s production, and moreover, to place those events into the historical context of Seongbulsa Temple. The rich details provided by these records were not conveyed by the inscription on the bottom of the painting. In particular, we owe a debt to the monks who collaborated to create the Hyeonwangdo from Seongbulsa Temple, for deciding to commemorate their creation by installing Hyeonwangtaeng wonmun inside the rod of the scroll. Thanks to them, we now know important details about the production of this painting in the late eighteenth century, as well as about the history of Seongbulsa Temple. Conclusion Buddhist paintings often served a ritual function, and thus reflect the religious practices of the time. In this context, I have examined the function and role of Hyeonwangdo from the late Joseon period, particularly their use in ritual and their enshrinement in Daeungjeon Hall. The fact that Hyeonwangdo were produced and enshrined independently from Siwangdo indicates the emergence of a new ritual that could not be performed with existing paintings. Details of the systemization and organization of this ritual were drawn from Hyeonwangcheong. According to Sutra on the Ten Kings, Shakyamuni predicted that King Yama would be reborn as Hyeonwang, and this prediction is reflected in the yesujae ritual of the late Joseon period. Notably, however, Book of Yesujae Formalities, which documents the origin of the yesujae ritual, simply refers to King Yama as one of the Ten Kings of the Underworld, and does not grant him any elevated status. The first record of an independent cult of King Yama, or Hyeonwang, comes from Hyeonwangcheong in Gwongong jebanmun, a liturgy of the ritual for Hyeonwang, published in 1574. Hyeonwangcheong was also published in Jebanmun (1624), Jakbeopgwigam (1827), Yojimmun (eighteenth century), Cheongmunnyojip (various versions from sixteenth through the eighteenth centuries), and Seongmun uibeom (1931). According to Hyeonwangcheong, Hyeonwang was believed to be able to save the soul of a deceased person on the third day after the person’s death, enabling the soul to be reborn in the Pure Land. Hyeonwang was accompanied by the attendant bodhisattvas Daeryunseongwang and Jeollyunseongwang, on the left and right respectively. The practice of producing Hyeonwangdo and enshrining them in the main hall of a temple became popular during the late Joseon period, reflecting the contemporaneous belief in Hyeonwang’s capacity for prompt salvation. This article examines the Hyeonwang faith and the production of Hyeonwangdo from various perspectives, focusing on Hyeonwangdo from Seongbulsa Temple. While the painting was being conserved, a small group of objects was discovered to have been installed in quadrilateral cavities carved into the upper rod of the scroll. These objects included a document entitled Hyeonwangtaeng wonmun, which revealed that the painting was originally enshrined in Seongbulsa Temple in Hamheung, Hamgyeong Province. According to Josanggyeong Sutra, which describes the proper procedure for installing objects within Buddhist artifacts, a quadrilateral container should be used to install objects in a painting. However, in the case of Hyeonwangdo from Seongbulsa Temple, quadrilateral cavities were carved into the rod instead, which is a type of installation not previously seen in other Buddhist painting. In addition, documentation on the reconstruction of Seongbulsa Temple found in “Record of Reconstruction” provides evidence that this painting was produced during a period of reconstruction, after the temple was badly damaged by a fire in 1792. Hyeonwangdo from Seongbulsa Temple features several interesting iconographic discrepancies from other Hyeonwangdo. For example, the attendants Daeryunseongwang and Jeollyunseongwang are not differentiated from Hyeonwang’s other retainers. In addition, a celestial maiden is shown holding an unusual blue creature with a red jewel on its head. The animal seems to be reminiscent of both Bangyu (蚌魚), as described in Sancai Tuhui, and Makara, a mythological creature from Hinduism. In the later Hyeonwangdo from Yongheungsa Temple (1875), this creature is depicted as a toad, often associated with the Taoist master Liu Haichan. This change seems to reflect the simplification and appropriation of unfamiliar motifs that occurred over time. After Hyeonwangdo from Seongbulsa Temple, Hyeonwangdo began to take on some of the characteristics of Siwangdo by featuring depictions of scenes of hell, or by representing Hyeonwang as a judge of hell, like the ten kings. Hyeonwangcheong established the ritual for Hyeonwang, which was conducted in Daeungjeon Hall in front of the altar for Hyeonwang. Over time, however, the originality and independent meaning of the Hyeonwang faith weakened, and Hyeonwangdo were eventually subsumed back into the overall genre of Buddhist paintings of the underworld.

The middle period of the Goryeo Dynasty extends from the reign of King Munjong (文宗, r. 1046-1083), when Goryeo shared a close relationship with the Liao (907-1125) and Song (960-1279) Dynasties of China, through the Goryeo Military Regime (1170-1270). The Buddhist sculpture of this 200-year period is especially notable, due to the emergence of new forms and styles, such as sculptures showing bodhisattvas in the relaxed posture of royal ease; increased popularity of wooden sculptures and the dry lacquer technique; the insertion of rock crystal for the eyes; and the use of a pliable substance called gamtang to depict hair and jewels. Some of these new characteristics represent the influence of Song Dynasty. Diplomatic relations between Goryeo and Song began during the reign of King Gwangjong (光宗, r. 949-975) and continued through 1030, when Goryeo sent a delegation of 293 envoys to Song. After that, relations seem to have been temporarily discontinued, as there are no known records of any exchanges until the sixth month of 1072 (26th year of King Munjong), when Song sent medical officials, a Royal Rescript, and official gifts to Goryeo, as reported in Goryeosa(高麗史, History of Goryeo). It would seem that Goryeo and Song began full diplomatic relations at this time, as accounts of further cultural exchanges between Goryeo and Northern Song (960-1127) can be found in several other historical documents, including Goryeosa and Xuanhe fengshi Gaoli tujing (宣和奉使高麗圖經, Illustrated Record of the Chinese Embassy to the Goryeo Court in the Xuanhe Era). Various records by monks also provide evidence of relations between Goryeo and Song, including the travel records of Monk Controller Uicheon (義天, 1055-1101) describing his travels to Bianjing (capital of Northern Song, present-day Kaifeng, in Henan province) and Mingzhou (present-day Ningbo, in Zhejiang province) in 1085. This paper examines characteristics of Buddhist sculpture from the mid-Goryeo period to ascertain the influence of the arts of the Song Dynasty. In addition, the route by which iconographic characteristics may have been imported from Song will be theorized through an examination of excavation sites, maritime routes, and beliefs in Avalokitesvara Bodhisattva. Influence of Song Buddhist Sculpture during the Mid-Goryeo Period Compared to the early Goryeo period (918-c. 1070), there are relatively few extant Buddhist sculptures from the mid-Goryeo period (c. 1070 to 1270); those with engraved inscriptions are especially scarce. Two of the more notable characteristics of these rare sculptures are the relaxed postures of the bodhisattvas and the decorative techniques used to render hair, jewels, and crowns. Relaxed Bodhisattva: Posture of Royal Ease The “posture of royal ease” refers to bodhisattvas in a stable, seated position, with the right arm extended out to rest casually on the upraised right knee, the left leg bent and tucked in against the body, and the vertical left arm firmly planted back on the pedestal. Three mid-Goryeo statues of bodhisattvas seated in this stance are currently known: a bronze seated bodhisattva from Goseongsa Temple in Gangjin, South Jeolla Province (Fig. 1); a gilt bronze seated bodhisattva from Daeheungsa Temple in Haenam, South Jeolla Province (Fig. 2); and a gilt bronze seated bodhisattva in the collection of the National Museum of Korea (Fig. 3). The same pose can also be seen in a bronze seated bodhisattva from Seocheon, South Chungcheong Province (Fig. 4), which is known only from a photograph in Joseon Gojeok dobo (朝鮮古蹟圖譜, Album of Ancient Sites and Monuments of Korea), a catalogue series published by the Japanese Government-General of Korea (1910-1945). Thus, there are currently four known examples of bodhisattva sculptures from the mid-Goryeo period depicting the bodhisattva in the posture of royal ease. Fig. 1. Seated bronze bodhisattva from Goseongsa Temple in Gangjin, South Jeolla Province. Late twelfth - early thirteenth century. Height: 51.0 cm. (Author’s photograph). Fig. 2. Seated gilt bronze bodhisattva from Daeheungsa Temple in Haenam, South Jeolla Province. Late twelfth - early thirteenth century. Height: 49.3 cm. (Research Institute of Buddhist Cultural Heritage). Fig. 3. Seated gilt bronze bodhisattva. Eleventh - twelfth century. (National Museum of Korea). Fig. 4. Seated bronze bodhisattva from Seocheon, South Chungcheong Province. Eleventh - twelfth century. Height: 15.0 cm. Originally printed in Joseon Gojeok dobo (朝鮮古蹟圖譜) and reprinted in Study of Late Goryeo Buddhist Sculpture (고려후기 불교조각 연구) by Jeong Eunwoo. (PhD dissertation, Hongik University, 2001, Fig. 3). Popular from the Goryeo into the early Joseon period, the royal ease posture is typically reserved for the Water-Moon Avalokiteshvara. In China, the earliest known image of a figure seated in this pose is a bodhisattva incised on a mirror dated 985 (currently housed in Seiryoji Temple, Kyoto, Fig. 5). This posture seems to be almost exclusively associated with the southern regions of China, and is rarely seen among Liao sculptures from the northern regions of China. In fact, there is only one known sculpture of a bodhisattva with this pose from the Liao Dynasty, along with a few from the Jin Dynasty (1115-1234), also in China’s northern regions, some of which are found in the Shihongsi Caves in Yanan, Shaanxi Province. The prevalence of this posture in sculptures from China’s southern regions is likely due to the area’s association with the Water-Moon Avalokiteshvara, who was believed to reside on Mt. Potalaka, a spiritual mountain believed to be located in the seas south of India. Images of this bodhisattva in the posture of royal ease continued to develop and spread into Korea and Japan. For example, in Japan, volume 22 of Besson zakki (別尊雜記, Assorted Notes on Individual Divinities), a work from the twelfth century, includes a painting of Avalokiteshvara Bodhisattva in this pose (Fig. 6), said to have been made by an artist from Quanzhou, China. Another Avalokiteshvara with the same pose can be seen in the Japanese painting titled Kegon kaie zenchishiki mandara (華嚴海會善知識曼多羅, The Spiritual Mentors of the Avatamsaka Ocean Assembly) from the thirteenth century. Images of bodhisattvas in this posture also seem to have been disseminated to Korea in the eleventh and twelfth centuries, likely through maritime routes from Quanzhou and Mingzhou. Today, there are approximately ten known Goryeo statues of bodhisattvas seated in the royal ease posture, as well as roughly ten mirrors with incised images of bodhisattvas assuming the same pose. Fig. 5. Bodhisattva incised on a mirror. 985. Seiryoji Temple, Kyoto. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, p. 24). Fig. 6. Avalokiteshvara bodhisattva from the 22nd volume of Assorted Notes on Individual Divinities (別尊雜記). Twelfth century. Ningbo, the Holy Place (聖 地寧波). (Nara: Nara National Museum, 2009, Fig. 94). Based on such evidence, the motif of bodhisattvas in the royal ease posture likely came to Goryeo during the Northern Song period. There are at least two early examples of statues in this posture from the Goryeo Dynasty. The first is a seated bronze bodhisattva from Seocheon, South Chungcheong Province (Fig. 4), which is known only from a photograph in Joseon Gojeok dobo (朝鮮古蹟圖譜, Album of Ancient Sites and Monuments of Korea). Based on the photo, the sculpture’s lean body and face indicate that it was probably produced in eleventh or twelfth century. The second example is the seated gilt bronze bodhisattva in the collection of the National Museum of Korea (Fig. 3) which has a lean body, slender waist, and a number of decorative details, including a tall leaf-shaped crown, Buddhist prayer beads in the right hand, bracelets on the wrists, and a robe with a double-pointed lower edge, like a swallow’s tail. The features of this sculpture are visually similar to those of a standing gilt bronze bodhisattva in the collection of Dongguk University Museum, as well as to a bodhisattva from a gilt bronze Vairocana triad at Yeongtapsa Temple, both of which are believed to have been produced during tenth or eleventh century. Based on those features, both the Seochon bodhisattva in Figure 4 and the seated gilt bronze bodhisattva in Figure 3 probably were produced around the same time, in the eleventh or twelfth century (Jeong Eunwoo 2006). The bronze seated bodhisattva from Goseongsa Temple in Gangjin (Fig. 1) and the gilt bronze seated bodhisattva from Daeheungsa Temple in Haenam (Fig. 2), both from South Jeolla Province, resemble each other but differ from the two bodhisattvas mentioned above. The Gangjin bodhisattva is 51.0 centimeters tall, making it the largest known Goryeo example of a bodhisattva seated in the posture of royal ease. Its head is relatively large in comparison to its body and it displays a very natural facial expression. The body is somewhat plump, with a voluminous chest and round belly. Around its neck is a simple yet elegant beaded necklace, highlighted by a small f lower pendant that appears in the center of the chest. At 49.3 centimeters, the gilt bronze seated bodhisattva from Haenam is almost as tall as the one from Gangjin. They are also very similar in terms of the overall form and decorative details. For instance, both figures have their hair arrayed in a similar fashion, with braids that loop around the ears and hang down across the shoulders. A number of the other details are also quite similar, including the voluminous chest, round belly, nipples, beaded necklace, the creases in the robe around the right knee, the big toes (either raised or lowered), and the way the left arm is planted gently on a fold from the robe. Some of these features—such as the necklace, hair, voluminous body, round belly, and loose clothing—can be found in other sculptures of bodhisattvas from the Song Dynasty. For example, the style of the necklace resembles that of the Samantabhadra Bodhisattva from the Hall of Sakyamuni in Qingliansi Temple (靑蓮寺) in Jincheng, Shanxi Province, which was produced in the Northern Song period (Fig. 7). A similar necklace can also be seen on the wooden seated bodhisattva currently housed in Kanagawa Prefectural Museum of Cultural History in Japan (Nara National Museum 2009, 93), which was produced during the Northern Song period. The long braids of hair draped over the shoulders are characteristic of Buddhist art of the Southern Song period (1127-1279), such as the wooden seated bodhisattva from Sennyuji Temple in Kyoto, Japan (Fig. 8). Notably, the bodhisattva in Figure 4 also features a voluminous body and a somewhat stubby nose, much like the bodhisattva from Goseongsa Temple (Fig. 1a). The round belly and loose clothing can also be seen in a pair of Song-Dynasty sculptures now housed in Japan: the wooden bodhisattva from Seiunji Temple, Kanagawa (Fig. 9) and the wooden seated bodhisattva from Hoonji Temple, Hyogo (Fig. 10). Based on these examples, the loose clothing, voluminous chest, and round belly have been categorized as the features of the late Northern Song and Southern Song periods. Fig. 7. Samantabhadra Bodhisattva. Northern Song. Hall of Sakyamuni of Qingliansi Temple in Jincheng, Shanxi Province. Buddhist Colored Statues from Shanxi (山西佛敎彩塑). (Beijing: Chinese Buddhism Association, 1991, Fig. 77). Fig. 1a. Side view of Fig. 1. (Author’s photograph). Fig. 8. Wooden seated Avalokiteshvara Bodhisattva from Sennyuji Temple, Kyoto, Japan. Southern Song Dynasty. Height: 113.8 cm. (Author’s photograph). Fig. 9. Wooden bodhisattva from Seiunji Temple, Kanagawa. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, Fig. 91). Fig. 10. Wooden bodhisattva from Hoonji Temple, Hyogo. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, Fig. 92). As previously mentioned, the two bodhisattvas from Seocheon and the National Museum of Korea (Figs. 3 and 4) are believed to have been produced in the eleventh or twelfth century. The other two bodhisattvas, from Goseongsa Temple in Gangjin and Daeheungsa Temple in Haenam (Figs. 1 and 2), are thought to date from the late twelfth or early thirteenth century. Goseongsa Temple is a branch temple of Daeheungsa Temple, and the two are located relatively close to one another. In Beomugo (梵宇攷, On Buddhist Temples, Past and Present), published in 1799, Goseongsa Temple is recorded as “Goseongam Temple.” In addition, a stele on Mt. Cheonbul provides further details about Hwaeomsa Temple (千佛山華嚴寺事蹟碑), stating that a monk named Yo Se (了世, 1163-1245) built Goseongsa Temple at the same time that he rebuilt Baengnyeonsa Temple, from 1211 through 1216. Given that the sculptures would have been enshrined immediately after the temples were constructed, these records provide us with approximate dates for the statues. Decorative Techniques The three most representative twelfth- and thirteenth-century Goryeo sculptures of the Buddha are a wooden seated Buddha in Gaeunsa Temple, Seoul (produced before 1274, originally located in Chukbongsa Temple, Asan, South Chungcheong Province); a wooden seated Amitabha Buddha in Gaesimsa Temple, Seosan, South Chungcheong Province (before 1280); and a dry-lacquer seated Buddha in Simhyangsa Temple, Naju, North Jeolla Province (Jeong Eunwoo 2008; Choe Seongeun 2008). The approximate date of the sculpture from Chukbongsa Temple is known from a written prayer installed inside the statue, which records that the statue was repaired in 1274. Similarly, a handwritten ink document found inside the statue from Gaesimsa Temple states that it was repaired in 1280. All three of these statues display features that are characteristic of sculptures of the Buddha from the Southern Song period (Mun Myeongdae 1996; Choe Seongeun 2008). The two most representative twelfth- and thirteenth-century Goryeo sculptures of bodhisattvas are a wooden seated Avalokiteshvara Bodhisattva from Bongjeongsa Temple (1199, 104.0 centimeters in height, Fig. 11) and one from Bogwangsa Temple (c. thirteenth century, 113.6 centimeters in height, Fig. 12), both in Andong, North Gyeongsang Province. Both of these sculptures display the formal stylistic features associated with the Southern Song period, such as a bold, dignified facial expression, an outer robe that covers the shoulders, and billowing sleeves that drape across the knees. The bodhisattva sculpture from Bongjeongsa Temple is made from pine wood; a plaque at the temple states that it was made in 1199. The bodhisattva statue from Bogwangsa Temple is made from juniper wood and is believed to date to the thirteenth century, based on devotional and dedicatory objects found inside the sculpture (Son Yeongmun 2009). Fig. 11. Wooden seated Avalokiteshvara Bodhisattva from Bongjeongsa Temple, Andong, North Gyeongsang Province. 1199. (Research Institute of Buddhist Cultural Heritage). Fig. 12. Wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple in Andong, North Gyeongsang Province. (Author’s photograph). Both of these bodhisattva sculptures feature several notable technical features that reflect the influence of the arts of the Song Dynasty. First, although both sculptures are made from wood, the hair and jewels of each are formed of a material called gamtang (Fig. 12a). The exact ingredients of gamtang are unknown, but based on the yellow residue that remains where the gamtang was applied (Fig. 12b), it is believed to be consist of a mixture of wax, pine resin, and other ingredients. Before gamtang was applied, the surface of a wooden statue was coated with lacquer and then covered with pieces of hemp cloth. Layers or pieces of gamtang were then attached to simulate hair or jewels. For example, strips of gamtang were laid across the head and then incised with fine lines to represent strands of hair (Fig. 12c). The jewels were molded separately from gamtang and appliqued; some of them have become detached and disappeared over the years. The strips of cloth used to affix the gamtang can be clearly seen in X-ray photos (Fig. 12d). X-ray photography has confirmed that gamtang was also used to form the hair of other sculptures made about the same time, including a dry-lacquer seated bodhisattva from Cheongnyangsa Temple (Fig. 13), a wooden seated Buddha from Suguksa Temple (Fig. 13a), and a late-Goryeo bodhisattva in Japan’s Okura Collection (Fig. 13b). Fig. 12a. Gamtang of wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple. (Author’s photograph). Fig. 12b. Head of wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple. (Author’s photograph). Fig. 12c. Back of wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple. (Author’s photograph). Fig. 12d. X-ray photos of wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple, showing gamtang technique. (Research Institute of Buddhist Cultural Heritage). Figs. 13, 13a, and 13b. X-ray photos of statues from Cheongnyangsa Temple, Suguksa Temple, and Okura Collection (respectively), showing gamtang technique. (Research Institute of Buddhist Cultural Heritage). A second important decorative feature of both statues is the insertion of rock crystal to represent the eyes, a technique that has also been found among other dry-lacquered and wooden Buddhist statues from the mid-Goryeo period (Figs. 14, 14a, and 14b). The earliest known example of a Goryeo sculpture with rock-crystal eyes is a wooden seated Avalokiteshvara Bodhisattva from Bongjeongsa Temple. Table 1 shows a total of ten such sculptures, dating from the mid-Goryeo to the early Joseon period, and mostly located in North Gyeongsang Province and South Jeolla Province. Fig. 14. Crystal eye insertion shown from the inside of the head of the statue from Cheongnyangsa Temple. (Cheongnyangsa Temple). Figs. 14a and 14b. X-ray photos of statues from Suguksa Temple and Okura Collection (respectively) showing how the crystal eyes were inserted. (Suguksa Temple and gift from a Japanese scholar). Table 1. Statues with crystal eyes. Name Date Location Height (cm) Notes 1 Wooden seated Avalokiteshvara Bodhisattva from Bongjeongsa Temple 1199 Andong, North Gyeongsang Province 104.0 Treasure #1620 2 Dry-lacquer seated bodhisattva from Cheongnyangsa Temple Goryeo Bonghwa, North Gyeongsang Province 90.0 3 Dry-lacquer seated Bhaisajyaguru from Cheongnyangsa Temple Goryeo Bonghwa, North Gyeongsang Province 92.5 4 Wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple Goryeo Andong, North Gyeongsang Province 113.6 Treasure #1571 5 Wooden dry-lacquer seated Buddha from Simhyangsa Temple Goryeo Naju, South Jeolla Province 136.0 Treasure #544 6 Wooden seated Buddha from Suguksa Temple Goryeo Seoul 104.0 Treasure #1580 7 Dry-lacquer seated Amitabha Buddha from Seonguksa Temple Goryeo Namwon, South Jeolla Province 132.0 Treasure #517 8 Dry-lacquer seated Vairocana Buddha from Bulhoesa Temple Goryeo Naju, South Jeolla Province 128.0 Treasure #545 9 Dry-lacquer seated Buddha from Jungnimsa Temple Joseon Naju, South Jeolla Province 116.0 10 Dry-lacquer seated Avalokiteshvara Bodhisattva from Naksansa Temple Joseon Yangyang, Gangwon Province 143.0 Treasure #362 11 Wooden seated Buddha from Gaesimsa Temple Goryeo (before 1280) Haemi, South Chungcheong Province 120.5 Notably, the pieces of rock crystal were not simply attached to the exterior surfaces of the statues. Rather, the hollow statues were made with empty eye sockets, into which the crystals were inserted via the interior. After the holes for the eye sockets were made and the outer surfaces smoothed down, the rock-crystal pieces were inserted into the interior and affixed into the sockets with a mixture of pine resin, crude lacquer, or wax. The crystals were often painted to represent the eye’s iris and pupil. X-ray photography of the dry-lacquer seated bodhisattva from Cheongnyangsa Temple shows that its rock-crystal eyes were held in place with a lacquered hemp cloth (Fig. 14). A similar method was likely used for the other statues as well, although this has not been confirmed. Irises and pupils were sometimes drawn or painted onto the exterior surface of the crystals, but there is no extant example of a statue showing this technique. Although there are some statues with painted pupils, the pupils were added sometime after the statue was made, so it is difficult to know the exact technique that was originally used. Some Korean Buddhist sculptures have stones other than rock crystal inserted to represent the eyes. The earliest example of such a sculpture is the stone standing bodhisattva from Gwanchoksa Temple in Buyeo, South Chungcheong Province, which features a black stone (perhaps obsidian or shale) that was smoothed and inserted from the outside (Fig. 15). The method of using rock crystal or another stone to simulate the eyes adds a much more naturalistic, lifelike quality than simply painting or carving the eyes, and thus is probably associated with the belief that the Buddha was a living person with a corporeal body. It is also related to the practice of installing objects inside a statue. Yi San (李㦃), a fourteenth-century Goryeo literati, wrote a text entitled “On Gold-Painting the Maitreya Triad from Geumjangsa Temple in Mt. Yongdu” (龍頭山 金藏寺 金堂主彌勒 三尊改金記), in which he described several steps involved in the restoration process: “...shaping clay, sculpting wood, gilding the triad, replacing the eyes with blue beads, replacing and renewing everything, including the crown, necklace, robe, etc…..” (Dongmunseon). Thus, it would seem that blue beads were sometimes used to make the eyes, indicating the prevalence of this practice during the Goryeo Dynasty. Fig. 15. Stone standing bodhisattva from Gwanchoksa Temple. Goryeo Dynasty. (Author's Photograph). In Japan, the eyes are sometimes represented with black ink or colored pigments, but some sculptures have eyes made of rock crystal or glass, which are held in place with bamboo pegs (Mitsumori Masashi and Okada Ken 1996, 105-106). Within China, there are very few extant wooden Buddhist sculptures from the Song Dynasty, and the history of this technique has not been the subject of much research, so it is difficult to know just when the technique first appeared. The only known instance is a sculpture of the Buddha from the Liao Dynasty, which has ceramic pieces inserted for the eyes, reminiscent of the monumental sculpture of the Buddha from Cave 20 of the Yungang Grottoes (near Datong, Shanxi province) of the Northern Wei Dynasty. In fact, the finest extant examples of wooden Buddhist sculptures from the Southern Song period are now housed in Japan, and they have been examined in detail. Based on evidence from extant Buddhist sculptures, it is believed that the gamtang technique and the use of rock crystal for the eyes were both imported into Goryeo from Song. For example, two sculptures from the Southern Song period that are currently in Japan—a wooden bodhisattva from Seiunji Temple in Kanagawa (Fig. 9) and a wooden seated Avalokiteshvara Bodhisattva from Sennyuji Temple in Kyoto (Figs. 8a and 8b)—have hair made with a material similar to gamtang. In particular, the hair of the sculpture from Sennyuji Temple closely resembles that of the ones from Goryeo made with conjoined strips of the pliable material. The bodhisattva from Seiunji Temple also has green rock crystal or green glass inserted from the inside the forehead, which is similar to Yi San’s description of using a colored stone or beads to represent the eyes. Thus, the extant examples would seem to indicate that the use of gamtang and rock-crystal eyes first appeared during the Goryeo Dynasty in Korea and during the Southern Song period in China. Nonetheless, more precise details about the differences and affinities between techniques and materials have yet to be determined. Fig. 8a. Head of Fig. 8. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, Fig. 87). Fig. 8b. Side view of Fig. 8. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, Fig. 87). Form and Technique of Crowns Two representative crowns for mid-Goryeo-period sculptures of bodhisattva can be seen in the gilt bronze seated bodhisattva from Daeheungsa Temple in Haenam (Fig. 2a) and the wooden seated Avalokiteshvara Bodhisattva from Bogwangsa Temple in Andong (Fig. 12e). Both the form and the technique for making these crowns is unique among Goryeo Buddhist sculptures. The gilt crown of the sculpture from Bogwangsa Temple is made primarily from copper (83.1%) and zinc (10.86%). It is extravagantly decorated with scroll and lotus designs and various inset jewels, including sardonyx and turquoise, though most of the jewels have been lost. The tiny Nirmana-Buddha was made separately and attached to the small wooden panel at the center of the crown. The front of the crown consists of three separate pieces that were joined together with rivets and then attached to the back of the crown. The lower rim (Fig. 12f) is decorated with a pattern of squares and several lotus blossoms, which were made separately and appliqued, exemplifying advanced metalworking techniques. The lotus blossoms are depicted as buds about to burst into full bloom, which is characteristic of lotus designs from the Goryeo Dynasty. Fig. 2a. Crown of Fig. 2. (Research Institute of Buddhist Cultural Heritage). Fig. 12e. Crown of wooden seated Avalokiteshvara Bodhisattva from Bogwangsa. (Research Institute of Buddhist Cultural Heritage). Fig. 12f. Lower rim of Fig. 12e. (Research Institute of Buddhist Cultural Heritage). Although no crown with this exact form has ever been found in China, the overall composition of the scroll and lotus design was a common motif from the Northern and into the Southern Song period. For instance, similar designs can be seen on the crowns of bodhisattvas sculpted in the grottoes of Anyue and Dazu in Sichuan Province. Notably, those crowns also feature an arrangement of lotus flowers on the front. In addition, the overall style of crown, with multiple pieces and points spreading outward (like a blooming f lower), is reminiscent of the crowns of two different Buddhist sculptures from China: the bodhisattva statue from the Hall of Three Great Bodhisattvas in Chongqingsi Temple in Zhangzi, Shanxi Province (1079) and the bronze Samantabhadra Bodhisattva statue from Wanniansi Temple on Mt. Emei, Sichuan Province (980) (Fig. 16). A similar crown also appears in the depiction of a bodhisattva in Pure Land of Amitabha from Chionin, Kyoto (1180, Fig. 17). Other crowns that feature a lower band decorated with the quadrilateral design include that of the Avalokiteshvara Bodhisattva in Cienyan (賜恩岩) Temple in Mt. Qingyuan, Quanzhou, and that of a stone bodhisattva from Nantiansi Temple, Jinjiang, Fujian Province (1216). In addition, the crown of a wooden seated Avalokiteshvara Bodhisattva from Sennyuji Temple, Kyoto (Fig. 8) has its lower rim decorated with circular motifs rather than lotus flowers. This evidence indicates that the style of crown of Goryeo bodhisattva sculptures reflects the influence of the Northern and Southern Song. Fig. 16. Bronze Samantabhadra Bodhisattva statue from the Hall of Amitabha Buddha in Wanniansi Temple on Mt. Emei, Sichuan Province. 980. Chinese Sculpture by Angela Falco Howard, Li Song, Wu Hong, Yang Hong. (New Haven: Yale University Press, 2006, plate 4.23). Fig. 17. Pure Land of Amitabha from Chionin, Kyoto. Southern Song Dynasty, 1180. Colors on silk, 150.5 x 92.0 cm. Ningbo, the Holy Place (聖地寧波). (Nara: Nara National Museum, 2009, Fig. 57). Very few extant bodhisattva statues of the mid-Goryeo period have crowns. However, related examples can be found among sculptures from the late Goryeo and early Joseon periods, including a seated bodhisattva from Gaesimsa Temple, a gilt bronze standing bodhisattva in the National Museum of Korea, a dry-lacquer seated Avalokiteshvara Bodhisattva from Naksansa Temple, and a copper crown currently located in Takuzudama Shrine in Tsushima Island, Japan (Fig. 18). All of these crowns were produced between the fourteenth and fifteenth centuries, and are comparable to the previously discussed crowns, in that they are formed by combining separate pieces and that they feature an openwork scroll design. Hence, the style and form of crowns that were likely introduced from the Song Dynasty during the mid-Goryeo period continued to be widely produced in the late Goryeo period. Fig. 18. Copper crown, Tsushima Island, Japan. Goryeo Dynasty. Art of Tsushima Island (對馬の 美術). (Fukuoka: West Japan Culture Association, 1978, plate 55). Selection of the Maritime Route and Iconography Most of the Goryeo Buddhist sculptures discussed thus far have been sculptures of Avalokiteshvara Bodhisattva. Figure 19 shows the sites where most of those Goryeo Avalokiteshvara sculptures were excavated and enshrined. Notably, almost all of these sites—including Gangjin and Haenam (South Jeolla Province), Seocheon and Haemi (South Chungcheong Province), and Gaeseong (Goryeo’s capital)—were prominent port cities that regularly received ships from Song. Two other excavation sites—Andong and Bonghwa (North Gyeongsang Province)—are located in an area of the Namhan and Nakdong Rivers, near the Joryeong Pass, which was the primary access route for ground traffic moving from the south to Hanyang (present-day Seoul) and Gaeseong. During the Goryeo Dynasty, this area was known to have many temples that were dedicated to deceased members of the royal family, including Bongjeongsa, Bogwangsa, and Yongmunsa Temples in Andong. Similar temples could also be found in Seocheon, Gangjin, and Haenam, all of which are located near the coast. Thus, all of these areas were located in proximity to maritime transportation routes, and all had various temples dedicated to deceased members of the royal family or the aristocracy. Fig. 19. Excavation or enshrinement sites of Goryeo Avalokiteshvara Bodhisattva statues. The fact that so many of these sculptures display aesthetic features from the Northern and Southern Song can be attributed to the location of the cities along maritime routes to China. During the early Goryeo period, there were at least two main routes to China. In the early Goryeo, the primary route went through Dengzhou (Shandong Province) and then on to Kaifeng (the capital of Northern Song, in Henan Province). This was the route traversed by the Great Monk Uicheon for his trip from Goryeo to Song in 1085. It is also known that a Goryeo official named Kim Je (金悌) traveled to Song via Dengzhou in 1071 in order to present a tribute prior to the resumption of full diplomatic relations in 1072. However, as evinced by a passage in Goryeosa1, a southern route was also in use around the same time, starting from the Yeseong River (Gaeseong), to Heuksando Island, and then passing through Mingzhou (present-day Ningbo, Zhejiang Province) and Quanzhou (Fujian Province), eventually arriving at Kaifeng (Fig. 20). According to Goryeosa, a Song merchant named Huang Shen delivered a diplomatic document to Fujian, which was located south of the aforementioned route through Shandong Province. As such, it would seem that the maritime routes through both Shandong and Fujian were in use at that time. Fig. 20. Maritime routes from Goryeo to Song. Some of the posts along the routes included Gangjin, Haenam, and Seocheon, which also happened to be the main production sites of Goryeo ceramics. In particular, during the Goryeo Dynasty, Jeolla Province served as the primary point of departure for routes crossing south of the Yellow Sea or East China Sea, which were the preferred ways into Hangzhou or Mingzhou (Zhejiang Province) (Yun Myeongcheol 2009, 223-225). Gangjin was not only a main port of Jeolla Province, but also provided access to other major ports, including Haenam, Jangheung, Yeongam, and Wando. As the center of the Goryeo ceramic industry and a crucial transportation hub, Gangjin was ideally suited to receive aesthetic influence from Song. Notably, Buddhist sculpture produced in Mingzhou made its way into Japan. For example, a wooden sculpture of a seated Avalokiteshvara Bodhisattva (Fig. 8) was brought from Bailiansi Temple in Mingzhou to Sennyuji Temple in Kyoto by a Japanese monk named Tankai (湛海). Records show that Tankai went to China in 1230 to commission the sculpture, which he then brought back to Japan in 1250. Furthermore, the record states that the statue was originally f lanked by standing sculptures of Skanda and Somachattra, and that the hall where it was enshrined was decorated like Mt. Potalaka, confirming that the depicted figure is the Water-Moon Avalokiteshvara. Another wooden bodhisattva from Hoonji Temple in Hyogo (Fig. 10) is known to have been made by a Mingzhou sculptor named Shin Ichiro (沈一郞), and brought to Japan in 1237 (Tsuda Tetsuei and Sarai Mai 2006, 48-54; Jeong Eunwoo 2008, 194-198). Furthermore, Wudenghuiyuan (五燈會元, Compendium of the Five Lamps), written by a Song monk named Puji (普濟, 1179-1253), states that someone from Goryeo went to Mingzhou and ordered the production of a sculpture of Avalokiteshvara Bodhisattva. Although he intended to take the statue back to Goryeo, he eventually had to stay in Kaiyuansi Temple in Mingzhou because of problems with his ship (Jang Dongik 2000, 413-414).2 This record demonstrates that Avalokiteshvara sculptures from Mingzhou were brought into Goryeo; it is believed that this was a fairly common practice in East Asia at the time. There is a similar story involving the Japanese monk Egaku (惠萼), who in 858 acquired a sculpture of Avalokiteshvara Bodhisattva from Mt. Wutai in Shanxi Province, and attempted to take it back to Japan. However, the ship that he took from Mingzhou became stranded, and Egaku interpreted this as a sign that the statue should not be taken to Japan. Thus, he enshrined the statue in a small temple near Mt. Putuo (Taniguchi Kosei 2009, 14-15). These stories demonstrate the overall popularity of faith in Avalokiteshvara Bodhisattva, the large-scale production of Avalokiteshvara sculptures during the Song Dynasty, and the tendency of people from Goryeo and Japan to acquire such sculptures and take them back to their home countries. It is crucial to note that both of these historical records document sculptures of Avalokiteshvara Bodhisattva, and that the extant sculptures in Korea and Japan are also Avalokiteshvara Bodhisattva. The wide production of Avalokiteshvara sculptures in Mingzhou can be attributed to the city’s close proximity to Mt. Putuo and Mt. Luojia, two holy sites of the Avalokiteshvara faith named after the legendary Mt. Potalaka. Mt. Putuo in particular has flourished as a sacred location amongst the followers of Avalokiteshvara since the Southern Song period. Mt. Putuo was believed to have strong auspicious powers, so people regularly went there to pray for a safe voyage, particularly when traveling by sea. Hence, people sailing to Korea or Japan likely would have stopped there to pray for a safe journey. In fact, sculptures of Avalokiteshvara were thought to have protective powers, and, as such, were often prominently placed on the prow of a ship. It seems likely that these beliefs were also prevalent in Goryeo, given the state’s reverence for both the Lotus Sutra and Avalokiteshvara Bodhisattva, and the wide production of Avalokiteshvara sculptures. During the Goryeo Dynasty, paintings of Avalokiteshvara were also widely used, as documented in Gujin huajian (古今畫鑑, Examination of Painting) by Tang Hou (湯垕, active in the early fourteenth century), and Tuhui baojian (圖繪寶鑑, Precious Mirror for Examining Painting) by Xia Wenyan (夏文彦, active in the late fourteenth century) of the Yuan Dynasty (1279-1368). According to those records, “Avalokiteshvara paintings of Goryeo are very elaborate. They are based on the masterful strokes of Weichi Yiseng (㷉遟乙僧, active in the seventh century), a monk painter of the Tang Dynasty (618-907), although they achieve their own subtlety and beauty”3 (Huang Binhong and Deng Shi 1986, 1389). These records reflect the popularity of Avalokiteshvara Bodhisattva and the technical refinement and development of Goryeo artists. Conclusion This paper has examined some features of Buddhist sculptures from the mid-Goryeo period that reflect the influence of the Song Dynasty, including bodhisattvas depicted in the posture of royal ease, the use of gamtang to form hair and jewels, the insertion of rock crystal for the eyes, and the composite crowns made from multiple thin pieces of metal. Through maritime routes, the Avalokiteshvara faith associated with Mt. Putuo in China was introduced to Goryeo, leading to the widespread production of sculptures of Avalokiteshvara Bodhisattva. Goryeo Buddhist sculpture is world-renowned for its beauty and refinement, which is directly related to the afore-mentioned features. For example, the application of gamtang on wooden or dry-lacquer surface allowed for a delicate, elaborate rendering of hair and jewels. The crowns, assembled from several thin sheets of copper, demonstrate amazing expertise and precision in metalworking. Glass, rock crystal, and other dazzling jewels were carefully inserted to represent the eyes. Used together, these techniques yielded the overall excellence for which Goryeo Buddhist sculpture is renowned. All of these stylistic and technical elements likely were introduced to Goryeo from Song via maritime routes. Of course, these new methods were refined by Goryeo artisans and combined with existing techniques to create a new aesthetic that was unique to Goryeo. Mid-Goryeo and Song Buddhist sculptures are alike in terms of their formal and stylistic elements, as well as their production techniques. Such affinities confirm that Goryeo Buddhist sculpture was not based solely on imported Buddhist sculptures or albums of Buddhist paintings. Goryeo sought sculptors from Song, and dispatched their own artisans to there to copy Buddhist murals. Goryeo sculptors actively and directly incorporated elements of Song Buddhist sculpture into their own culture, and this passion still reverberates today through the extant Buddhist sculptures of Goryeo.

Buddhist images, whether sculptures or paintings, only become objects of faith and worship through two rituals: jeoman (點眼), wherein the pupils of the Buddha’s eyes are painted in the final stages of creating a Buddhist image, and bokjang (腹藏), the ritualistic installation of various objects inside a Buddhist statue or painting. Virtually every country where Buddhism has been introduced has a tradition similar to the Korean bokjang tradition. In Korea, the objects to be installed in the image in the bokjang ritual are called bokjangmul (裝藏物 in China and 納入品 in Japan). The exact origins of the bokjang ritual are not known, but the oldest evidence of such practices are the sixth-century Buddhas of Bamiyan, in Afghanistan, which were destroyed by the Taliban in 2001. Such evidence suggests that, contrary to previous assumptions, the practice of installing sutras or sarira (beads, crystals, or relics believed to be corporeal) within Buddhist sculptures likely began much earlier than the eighth century. In Korea, the earliest known example of bokjang is thought to be an agalmatolite jar dated by inscription to the second year of Yeongtae (永泰, 766), which was discovered inside the pedestal of a stone statue of Vairocana Buddha. Bak Gyeongwon and Jeong Wongyeong (1983) argued that this jar served as a sarira case enshrining the Pure Light Dharani Sutra (無垢淨光大陀羅尼經), similar to the practice of installing sutras and miniature pagodas inside a pagoda. The first known use of the term bokjang comes from Naksan Gwaneum Byeongsong (洛山觀音幷頌), from volume 25 of Dongguk Isanggukjip (東國李相國集, Collected Works of Minister Yi of Korea, 1241) by Yi Gyubo (李奎報, 1168-1241). The term also appears in some votive inscriptions from Buddhist sculptures dating to the Goryeo Dynasty (918-1392). Based on this evidence, the practice of bokjang is thought to have been established as a Buddhist ritual during the Goryeo Dynasty. Thus far, a wide variety of bokjangmul have been discovered in Korea, including votive inscriptions, Buddhist texts, sutras, dharani (sacred words or short phrases), as well as textiles and traditional Korean paper. Other items used as bokjangmul include grains, scents, medicines, and cloth representations of parasols and vajras, all of which are typically found in groups of five, corresponding to the four cardinal directions and the center. Other constituent elements of the bokjang ritual in Korea are obobyeong (五寶甁, five wrappings for enshrining bokjangmul); huryeongtong (喉鈴筒, case for obobyeong); and hwangchopokja (黃綃幅子, yellow silk in which to wrap the huryeongtong). All three items were used to enshrine bokjangmul during both the Goryeo and Joseon (1392-1910) Dynasties according to a procedure described in the Josanggyeong Sutra (造像經). This differs from early bokjang practices in China and Japan, in which miniature representations of internal organs were installed inside sculptures of the Buddha. To date, studies of Korean bokjang have largely focused on the individual types of bokjangmul (e.g., sutras, dharani, textiles, votive inscriptions, etc.). Research has been limited by the difficulty of examining bokjangmul once they have been installed inside Buddhist sculptures or paintings, as well as by theft, damage, and the installment of new bokjangmul. This paper examines the evolution of the Korean bokjang tradition by comparing five existing versions of the Josanggyeong Sutra, with specific focus on references to the main elements of bokjang from the Goryeo and Joseon Dynasties (i.e., huryeongtong, obobyeong, and hwangchopokja). Those records are compared to surviving examples of huryeongtong, obobyeong, and hwangchopokja to illustrate how the procedures and contents of the bokjang ritual changed over time. Josanggyeong Sutra and Changes in Bokjangmul Comparison of bokjang terms from five versions of the Josanggyeong Sutra The Josanggyeong Sutra, a non-canonical sacred text of Buddhism, details various Buddhist rituals, including instructions for creating and enshrining bokjang, and the process of making sculptures of Buddhas and bodhisattvas. Today, five original versions of the Josanggyeong Sutra have survived in Korea: one each from Yongcheongsa Temple (1575), Neunggasa Temple (1697), Hwajangsa Temple (1720), Geumnyongsa Temple (1746), and Yujeomsa Temple (1824). There is also a transcription of an unknown original, which dates from the late Joseon period. The five versions and the transcribed copy have different titles, but the main contents are similar. For instance, all five versions include the following three chapters: “Daejangillamgyeong” (大藏一覽經), which compiles sutras related to the production of Buddhist statues; “Jebulbosal bokjangdan uisik” (諸佛菩薩腹藏壇儀式), which lists the proper bokjangmul for each of the cardinal directions and explains each object; and “Myogilsangdaegyowanggyeong” (妙吉祥大敎王經), which documents the procedures of the bokjang ritual. The Yujeomsa Temple version contains the most systematic description of the bokjang ritual. It is also the latest known original edition of the Josanggyeong Sutra, and thus the most significant for our purposes. The Yujeomsa Temple version seems to have been produced with the goal of systematizing the description of the bokjang ritual and standardizing the terminology. All five versions of the Josanggyeong Sutra list the bokjangmul for the cardinal directions and the center, but the names or types of bokjangmul vary among the versions. However, only the Yujeomsa Temple version uses the word huryeongtong, a term that appears to have been in widespread use even before the publication of this version, as the expression huryeongtong siju (喉鈴筒施主, donation of huryeongtong) is found in three copies of votive inscriptions in the sculptures of Three Buddhas in Beopjusa Temple (1626). Other versions, however, include a variety of names for bokjang containers, e.g. palyeopgae1 (八葉箇, huyreongtong with a cover decorated with eight-petal motifs), eunhaptong (銀合筒, covered silver bowl), and huryeong palyeopgae (喉鈴八葉箇). For example, the Neunggasa Temple version uses huryeong palyeoptong eunhap (喉鈴八葉筒銀合), and huryeong eunhap palyeoptong (喉鈴銀合八葉筒). The Yujeomsa Temple version introduces several new terms that replace or unify miscellaneous terms used in the other versions. For example, while the earlier four versions use the terms ogyeong (five mirrors) and osaekseung (五色䋲, string of five colors), the Yujeomsa Temple version replaces those terms with obanggyeong and osaekseon (五色線, threads of five colors), respectively. Also, the Yujeomsa Temple version uses the term obobyeong in place of obogwabyeong (五寶裹甁) and obogwabaek (五寶裹帛), which are used in the other versions. Josanggyeong Sutra and Changes in Bokjangmul In Korea, the central element of the bokjang ritual is the huryeongtong. Typically, the bokjangmul (e.g. vajras, flags of five colors, parasols, grains, medicines, etc.) are wrapped inside the obobyeong, which are then put into the huryeongtong. The huryeongtong is wrapped in hwangchopokja, and sabangju (四方呪, Sanskrit words for the four directions written on outer surface of huryeongtong) and sabanggyeong (四方鏡, mirrors for the four directions) are attached to the huryeongtong. Table 1 shows how to prepare a huryeongtong, using the huryeongtong found inside the Amitabha Buddha from the statue of Three Buddhas in Sudeoksa Temple (1639). Over time, the procedure for preparing and installing huryeongtong changed, with variations in the types of objects installed, the material of the huryeongtong, and the place of enshrinement. Despite these procedural changes, the general concept of huryeongtong was applied in all bokjang practices. Table 1. General procedure for preparing huryeongtong. 1. HURYEONGTONG: Huryeongtong are meant to represent the heart of a Buddhist sculpture, and they are also at the very heart of the bokjang tradition. All of the records about bokjang from Josanggyeong Sutra focus on the placement of the objects in obobyeong, which are then placed within a huryeongtong. According to the version of Josanggyeong Sutra from Neunggasa Temple, a huryeongtong was a silver bowl with a cover (銀盒), and such bowls have been found inside various Korean Buddhist sculptures. For example, a silver covered bowl, now in the collection of the Onyang Folk Museum (Fig. 1), was discovered inside a sculpture of Amitabha Buddha (1302), while another such bowl was found inside a giltbronze statue of the Medicine Buddha from Janggoksa Teample (1346). However, a gilt-bronze statue of Amitabha Buddha from Munsusa Temple (1346) was found to contain a wooden palyeoptong (Fig. 2), and another wooden container (Fig. 3) was found inside a late Goryeo wooden sculpture of Amitabha Buddah from Jaunsa Temple. Most huryeongtong of the Joseon Dynasty were metal, as evinced by those found inside a wooden statue of the Youthful Manjushri from Sangwonsa Temple (1466), sculptures of the Four Heavenly Kings from Songgwangsa Temple (1628), and a wooden sculpture of Three Buddhas from Sudeoksa Temple. However, a wooden statue of Mahasthamaprapta was found to contain a paper huryeongtong (1778, now in the collection of Onyang Folk Museum), and a bamboo huryeongtong of unknown origin has also been discovered. Fig. 1. Silver bowl from inside a statue of Amitabha Buddha, from Onyang Folk Museum. 1302. Silver. Height - 4.3 cm. (Author’s photograph). Fig. 2. Palyeoptong from inside a gilt-bronze statue of Amitabha Buddha at Munsusa Temple. 1346. Wood. Height - 7.0 cm. (Sudeoksa Temple Museum). Fig. 3. Wooden container from inside a wooden statue of Amitabha Buddah at Jaunsa Temple. Late Goryeo Dynasty. Wood. Height - 17.2 cm. (Sudeoksa Temple Museum). Regarding the shape of the huryeongtong, all five versions of the Josanggyeong Sutra refer to it as either a hap (盒), meaning a covered bowl, or a tong (筒), simply meaning a container, without more detailed explanation. Among huryeongtong from the Goryeo Dynasty, there are at least three in the form of a covered bowl: the aforementioned silver covered bowls from the statue of Amitabha Buddha (1302) and the gilt-bronze statue of Medicine Buddha in Janggoksa Temple (1346), as well as the wooden palyeoptong discovered inside the gilt-bronze statue of Amitabha Buddha from Munsusa Temple (1346). A wooden statue of Amitabha Buddha from Jaunsa Temple contained a covered wooden huryeongtong, but it was rectangular in shape, with curved sides (Fig. 3). Another huryeongtong of unknown origin, believed to be from the late Goryeo or early Joseon period, is a short, round wooden container with a flat bottom, but it differs from the previously mentioned covered bowls. A document called Mita bokjang ip mulsaek gi (彌陀腹藏入物色記) (Fig. 4)2 features a list of bokjangmul installed inside the gilt-bronze statue of Amitabha Buddha from Munsusa Temple. The list includes saridong (舍利同, sarira case) and palyeopdong (八葉同), which refer to containers in the form of covered bowls. Here, the word dong (同) seems to have the same meaning as tong (筒). Since these objects have the same shape as other objects called hap (盒), it is believed that, at that time, hap and tong (筒) were used interchangeably to refer to vessels of the same shape. Fig. 4. Mita bokjang ip mulsaek gi (彌陀腹藏入物色記) document from inside a gilt-bronze statue of Amitabha Buddha at Munsusa Temple. 1346. Paper. 35.1 × 35.9 cm. (Sudeoksa Temple Museum). According to the Yujeomsa Temple version of the Josanggyeong Sutra, the shape of huryeongtong of the Joseon Dynasty depended on where they were to be installed; round containers were placed inside Buddhist statues, while rectangular containers were used for Buddhist paintings.3 Large cylindrical huryeongtong with covers, a shape not seen in previous eras, were also common during the Joseon Dynasty. The huryeongtong found inside the statue of the Youthful Manjushri from Sangwonsa Temple (1466, Fig. 5) exemplifies this style of huryeongtong. Thus, the shape of huryeongtong changed over time, and the most notable changes involved the cover. Among huryeongtong from the Goryeo Dynasty, the covered silver bowl of Onyang Folk Museum (1302, Fig. 1) and the palyeoptong from Munsusa Temple (1346, Fig. 2) have flat covers with no openings. However, the rounded rectangular container from Jaunsa Temple (1346, Fig. 3) has a cover with an opening (1.5 cm in diameter) that is lined with a metal rim (Fig. 6). In contrast, the huryeongtong from Sangwonsa Temple (1466, Fig. 5), and the two huryeongtong discovered inside the wooden sculptures of Vairocana Buddha from Daejeokgwangjeon Hall (大寂光殿) and Beopbojeon Hall (法寶殿) of Haeinsa Temple (1490, Fig. 7) have covers with an extended open spout. Fig. 5. Huryeongtong from inside a statue of Young Manjushri at Sangwonsa Temple. 1466. Iron. Height - 10.8 cm. Catalogue of Woljeongsa Temple Museum (월정사성보박물관 도록). (Pyeongchang: Woljeongsa Temple Museum, 2002). Fig. 6. Upper view of Fig. 2, showing huhyeol. (Author’s photograph). Fig. 7. Huryeongtong from inside a wooden statue of Vairocana Buddha in Beopbojeon Hall of Haeinsa Temple. 1490. Copper. Height - 33.8 cm. Vow: Special Exhibition on Bokjangmul from the Statues of Vairocana Buddha in Haeinsa (해인사 비로자나불 복장유물 특별전-誓願). (Hapcheon: Haeinsa Temple Museum, 2008). This spout is assumed to have been called a huhyeol (喉穴), a term that first appears in the Yujeomsa Temple version of the Josanggyeong Sutra, which instructs “Make a huhyeol over the center of the palyeopgae” (八葉蓋, eight-petal cover).4 This record offers no additional details about the huhyeol, but several extant huryeongtong from the Joseon Dynasty have covers with an open spout on top of the palyeopgae. In some examples, such as the huryeongtong found inside the sculpture of Shakyamuni Buddha in Gounam of Woljeongsa Temple (1710, Fig. 8), the threads of five colors that were wound around the obobyeong were pulled through the opening in the center of the palyeopgae and up through the spout, so that they protruded to the outside. Thus, the term huhyeol probably refers to the extended spout. Fig. 8. Palyeop and huryeongtong from inside a statue of Shakyamuni Buddha in Gounam of Woljeongsa Temple. 1710. Copper. Catalogue of Woljeongsa Temple Museum (월정사성보박물관 도록). (Pyeongchang: Woljeongsa Temple Museum, 2002). As the style of huryeongtong changed from the covered bowls of the Goryeo Dynasty to the large cylinders of the Joseon Dynasty, the covers were also undergoing a transformation from flat lids to rounded covers with an extended spout. The first relevant records concerning this change do not appear until 1824, in the Yujeomsa Temple version of the Josanggyeong Sutra, but it is believed that cylindrical huryeongtong with huhyeol had become prevalent by the early Joseon period and remained in use through the remainder of the dynasty. Another important component of huryeongtong is the palyeopgae. The term palyeop first appeared as palyeopdong (八葉同) in a document entitled Geoan (擧案), found inside a statue of the Thousandarmed Avalokitesvara (1322, private collection), as well as in the Mita bokjang ip mulsaek gi document from Munsusa Temple (1346, Fig. 4). Of the five versions of the Josanggyeong Sutra, only the Yujeomsa Temple version states that the sarira case and huryeongtong should have a palyeopgae. But the record is quite vague about how exactly the palyeopgae should be used in relation to the sarira case and huryeongtong. The Yujeomsa Temple version includes palyeop daehongryeon jido (八葉大紅蓮之圖), a design of daehongryeon (large red lotus inscribed with Sanskrit words for directions inside eight petals) and states that palyeopgae, which is different from daehongryeon (大紅蓮), should serve as the cover for huryeongtong.5 The huryeongtong installed inside the hanging scroll painting of Shakyamuni Buddha from Seomamsa Temple (1753, Fig. 13) has a cover shaped like a large lotus with eight petals, but it was originally found wrapped in a palyeop daehongryeon jido (八葉大紅蓮之圖), as was the huryeongtong from inside the painting of Sakra in Seonamsa Temple. Based on this evidence, palyeopgae and daehongryeon are thought to be separate objects. If palyeopgae can be regarded as the cover of huryeongtong, then the palyeoptong from Munsusa Temple (1346, Fig. 2), which thus far has been considered a covered wooden bowl, would be the earliest example. On both the cover and body of this bowl, lotus seeds and a double layer of eight petals were painted in red, probably in order to express palyeopdong (八葉同), as mentioned in Mita bokjang ip mulsaek gi. The huryeongtong from Sangwonsa Temple (1466, Fig. 5) has a cover shaped like a lotus with eight petals, while another cylindrical huryeongtong of unknown origin has a lotus-shaped cover, as well as an attached gilt-bronze plate shaped like an eight-petal lotus (Fig. 9). In later periods, metal plates in the shape of an eight-petaled lotus were often found inside the huryeongtong, atop the obobyeong (Fig. 10). This suggests that palyeopgae, which were once attached to the outside of the huryeongtong, eventually became separate objects that were placed inside the huryeongtong. Fig. 9. Palyeop on huryeongtong. Early Joseon Dynasty. Copper. Height - 13.3 cm. (Sudeoksa Temple Museum). Fig. 10. Palyeop from inside a wooden statue of Three Buddhas at Sudeoksa Temple. 1639. Copper. (Author’s photograph). 2. OBOBYEONG: Obobyeong literally translates as “five treasure bottle[s],” which symbolize the four cardinal directions and the center. If the huryeongtong represent the heart of Buddhist sculptures, then the obobyeong are the chambers of the heart. According to the Josanggyeong Sutra, the obobyeong represent the five treasures of wisdom of Vairocana Buddha. Specifically, the north is represented by a glass bottle, the south by a jewel bottle, the east by an onyx bottle, the west by a coral bottle, and the center by a rock-crystal bottle. The Yujeomsa Temple version states that obobyeong could be made from papers of five colors, if the seven treasures were not available.6 In actuality, though, most surviving obobyeong from the Goryeo and Joseon Dynasties are made from cloth. Indeed, the other four versions of the Josanggyeong Sutra use the term obogwabaek (五寶裹帛, silk for wrapping five treasures), which probably explains why textiles were commonly used. The exact shape of obobyeong does not appear to have been standardized in the Goryeo Dynasty. For example, the obobyeong discovered inside the giltbronze statue of Amitabha Buddha from Munsusa Temple (1346, Fig. 11) were made from cloth shaped as a heart, square, circle, and triangle. Even so, there is still not enough evidence to make a conclusive statement about the shape of obobyeong from the Goryeo Dynasty. At the same time, almost all of the extant obobyeong from the Joseon Dynasty are square, as exemplified by the obobyeong found inside the wooden statues of Three Buddhas from Sudeoksa Temple (1639). There are a few exceptions, such as the obobyeong from the wooden statue of Vairocana Buddha at Haeinsa Temple, estimated to have been installed in 1490, which features textiles of various shapes, including a square, circle, and semicircle. Fig. 11. Obobyeong from inside a gilt-bronze statue of Amitabha Buddha at Munsusa Temple. 1346. Silk. (Sudeoksa Temple Museum). In the Goryeo and Joseon Dynasties, various types of bokjangmul were placed inside obobyeong and then tied with the threads of five colors. Obobyeong of the Goryeo Dynasty were tied like pouches (Fig. 12), as seen in the obobyeong from the statue of Amitabha Buddha (1302, Fig. 1) and the gilt-bronze statue of Amitabha Buddha from Munsusa Temple (1346, Fig. 2). Every version of the Josanggyeong Sutra, except the one from Yujeomsa Temple, uses the term obogwabyeong (五寶裹甁, five treasure wrapping bottles). Gwa (裹) means to wrap in a cloth, so this term likely applies to the pouch-shaped obobyeong of the Goryeo Dynasty. Fig. 12. Obobyeong from inside a gilt-bronze statue of Amitabha Buddha at Munsusa Temple. 1346. Silk. (Sudeoksa Temple Museum). In the Joseon Dynasty, obobyeong were made by placing bokjangmul in a square cloth, folding the cloth into a triangle, rolling the cloth into the form of a tube, and then tying it with the threads of five colors (Table 1). The transition from pouch-shaped obobyeong to cylindrical obobyeong corresponds to the shift from bowl-shaped to bottle-shaped huryeongtong. In the Joseon Dynasty, the five bobyeong were wrapped and tied together with the threads of five colors and then encased in the huryeongtong to make an obobyeong. However, it is not currently known if this same procedure was also used during the Goryeo Dynasty. Most obobyeong of the Goryeo and Joseon Dynasties have been found inside the huryeongtong, but there are some notable exceptions, such as the obobyeong found in the wooden statue of the Ten Kings of Hell from Munsusa Temple in Go chang (1653), which was tied outside the huryeongtong. Moreover, both the hanging scroll painting of Shakyamuni Buddha (1753, Fig. 13) and the wooden statues of the Three Buddhas in Donghwasa Temple (1727) had huryeongtong containing six metal cylinders (five obobyeong and one sarira case). This technique was likely used to include both sarira cases and obobyeong, in accordance with the Josanggyeong Sutra. Fig. 13. Huryeongtong from inside a hanging scroll painting of Shakyamuni Buddha at Seomam Temple. 1753. Copper. Height - 7.2 cm. (Author’s photograph). The Josanggyeong Sutra states that 75 objects should be enshrined in the obobyeong, including five medicines, five grains, five scents, five parasols, five vajras, five leaves, five flags of five colors etc.7 Obobyeong of the Goryeo Dynasty are often found to contain small amounts of different grains, medicines and scents, but by the Joseon Dynasty, the grains, medicine, and scents were joined by a variety of other objects, including parasols, vajras, and flags of five colors.8 Also, the wooden statue of Amitabha Buddha from Heukseoksa Temple (1458) contained four leaf-shaped textiles, which is unique among examples of bokjang. 3. HWANGCHOPOKJA: The Josanggyeong Sutra refers to a hwangchopokja, which is a piece of yellow cloth used to wrap huryeongtong. The Mita bokjang ip mulsaek gi document includes the term hwangpokja (黃幅子), which seems to be synonymous with hwangchopokja. The only detail that the Josanggyeong Sutra provides about the hwangpokja is that it should be about 45.0 cm in length. The gilt-bronze statue of Amitabha Buddha from Munsusa Temple (1346) contained five pieces of cloth, including the yellow cloth used to wrap the palyeoptong. Meanwhile, the wooden container found inside the wooden statue of Amitabha Buddha in Jaunsa Temple was wrapped with pieces of blue, yellow, and red cloth (Fig. 14). The covered bowl found inside the gilt-bronze statue of the Medicine Buddha from Janggoksa Temple (1346) was wrapped in cloths of five colors, as well as a yellow cloth. Thus, it seems that during the Goryeo Dynasty, the yellow hwangpokja was used together with cloths of other colors. In contrast, most huryeongtong from the Joseon Dynasty were wrapped with a single yellow cloth (Mun Myeongdae 1968, 366-367). There are some exceptions, such as the huryeongtong from the wooden statue of Buddha from Gosansa Temple in Hongseong (1543), which was wrapped in various colors of silk (i.e. orange, white, sky-blue, purple and yellow), as well as the threads of five colors. However, most of the evidence suggests that cloths of multiple colors, including hwangchopokja, were used to wrap huryeongtong of the Goryeo Dynasty, but only hwangchopokja were used in the Joseon Dynasty (Yi Seonyong 2005, 80). Fig. 14. Discovery of hwangchopokja from inside a wooden statue of Amitabha Buddah at Jaunsa Temple. Late Goryeo Dynasty. Silk. (Jaunsa Temple). The version of the Josanggyeong Sutra from Yujeomsa Temple states that after objects are enshrined within the huryeongtong, it should be wrapped with a hwangchopokja.9 The exact wrapping procedure from the Goryeo Dynasty is not known, but during the Joseon Dynasty, the huryeongtong was placed in the center of the hwangchopokja, which was then gathered around the huhyeol and tied with the threads of five colors. Some huryeongtong found inside Buddhist paintings were wrapped at the top and bottom with paper illustrations from the Josanggyeong Sutra, such as palyeop daehongryeon jido (八葉大紅蓮之圖), junjeguja cheonwonjido (准提九字天圓之圖, design of nine letters and letters of the four directions), or yeolgeumgang jibangjido (列金剛地方之圖, design of Vajra Buddha for the four directions), before finally being wrapped with hwangchopokja. However, in most cases, huryeongtong were wrapped with hwangchopokja in a way that was much simpler than the procedure described in the Josanggyeong Sutra. During the Joseon Dynasty, the four corners of the hwangchopokja would typically be marked with the four directions in black or red characters (Fig. 15). No such markings have been found on hwangchopokja from the Goryeo Dynasty, indicating that this practice accompanied the other changes in the bokjang tradition. Fig. 15. Hwangchopokja from inside a wooden statue of Avalokitesvara at Samgilam. 1726. Cotton. 18.9 × 17.6 cm. (Sudeoksa Temple Museum). Changes in the Bokjang Tradition From the Goryeo to the Joseon Dynasty, the shape of huryeongtong changed from a covered bowl into a cylinder, and the shape of obobyeong changed from a pouch into a cylindrical roll, to fit more easily into the huryeongtong. To better understand these changes, the purpose of bokjang and the interrelationship between the bokjang practices of each period must be explored. The practice of bokjang is closely tied to the worship of sarira. According to Josangyangdogyeong Sutra (造像量度經), which was published after the Josanggyeong Sutra, sarira were to be enshrined in the crown on the head of a Buddhist statue. In fact, sarira were discovered inside the urna (third eye) of the statue of the Thousand-armed Avalokitesvara in Jikido Hall (食堂, dining hall) of Toji Temple (東寺). The Buddhas of Bamiyan in Afghanistan were found to contain sarira, textiles, and fragments of the Sutra on Dependent Arising (緣起經) written in the characters of the fifth to sixth century (Fig. 16). Thus, it would seem that, early on, the practice of bokjang was not distinct from the practice of enshrining sarira and sutra (i.e. dharma sarira) inside a pagoda or sculpture. Fig. 16. Sariras and fragments of a sutra. Fifth - sixth century. Exhibition of Gandhara Art and Bamiyan Ruins (ガンダ-ラ美術とバ-ミヤン遺蹟展). (Shizuoka Prefecture: Shizuoka Shinbunsha, 2008). Indeed, the inscription on the agalmatolite jar found inside the pedestal of a stone statue of Vairocana Buddha (766) states that the Pure Light Dharani Sutra (無垢淨光大陀羅尼經) was originally enshrined within the jar. Overall, however, in Korea, the practice of enshrining sarira differs from the practice of bokjang. In early bokjang practices in China and Japan, representations of internal organs were installed inside statues of Buddha, in accordance with the belief that enshrining sarira inside a statue of Buddha animates the sculpture. This practice is exemplified by the standing image of Shakyamuni Buddha in Seiryoji Temple (淸凉寺) in Kyoto, Japan (Fig. 17). It is uncertain whether the practice of bokjang prior to the Goryeo Dynasty may have been driven by similar beliefs. But for the most part, the Korean tradition of bokjang, focusing on the use of huryeongtong and the belief in five directions, is more concerned with Buddhist rituals involving the five directions, rather than trying to animate Buddhist statues (Bak Gyeongwon and Jeong Wongyeong 1983, 57). Fig. 17. Representations of internal organs from inside the statue of Shakyamuni Buddha in Seiryoji Temple, Japan. 985. Silk. Arts of Japan (日本の美術) 86. With the publication of the Josanggyeong Sutra, the practice of bokjang of the Joseon Dynasty differentiated itself from that of the Goryeo Dynasty. The Josanggyeong Sutra described the concept of the five directions in great detail and explained the meaning of the various objects used as bokjangmul. Thus, the ritualistic aspect of bokjang can ostensibly be traced back to the Josanggyeong Sutra. In the Goryeo Dynasty, fewer objects were placed in the obobyeong, and fewer types of objects were used as bokjangmul. Also, the directions were marked by oryunjongja and jinsimjongja, but the sabangju and sabanggyeong were not used. The concept of five directions became much more important in the bokjang practice of the Joseon Dynasty, when five varieties of objects (e.g. grains, medicines, scents, etc.) were placed in the obobyeong, corresponding to the five directions. At the same time, new objects (e.g. vajras, parasols, flags, etc.) appeared, which had not been used in the Goryeo Dynasty. During the Goryeo Dynasty, the concept of the five directions may not have been clearly expressed inside the huryeongtong, but it was represented by wrapping the huryeongtong in various textiles symbolizing the five directions, including hwangchopokja. The concept of the five directions appears to have been standardized and established as an element of bokjang practice in the sixteenth century, as printed copies of the Josanggyeong Sutra became more widely distributed and available. Hence, in the Joseon Dynasty, the five directions came to be represented inside the huryeongtong, and the directions were simply marked on the hwangchopokja outside the huryeongtong. However, the bokjang practices of the Goryeo and Joseon Dynasties did not develop separately from one another. There are two known documents from the Goryeo Dynasty that list objects that may be used as bokjangmul: the Geoan (擧案) document found inside the statue of Thousand-armed Avalokitesvara (1322), and the Mita bokjang ip mulsaek gi document found inside the gilt-bronze statue of Amitabha Buddha in Munsusa Temple (1346). The objects listed in Geoan include palyeopdong, huryeong, lotus pedestal, geumjwaja (金座子, golden stand), seven treasures, five scents, and five medicines. Meanwhile, Mita bokjang ip mulsaek gi lists such objects as five scents, five medicines, five treasures, five medicinal herbs, huryeong (喉鈴), five grains, hwangpokja, saridong (舍利同), and palyeopdong (八葉同). Notably, all of the objects for bokjangmul that are described as the contents of obobyeong in the Josanggyeong Sutra (e.g., the seven treasures, five scents, and five medicines) can be found in one of these two lists. Thus, the bokjang practices of both the Goryeo and Joseon Dynasties are closely related, particularly since both focus on enshrining obobyeong inside a huryeongtong. In addition, the Neunggasa Temple version of the Josanggyeong Sutra (1697) includes the terms huryeong palyeoptong eunhap (喉鈴八葉筒銀盒) and huryeong eunhap palyeoptong (喉鈴銀盒八葉筒), corresponding to the earlier terms huryeong and palyeopdong. This would seem to suggest that huryeong and palyeopdong, which were separate items in the Goryeo Dynasty, were integrated in the Joseon Dynasty. Such change is supported by extant examples of bokjang from the two dynasties. Table 2 shows that huryeong, palyeop, tong (筒), and hap (盒), as mentioned in Geoan, Mita bokjang ip mulsaek gi, and the Josanggyeong Sutra, were initially separate items that were eventually merged into a single object known as huryeongtong. In particular, the item huryeong is no longer listed among the objects for bokjang of the Joseon Dynasty, but the Yujeomsa Temple version uses the single word huryeongtong, indicating the merging of items that had occurred by the Joseon Dynasty. Thus far, the earliest known version of the Josanggyeong Sutra is the Yongcheonsa Temple version (1575), but there may have been relevant sutras describing bokjang from the Goryeo Dynasty. Therefore, the differences between the bokjang practices of the Goryeo and Joseon Dynasties do not represent a rupture in the tradition, but instead reflect the unique characteristics of Korean Buddhist culture that developed over time based on the sutras. Table 2. Changes in the shape of tong (hap), huryeong, and palyeop. Conclusion This paper has examined changes in the tradition of bokjang during the course of the Goryeo and Joseon Dynasties, focusing on objects related to the huryeongtong. In earlier practices from the Goryeo Dynasty, huryeongtong were typically shaped like a covered bowl or bottle, and they had no definitive name. By the Joseon Dynasty, the shape had been standardized as a cylinder with an extended open spout. Furthermore, the palyeop, huryeong, and hap/tong, which had been separate items in the Goryeo Dynasty, were combined into the single object of huryeongtong by the early Joseon period. At the same time, the shape of the obobyeong changed from a pouch to a cylindrical form. Finally, the hwangchopokja, which had been composed of textiles of various colors during the Goryeo Dynasty, was replaced in the Joseon Dynasty by a single piece of yellow cloth marked with the five directions. It seems that practice of bokjang initially began in Korea as a way to enshrine Buddha’s relics and sutras, but with the publication of the Josanggyeong Sutra, the tradition gradually developed into a practice unique to Korea, incorporating the concepts of five directions and huryeongtong. Despite the numerous changes that took place within the bokjang tradition from the Goryeo to the Joseon period, the core element always remained the huryeongtong, demonstrating that the bokjang practices of the two dynasties were interconnected. Moreover, this congruity suggests that the bokjang practices of the Goryeo Dynasty were performed in accordance with sutras related with the Josanggyeong Sutra that was published in the Joseon Dynasty.

During the Bronze Age in the Korean Peninsula, a key transition from household-based production to craft specialization is believed to have occurred in the manufacture of such artifacts as red burnished pottery, ground stone daggers, and jade ornaments. This transition in the nature of craft production was accompanied by the introduction of a new class of objects made from bronze. Of these bronze objects, Liaoning-type bronze daggers and Korean-type bronze daggers are regarded as representative artifact types of the Korean Bronze Age (~1500 - 400 BCE), and thus are the central focus of this article. Bronze objects in general, and bronze daggers in particular, appear in the Korean archaeological record around the same time that significant transformations were taking place in the production of crafts. The nature of the production of bronze objects during the Korean Bronze Age has not yet been systematically researched, however, likely due to the limited number of extant bronze artifacts and the complete lack of known production sites on the Korean Peninsula. Due to the comparative lack of archaeological data, previous studies have primarily focused either on establishing typologies and schemes of classification or on interpreting the cultural meanings of the bronze objects based on the archaeological contexts from which they were recovered. More recently, researchers have begun to utilize metallurgical analysis to investigate the casting technology and the source of the bronze used on the Korean Peninsula (Choi Ju 1992; Choi Ju et al. 1996). The reconstruction of the bronze artifacts through experimental archaeology has also been attempted, albeit on a limited scale. These new research trends have sparked interest in the production and distribution of bronze objects in the ancient times, which in turn has led to discussions on the possibility that such bronze objects were manufactured by specialized craftspeople (Lee Youngmoon 1998; Cho Jinseon 2004). However, little has yet been done in terms of investigating the actual nature of bronze production in this period. Given that bronze daggers are the most common bronze artifacts from the period, this paper examines the characteristic features of bronze dagger production during the Korean Bronze Age. Specifically, Liaoning-type and Korean-type bronze daggers discovered in southern Korea are analyzed, focusing on the morphological attributes that are related to specialized production. Based on this analysis, the organization of the system of bronze production of this period is considered. Research Aims and Methodology Fig. 1. Liaoning-type (left) and Korean-type bronze dagger parts (revised from Cho Jinseon 1999, 84) Archaeological evidence of bronze production in the Korean Bronze Age is sparse, to say the least. The mining and smelting of copper ores, for example, are essential aspects of bronze production, but no archaeological evidence of such activities has yet been found in the Korean Peninsula. To date, the only evidence of bronze production comes from the numerous talc and sandstone molds that have been found (Cho Jinseon 2005). But the presence of such molds allows us to deduce that the entire process of bronze casting—finding the stone with which to make the molds, carving the stone molds, and using the molds to cast the bronze objects—indeed took place on the peninsula during the Bronze Age. In order to explicate further details about bronze production, the issue of production specialization is very important. By identifying how bronze production was organized (i.e., whether or not specialization had emerged), we can better understand the system for producing prestige goods, the relationship between craftspeople and elite members of society, and the overall nature of social complexity at the time. This paper examines and compares the degree of production specialization of Liaoning-type and Korean-type bronze daggers. Do the Korean-type bronze daggers evince a higher degree of production specialization? What is the nature and scale of the respective systems for producing Liaoning- and Korean-type bronze daggers? The morphological attributes of Liaoning- and Korean-type bronze daggers excavated from southern Korea are quantified and analyzed in order to determine if the degree of production specialization changed over time. This analysis allows us to consider the overall level of manufacturing of the bronze daggers and to examine the nature and development of the system for producing bronze daggers. Before we can consider the issue of craft specialization through an analysis of bronze daggers, we must first define “craft specialization,” and discuss how it can be identified in the archaeological record. Craft specialization generally is understood as a supra-household phenomenon, in which a specialist or part-time specialist produces more goods than he or she needs and then distributes the surplus in order to procure other desired goods or services. In other words, it entails a division of labor, which is why production specialization is regarded as a useful tool for identifying the degree of complexity of social and political organizations. Ideally, specialization in craft production should be confirmed through direct evidence of production obtained through the excavation of production sites. However, as such evidence is generally difficult to acquire, most studies, including this one, utilize indirect evidence of production (i.e. archaeological features and artifacts pertaining to production) to identify and examine craft specialization. It has been demonstrated that craft specialization can indeed be observed from secondary evidence (e.g. the nature of the exchange of products, standards of production technology, labor conditions, etc.), and not just from production facilities (Blackman et al. 1993; Costin and Hagstrum 1995). Relative to this, the “standardization hypothesis” has been proposed, positing that standardization—or at least a high degree of similarity in the use of raw materials, manufacturing techniques, and artifact form or decoration—can be regarded as an index of craft specialization (Costin 1991). The basic assumption of this hypothesis is that variability is reduced and products become more standardized as production becomes more specialized (Arnold 1993). At present, the standardization hypothesis has been used to analyze a wide range of archaeological materials, frequently in conjunction with ethnographic studies. The standardization hypothesis has not often been applied to the study of bronze artifacts, largely because, in metal casting, multiple items can be cast from a single mold, thereby distorting the actual degree of production standardization. However, the majority of bronze artifacts—and all of the bronze daggers—excavated from the Korean Peninsula appear to have been cast with different molds. Therefore, it is possible to use the standardization hypothesis to study the production specialization of Korean bronze daggers. Moreover, in this case, analysis of bronze daggers can also illustrate how the degree of production specialization changed over time. The earlier Liaoning-type bronze daggers and the later Korean-type bronze daggers differ only in terms of morphology, and thus share a common production process, which makes it valid to compare the degree of specialization in the production of the two dagger types. In order to apply the standardization hypothesis, a means for measuring the degree of standardization of a production group is required. Many archaeologists use the coefficient of variation (CV) as this tool. CV can be defined as the sample standard deviation divided by the sample mean, multiplied by 100 and expressed as a percentage. It allows comparisons between groups of different sizes or units, and therefore can be useful in pointing out the “relative” degree of variation in certain attributes of different products. Recently, various studies have demonstrated how CV can be used as an effective tool for measuring the degree of production standardization. For example, in ethnographic studies undertaken in Spain, Delhi, and Andhra Pradesh, Roux (2003) has suggested that ceramic assemblages presenting CV values below 3% may belong to high-scale (specialized) production, while ceramic assemblages presenting CV values from 6% to 9% indicate small- or very small-scale production. In addition, a CV value of 10% has been proposed as an arbitrary cut-off point between specialized and non-specialized production (e.g., VanPool and Leonard 2002), but such an empirical generalization is highly problematic. Therefore, this study compares the CV values of Liaoning-type and Korean-type bronze daggers in order to identify the “relative” degrees of standardization for these functionally equivalent types. This analysis then serves as the foundation for a discussion on production specialization in the Korean Bronze Age. Overview of the Data Set It has been suggested that comparative analyses to identify different degrees of standardization are most effective when they a limit the temporal or regional scope of the artifacts under study (Blackman et al. 1993). Therefore, in examining and comparing the degree of standardization evidenced in the bronze daggers of southern Korea, the data set must first be carefully chosen and defined. This requires a review of the previous studies undertaken on these bronze daggers, particularly those pertaining to chronology and typology, in order to select which Liaoning- and Korean-type bronze daggers are most appropriate for a comparative analysis in order to explore their differing degrees of production standardization. Liaoning-type bronze daggers have been found in almost every region of the Korean Peninsula, in archaeological contexts dating from the ninth to the eighth century BCE. Korean-type bronze daggers appear in the archaeological record at a later date, from the fifth to the fourth century BCE. Therefore, the Liaoning-type bronze dagger and Korean-type bronze dagger are considered to be the representative artifact types of the Middle and Late Bronze Age, respectively (Korean Archaeological Society 2012). Many researchers have attempted a type classification of Liaoning-type bronze daggers (Lee Youngmoon 1991; Kang Inuk. 2005; Miyamoto Kazuo 2002; Miyazato Osamu 2001). Of these, the framework for this study is provided by the chronological schemes established by Y.M. Lee and Kang Inuk, which are relatively detailed in nature. In order to identify the degree of standardization for Liaoning-type bronze daggers, the current analysis requires a single classificatory dagger type that was produced within a clearly defined period of time. Therefore, this study focuses on the 21 daggers belonging to the “Typical Ib type” of Lee’s chronological scheme, a data set that roughly corresponds to the “Yejeon-dong” type of Kang’s chronological scheme. These daggers date to the eighth century BCE and represent the majority of Liaoning-type bronze daggers that have been found in southern Korea (Fig. 2). Fig. 2. ‘Typical Ib’ Liaoning-type bronze daggers of southern Korea. Table 1. Measurements (in cm) of morphological attributes of Typical Ib-type Liaoning-type bronze daggers. No. Site Dagger body Dagger blade Dagger tang No. Site Dagger body Dagger blade Dagger tang Fig. 2-1 Jeokryang-dong ∙ 2.5 Fig. 2-12 Songguk-ri 33.4 30.1 3.3 Fig. 2-2 Jeokryang-dong ∙ 2.8 Fig. 2-13 Orim-dong ∙ 1.6 Fig. 2-3 Jeokryang-dong 35 32.5 2.5 Fig. 2-14 Birae-dong ∙ 2.65 Fig. 2-4 Jeokryang-dong 33 30 3.0 Fig. 2-15 Hwajang-dong ∙ 4.1 Fi)g. 2-5 Jeokryang-dong ∙ 2.8 Fig. 2-16 Undae-ri ∙ (4.2) Fig. 2-6 Usan-ri ∙ (4.0) Fig. 2-17 Wolnae-dong 35.7 33.05 2.65 Fig. 2-7 Usan-ri ∙ (1.4) Fig. 2-18 Wolnae-dong 43.4 40.4 3 Fig. 2-8 Jindong-ri ∙ 2.7 Fig. 2-19 Wolnae-dong (21.6) Fig. 2-9 Attributed to Muju 38.3 35 3.3 Fig. 2-20 Wolnae-dong (16.8) (14.4) 2.4 Fig. 2-10 Attributed to Muju 42 39.4 2.6 Fig. 2-21 Wolnae-dong ? ? 2.8 Fig. 2-11 Geumneung 40.8 38.1 2.7 Various studies have also attempted a type classification of Korean-type bronze daggers (Yoon Mubyeong 1966; Kim Jeonghak 1978; Lee Chongkyu 1982; Yi Kunmoo 2003; Cho Jinseon 2004). In most studies, the most important attribute for classifying daggers into different types is the length of the “edge on spine.” However, it has been argued that such attempts at classification fail to consider that some morphological attributes that were present at the time of casting (and are therefore indicative of the production process) may have been altered by the subsequent use of the dagger (Cho Jinseon 2004). Thus, to trace the development of Korean-type bronze daggers effectively, it is necessary to differentiate between those morphological attributes that represent the original casting process and those that may have been affected by usage (Fig. 3). As such, this study adopts the chronological scheme set forth by Cho Jinseon (2004). Specifically, the current analysis focuses on the 44 daggers belonging to the “Establishment phase” (around 300 BCE) and “Development I phase” (late third - early second century BCE) of Cho’s classificatory scheme. These Korean-type bronze daggers were found in most regions of southern Korea, so the data set is appropriate for a comparative analysis with the Liaoning-type bronze daggers (Fig. 4). Fig. 3. Development of Korean-type bronze daggers from the Korean Peninsula (from Cho Jinseon 2004, 96). Fig. 4. ‘Establishment-’ and ‘Development I-phase’ Korean-type bronze daggers of southern Korea. Table 2. Measurements (in cm) of morphological attributes of Establishment-phase Korean-type bronze daggers. No. Site Dagger body Dagger blade Dagger tang No. Site Dagger body Dagger blade Dagger tang Fig. 4-1 Dongseo-ri 1 31.8 29.3 2.5 Fig. 4-12 Namseong-ri d 26.2 24.5 1.7 Fig. 4-2 Dongseo-ri 2 36.8 33.5 3.3 Fig. 4-13 Namseong-ri e [30.2] [28.1] [2.1] Fig. 4-3 Dongseo-ri 3 25.1 22.5 2.6 Fig. 4-14 Namseong-ri f 21.2 18.9 2.3 Fig. 4-4 Dongseo-ri 4 28.1 24.7 3.4 Fig. 4-15 Namseong-ri g 20 17.7 2.3 Fig. 4-5 Dongseo-ri 5 26.9 24.6 2.3 Fig. 4-16 Namseong-ri h 29.5 26.5 3 Fig. 4-6 Dongseo-ri 6 31.8 28.3 3.5 Fig. 4-17 Yeonhwa-ri 1 20.4 18.4 2 Fig. 4-7 Dongseo-ri 7 31 28.5 2.5 Fig. 4-18 Yeonhwa-ri 3 26.8 24.3 2.5 Fig. 4-8 Dongseo-ri 8 27.5 24.8 2.7 Fig. 4-19 Yeonhwa-ri 4 30.8 27.9 2.9 Fig. 4-9 Namseong-ri a 31.5 28.9 2.6 Fig. 4-20 Goejeong-dong 32.4 29.9 2.5 Fig. 4-10 Namseong-ri b 28.5 25.7 2.8 Fig. 4-21 Cheongra 24 20.9 3.1 Fig. 4-11 Namseong-ri c 29.6 26.9 2.7 Fig. 4-22 Sumok-ri 24.3 22.5 1.8 Table 3. Measurements (in cm) of morphological attributes of Development I-phase Korean-type bronze daggers. No. Site Dagger body Dagger blade Dagger tang No. Site Dagger body Dagger blade Dagger tang Fig. 4-23 Sindong-ri (25.1) ∙ 2.3 Fig. 4-34 Gubong-ri 28.5 26.4 2.1 Fig. 4-24 Chopo-ri (25.8) ∙ 2.1 Fig. 4-35 Gubong-ri (28.5) ∙ 2.0 Fig. 4-25 Chopo-ri 32.7 30.3 2.4 Fig. 4-36 Gubong-ri 31.1 28.9 2.2 Fig. 4-26 Chopo-ri 31.3 29.2 2.1 Fig. 4-37 Gubong-ri ∙ ∙ 2.0 Fig. 4-27 Chopo-ri 28.5 26.1 2.4 Fig. 4-38 Gubong-ri 33.8 31.3 2.5 Fig. 4-28 Jangcheon-ri ∙ ∙ 2.2 Fig. 4-39 Gubong-ri ∙ ∙ 2.5 Fig. 4-29 Hyoja 4 31.8 29.2 2.6 Fig. 4-40 Gubong-ri 31.8 29.3 2.5 Fig. 4-30 Daegok-ri 24.7 23 1.7 Fig. 4-41 Gubong-ri 30.3 27.9 2.4 Fig. 4-31 Daegok-ri 29.5 27.3 2.2 Fig. 4-42 Gubong-ri 23.5 21.7 1.8 Fig. 4-32 Daegok-ri 32.8 30.7 2.1 Fig. 4-43 Gubong-ri (27) ∙ 2.7 Fig. 4-33 Tanbang-dong 32.5 30 2.5 Fig. 4-44 Gubong-ri (21.3) ∙ 2.0 Comparative Analysis of Liaoning-type and Korean-type Bronze Daggers Before undertaking the comparative analysis of Liaoning-type and Korean-type bronze daggers, we must first identify the morphological attributes common to both types that can be used to measure the standardization in the production process. For the purposes of this study, it is crucial to differentiate between those morphological attributes that are associated with spatial and diachronic variation, and those that ref lect production standardization. At present, the morphological attributes of Korean-type bronze daggers have been examined in more detail than those of Liaoning-type bronze daggers; thus, we will first consider the former to identify attributes that may be of most use in our analysis of production standardization. Those same morphological attributes will then be used to examine the degree of standardization evidenced in both Liaoning- and Korean-type bronze daggers. The following morphological attributes have been used to establish typologies of the Korean-type bronze dagger: blade form, blade length, spine height, cutting edge form, and tip thickness. Of these attributes, blade length can be seen to reflect the original casting process, since the length of a dagger’s blade does not change much, even after extended use. The blade form, on the other hand, might change drastically over time, as the blade is continuously ground in order to sharpen the cutting edge. Similarly, the height of the spine, the shape of the cutting edge, and the thickness of the tip also experience significant wear during use, and are thus considered to be attributes subject to alteration after casting (Cho Jinseon 2004). Hence, this study uses the length of the blade to measure production standardization. The other morphological attributes, which experience change during use, cannot be used to accurately evaluate the degree of production standardization. Some studies implementing the standardization hypothesis have also identified the entire length of the artifact as one of the most valid attributes of standardization (Blackman et al. 1993; Clark 1995). Therefore, this study will also examine the length of the dagger body (i.e., the blade and tang) as a morphological attribute that reflects standardization. Finally, the length of the tang—the area where the dagger body joins the handle—will also be considered. The tang is important because it represents a functional element, rather than a spatial or temporal element. To summarize, this study utilizes the length of the entire dagger body, the length of the dagger blade, and the length of the dagger tang as morphological attributes indicative of the degree of standardization for Korean-type bronze daggers. Typological studies of Liaoning-type bronze daggers have considered the following morphological attributes: cross-section shape of the dagger spine (i.e. flat or semi-circular); shape and position of the protrusion on the spine; shape of the lower section of the cutting edge; shape of the base of the blade; and length of the tip. However, this study compares Liaoning- and Korean-type bronze daggers, so morphological attributes common to both dagger types must be used as measures of standardization. Fortunately, the attributes previously identified as being indicative of standardization in the Korean-type bronze daggers—length of the entire body, length of the blade, and length of the tang—are equally relevant to the casting process of the Liaoning-type bronze dagger. Therefore, these three morphological attributes will allow us to compare the degree of standardization between Liaoning- and Korean-type bronze daggers. In addition, the length ratio of the body to the tang will also be examined, as it is deemed to be an important feature of standardization. In accordance with this approach, measurements were compiled of the length of the entire body, the blade, and the tang of the “Typical Ib type” Liaoning-type bronze daggers and the “Establishment-” and “Development I-phase” Korean-type bronze daggers. The resulting data are presented in the following three tables.1 Analysis of the Liaoning-type bronze daggers yielded the following results: the CV values are 0.11% for dagger body length, 0.12% for blade length, and 0.18% for tang length. The CV value for the body-to-tang length ratio is 0.16%, which is close to the CV value of the tang length. As for the Korean-type bronze daggers of the Establishment phase, the CV values are 0.15% for body length, 0.16% for blade length, 0.18% for tang length, and 0.17% for the ratio of the body-to-tang. Thus, the CV values for the Establishment-phase Korean-type bronze daggers are higher for the body length and blade length, and slightly higher for the ratio of the body-to-tang. The CV value of the tang length, on the other hand, is similar for the two dagger types. Finally, in the case of the Korean-type bronze daggers of the Development I phase, the CV values are 0.10% for body length, 0.10% for blade length, 0.12% for tang length, and 0.08% for the body-to-tang length ratio. Hence, all of the CV values for the Korean-type bronze daggers of the Development I phase are considerably lower than those of the other dagger types (Table 4). Fig. 5. Distribution of bronze daggers analyzed in this study. Table 4. Mean, SD, and CV of bronze daggers. Attribute No.of analyzed samples Mean (cm) Standard Deviation Coefficientof Variation (%) Liaoning-type bronze daggers Body length 8 37.7 4 0.11 Blade length 8 34.8 4.08 0.12 Tang length 17 2.79 0.51 0.18 Body-Tang length ratio 8 13.24 2.12 0.16 Korean-type bronze daggers (Establishment phase) Body length 21 27.82 4.31 0.15 Blade length 21 25.2 4.07 0.16 Tang length 21 2.62 0.48 0.18 Body-Tang length ratio 21 10.82 1.89 0.17 Korean-type bronze daggers (Development I phase) Body length 14 30.2 3.04 0.1 Blade length 14 27.95 2.83 0.1 Tang length 22 2.25 0.26 0.12 Body-Tang length ratio 14 13.49 1.04 0.08 Examining the results in more detail, we can see that, in the case of the Liaoning-type bronze daggers, the CV values of the body and blade are similar to the other dagger types, but the CV value of the tang is much higher. This is because the data set included eight “complete” daggers and nine “incomplete” daggers, which were either fragmented or had been recycled into smaller daggers or even arrowheads (see Fig. 3, No. 6, 7, 8, 9, 14, 15, 20). The tang length of the incomplete samples varied significantly (1.4 - 4.2cm), resulting in the high CV values for the tang length of the Liaoning-type bronze daggers. Consequently, it was confirmed that the selected CV values were valid measures of the degree of standardization only in the case of complete Liaoning-type bronze daggers. Additional analysis was therefore carried out on the eight complete samples, and a much lower CV value of 0.11% was obtained for the tang length. In contrast, a separate analysis of the nine incomplete daggers yielded a much higher CV value of 0.24% (Table 5). This indicates that, when the data were not skewed by the incomplete samples, the CV values of the body, blade, and tang lengths of Liaoning-type bronze daggers are indeed similar to those of the other dagger types. Table 5. Mean, SD, and CV of the tang length of Liaoning-type bronze daggers Attribute No.of analyzed samples Mean (cm) Standard Deviation Coefficientof Variation (%) Complete dagger Tang length 8 2.88 0.31 0.11 Fragmentary/recycled daggerbronze daggers (Establishment phase) Tang length 9 2.71 0.65 0.24 It has been suggested that the Korean-type bronze daggers can be divided into three size categories according to blade length (Cho Jinseon 2004, 68-70): small (18-21 cm), medium (22-26 cm), and large (27-34 cm). Therefore, additional analysis was carried out to evaluate this proposition. Notably, however, this study categorizes daggers by the length of the entire body, rather than the blade alone, since the former has been shown to be a more useful morphological attribute for assessing standardization. The body measurements of the Korean-type bronze daggers are presented in the histogram in Figure 6, which shows that the daggers can be usefully divided into the following three groups: small (less than 21.0 cm), medium (23.0-27.5 cm), large (greater than 27.5 cm). The results demonstrate that size was indeed an important feature in the production of the Korean-type bronze daggers. Fig. 6. Grouping of Korean-type bronze daggers according to the dagger length. Accordingly, the CV values of the Korean-type bronze daggers were re-calculated according to size group. In the case of the large dagger group (Table 6), the CV values of the Establishment-phase daggers were 0.08% for body length, 0.09% for blade length, 0.13% for tang length, and 0.14% for the body-to-tang length ratio. The CV values of the large Development I-phase daggers were 0.05% for body length, 0.06% for blade length, 0.08% for tang length, and 0.08% for the body- to-tang length ratio. These values are significantly lower than those obtained for the entire sample group of Korean-type bronze daggers (see Table 4), indicating that the CV values of middle- and large-size daggers were relatively higher. Therefore, it can be said that the large Korean-type bronze daggers exhibit a relatively high degree of morphological standardization. In fact, it appears that they were produced separately according to dagger size, and that the production process of the large Korean-type bronze daggers was relatively standardized. Table 6. Mean, SD, and CV of large Korean-type bronze daggers. Attribute No.of analyzed samples Mean (cm) Standard Deviation Coefficientof Variation (%) Establishmentphase daggers Body length 12 30.78 2.48 0.08 Blade length 12 27.91 2.46 0.09 Tang length 12 2.87 0.36 0.13 Body : Tang 12 10.87 1.47 0.14 Development I-phase daggers Body length 12 31.22 1.71 0.05 Blade length 12 28.88 1.66 0.06 Tang length 12 2.34 0.18 0.08 Body : Tang 12 13.44 1.07 0.08 The results of the foregoing analysis can be summarized in two graphs (Fig. 7). The graph on the left presents the CV values of all Liaoning-type and Korean-type bronze daggers, while the graph on the right features the CV values of the complete Liaoning-type bronze daggers andvv the large Korean-type bronze daggers. A number of observations can be made regarding the right graph. First, the CV value of the blade length decreases in the following order: Liaoning-type bronze dagger (0.12%) → Establishment-phase Korean-type bronze dagger (0.09%) → Development I-phase Korean-type bronze dagger (0.06%). This may reflect an increase in the degree of standardization over time. Second, all of the CV values of the Development I-phase daggers are noticeably less than the corresponding values for the Establishment-phase daggers. This can also be interpreted as a clear indication of an increase in the degree of standardization over time. Third, the CV value for tang length is noticeably higher for the Korean-type bronze daggers (particularly the Development I-phase daggers). It has been suggested that Liaoning- and Korean-type bronze daggers differed in terms of the means by which the handle was attached to the body (Oh Gangwon 2003, 12-13). This argument has yet to be confirmed, as no handle components of Liaoning-type bronze daggers have yet been found in southern Korea. However, if this indeed was the case, it might be regarded as a valid factor for explaining the pattern of CV values for tang length. Amongst Korean-type bronze daggers, the CV values for tang length are considerably lower with those from the later Development I-phase (0.08% as opposed to 0.13%). Again, this suggests increased standardization in the production of Korean-type bronze daggers on the Korean Peninsula. Finally, the CV values for the body-to-tang length ratio also demonstrate a similar pattern of decrease over time, further indicating an increase in the degree of production standardization. Fig. 7. Comparison of the CV values of Liaoning-type and Korean-type bronze daggers (left: CV values of all analyzed daggers, right: CV values of complete Liaoning-type daggers and large Korean-type bronze daggers). Conclusion This examination and comparative analysis of the CV values of the body length, blade length, tang length, and body-to-tang length ratio of Liaoning- and Korean-type bronze daggers from southern Korea has yielded meaningful results. Above all, the results indicate that the later Korean-type bronze daggers show a higher degree of standardization than the earlier Liaoning-type bronze daggers. Following other studies that have implemented the standardization hypothesis, the length of the body and blade (which best represent the entire dimensions of the artifact) were selected as key morphological attributes associated with production standardization. Notably, these two attributes also directly correlate with the total amount of molten bronze injected into the mold during the casting process. In other words, the standardization observed in these attributes is indicative of the standardization in the amount of bronze required to make a single dagger. In ancient times, bronze was an extremely valuable material that often had to be acquired over long distances, and standardization would have allowed craftspeople to make more precise estimates about the quantity of bronze required to manufacture the necessary number of daggers. The results also confirm that dagger size was taken into consideration from the earliest stages of production (i.e. when carving the mold). Then is it possible to regard this increase in the degree of standardization as an indicator of production specialization? As previously mentioned, standardization does not necessarily indicate the specialization of production. But in the case of the bronze daggers examined in this study, it seems highly likely that the production system of the daggers indeed became more specialized over time. There is clear evidence that the demand for and production of bronze items increased significantly in the Late Bronze Age. Therefore, it is reasonable to assume that the production system of bronze items, including daggers, became more specialized, and such specialization would have naturally led to product standardization. This specialization in the production system of bronze items would also have allowed for more diversity in the types of products being manufactured. The fact that the Korean-type bronze daggers examined in this study could be categorized into three different size groups might serve as evidence of such diversification,2 along with the results of some previous studies (e.g. Cho Jinseon 2004). The fact that different-sized Korean-type bronze daggers were made through separate processes of production indicates that size was an important feature in the production of bronze daggers, which in turn suggests that different-sized Korean-type bronze daggers served different functions. In particular, the high degree of morphological standardization shown for large Korean-type bronze daggers might be attributed to their use as actual weapons. Of course, additional studies must be carried out to establish the viability of this claim. It can also be noted that the diversification of bronze items that took place in the Late Bronze Age led to the production of new types of bronze items, such as bronze mirrors with coarse and fine design, and ritual implements in the form of pole-top bells and eight-branched bells. To summarize, the results of this study indicate that relative standardization was achieved in the production of Korean-type bronze daggers in the Late Bronze Age, and that the production system became more specialized around this time. This development can be attributed to the increase in the number, diversity, and technological standard of bronze objects produced in the Late Bronze Age in southern Korea. However, it must be stressed that this study simply aims to trace the general trend of diachronic change in the organization of production systems. Thus, rather than suggesting the existence of “specialized” bronze production in the Late Bronze Age, we interpret the results to indicate merely that bronze production in the Late Bronze Age, when the Korean-type bronze daggers were manufactured, was “relatively more specialized” than production in the Middle Bronze Age, when the Liaoning-type bronze daggers were manufactured,. Discussions of production specialization must be further developed by examining the molds used to cast bronze items, as well as by undertaking a detailed comparative analysis of a wider range of bronze artifacts. In addition, such research must be supplemented by metallurgical studies that examine production technology, and provenance studies based on the analysis of rare earth elements. When such a holistic approach is adopted, then we may begin to reconstruct the entire process of bronze production—from the procurement of raw materials to the manufacture of the actual product—and thus gain a deeper understanding of the production, distribution, and consumption of bronze items in the Korean Bronze Age.

The state of Goguryeo (37 BCE – 668 CE) was founded in the northwestern reaches of the Korean Peninsula, in the Jolbon region (present-day Benxi, Huanren County, Liaoning Province, China). Goguryeo had instituted a centralized system of government by the first century CE, at which time the rulers moved its capital to Gungnae Fortress (present-day Ji’an, Jilin Province, China); by the fourth century, it had emerged as one of the four great powers of East Asia. In 427, the capital was moved to Pyeongyang Fortress, and from this time on the state intensified its southern expansion. The pinnacle of Goguryeo’s power came in 475, when its forces conquered the Baekje capital of Hanseong (漢城). At its peak, Goguryeo territory stretched from the northeast region of present-day China to the central and southern regions of the Korean Peninsula. Historical records have been intensively studied in order to glean information about Goguryeo’s southern expansion following the move of its capital to Pyeongyang. Even so, written materials on the subject are exceedingly scarce, so the method of this expansion, not to mention its details, remain largely in question. Fortunately, beginning in the 1980s, a number of Goguryeo sites and artifacts discovered in southern Korea have been investigated, yielding sufficient archaeological material to shed new light on the topic. The 1979 discovery of the Jungwon Goguryeo Monument served as a catalyst for the study of Goguryeo’s southern expansion policy. In particular, the site generated considerable research on Goguryeo-Silla relations in the fifth century and on Goguryeo’ s expansion into the Chungju region. Beginning in the 1980s, tombs with structures similar to those of Goguryeo tombs were identified in southern Korea, but they could not be definitively characterized as Goguryeo tombs due to the lack of conclusive evidence. Thus, they were regarded merely as “Goguryeo-style” tombs, a perception that inhibited further research on them. Goguryeo sites and artifacts were unearthed within Mongchon Earthen Wall Fortress in 1988, leading to a renewed interest in and investigation of Goguryeo sites in southern Korea. Then, in 1994, some 20 Goguryeo forts were identified along Mt. Acha, on the northern banks of the Han River, and the excavation of those forts began in 1997. In 1999, a series of military fortification sites were also discovered in the Imjin-Hantan River region and in the Yangju basin. Goguryeo mountain fortress sites have also been identified and excavated in the Geum River region. More recently, since the 2000, an increasing number of “Goguryeo-style” tombs have yielded examples of Goguryeo pottery. As a result, many tombs once considered to be merely “Goguryeo-style” tombs have been reclassified as actual Goguryeo tombs. In addition, Goguryeo settlement sites have been identified in the vicinity of some of these tombs (Fig. 1). In light of these recent findings, this paper presents an overview of the research on the Goguryeo fortifications, tombs, and settlements of southern Korea, which can be regarded as evidence of Goguryeo’s southward expansion in the peninsula. Based on this archaeological evidence, the policy and actual process of Goguryeo’s southern expansion will also be examined. Fig. 1. Distribution of Goguryeo sites in southern Korea. (Author’s photograph). Goguryeo Fortification Sites in Southern Korea Location and Distribution of the Fortification Sites At present, approximately 50 Goguryeo fortification sites have been identified in southern Korea, and more sites are expected to be discovered through future investigations. The fortification sites, which comprise both f latland fortresses and mountain forts/fortresses, have been found in the following four areas: the Imjin-Hantan River region; the Yangju Basin and its environs; the lower reaches of the Han River and the area around Mt. Acha; and the Geum River region. The area including the Imjin and Hantan Rivers, which pass through Yeoncheon-gun and Paju-gun in Gyeonggi Province, is an intermediary point in approaching the Han River region from the north. Hence, approximately 20 fortification sites have been discovered along both sides of the rivers. Most of the fortifications in the Imjin River area face one another on the flatlands to the north and south of the river. Key examples of such flatland fortresses include Horogoro Fortress, Eundaeri Fortress, and Dangpo Fortress, all of which are situated on a triangular basalt plain formed by the Imjin River and its tributary, the Saet River. Two sides of the plain consist of high basalt cliffs overlooking the rivers, so a sturdy fortress wall was constructed along the plain’s third side. Situated between the Imjin and Han Rivers, the Yangju Basin was a strategically important point for transportation, and so a number of fortification sites were built there. To date, 28 mountain forts have been identified in the area, typically along mountain ridges, in lines that follow north-south axis. The forts were built on mountain ridges that offered good views of the transportation routes that passed through the flatlands below. Some of the forts were isolated, but most were built at intervals of 500 meters, forming a linear network of forts (Fig. 2). Fig. 2. Distribution of Goguryeo fortification sites in the Imjin-Hantan River region and the Yangju Basin. (Author’s photograph). Located in the central part of the Korean Peninsula, the Han River region sits on the main transportation route between the northern and southern parts of the peninsula. In addition, the wide plains of this region allow easy access to the West Sea. Due to this superior geographic location, control of the region played a key role in the geopolitics of the Three Kingdoms Period (57 BCE – 668 CE). A total of 21 forts have been identified on Mt. Acha and its environs, seven of which have been or are in the process of being excavated. The forts were placed at intervals of 400-500 meters, and they were most likely linked by structures such as wooden fences or stone walls (Fig. 3). Fig. 3. Distribution of Goguryeo fortification sites in the Mt. Acha region. (Institute for Archaeology and Environment of Korea University). Goguryeo fortifications identified in the Geum River region include Daemo Mountain Fortress in Jincheon, Namseonggol Mountain Fortress in Cheongwon, and the Wolpyeong-dong site in Daejeon. Excavations on sections of Namseonggol Mountain Fortress have provided relatively detailed information on the Goguryeo fortifications of this region. Notably, Namseonggol Mountain Fortress sits at the uppermost navigable point along the Geum River, a key point for transportation and logistics since ancient times. As large-scale mountain fortresses, the Goguryeo fortifications of the Geum River region are distinctive. In addition, unlike those in the three areas examined above, the fortification sites in the Geum River region are isolated, rather than arrayed in clusters. Structure of the Fortification Sites With the notable exceptions of Daemo Mountain Fortress and Namseonggol Mountain Fortress, both located in the Geum River region, virtually all of the Goguryeo fortification sites are small in scale (with diameters under 500 m). The majority of fortification sites consist of outer fortress walls made of stone or wood, with buildings within. During the early phases of occupation, wooden fences were often used as fortification, and stone walls were later added outside those fences. The method of constructing the fortress walls varies slightly from area to area. In the case of the flatland mountain fortresses of the Imjin River region (e.g., Horogoro Fortress, Eundaeri Fortress, and Dangpo Fortress), the walls consisted of a core of rammed clay, faced on either side with cut stone. (Fig. 4). The forts of the Han River region were constructed by erecting one or two stone walls about three to five meters in height on either side of the pre-existing wooden fence. The space between the structures was then filled with alternating layers of clay and saprolite, which were rammed or stamped. Fig. 4. Cross-section of the wall at Horogoru Fortress. (Korea Land Museum). The fortress walls of Namseonggol Mountain Fortress were built by erecting two parallel wooden fences and then packing clay into the space between. The entrances were the only parts of the fortress walls that were reinforced with stone. In order to improve their defenses, the majority of forts had additional chi (雉) structures—protruding sections of the walls that allowed soldiers to attack and defend from three sides. Some of the chi structures were built according to a different design, permitting them to be used as entrances (Fig. 5). Namseonggol Mountain Fortress also had an outer moat, which provided additional protection, as did Hongryeon Peak Fort 2. Fig. 5. Section of the wall and chi structure of Mt. Acha Fort 4. (Seoul National University Museum). These defensive sites included numerous buildings, in addition to water storage and drainage facilities (Fig. 6). Most of the buildings used as barracks stood above ground, although some pit dwellings have also been identified. These square or rectangular structures had gambrel roofs and walls made from a mixture of stone and clay. They usually measured five by seven meters, but some of the larger buildings could be up ten meters long and forty meters wide. The buildings contained L-shaped ondol facilities, which provided partial floor heating. This type of ondol consisted of a single flue through which heated air flowed. The flue was formed by long stone slabs, and the entire structure was plastered with a mixture of mud and straw. Fig. 6. Plan of Mt. Acha Fort 4. (National Research Institute of Cultural Heritage). All of the forts contained one or two water-storage facilities, consisting of square tanks that were carved into the weathered bedrock, with mud applied to the sides and bottom of the tanks for additional water-proofing. Wooden logs were stacked along each of the walls to form the inner face of the water tank, and the spaces between logs and walls were packed with mud. Although the overall size of the water tanks varies slightly, each held approximately the same volume of water, indicating that each fort was expected to store a similar amount of water. The forts also included a variety of other facilities (Fig. 7). At Mt. Acha Fort 4 and at Mt. Yongma Fort 2, for example, simple blacksmith workshops were found some distance away from the main cluster of buildings; these workshops were apparently used to make simple repairs to iron implements. Structures associated with firing pottery were identified at Hongryeon Peak Fort 2, while Mt. Acha Fort 3 yielded the remains of what appear to have been a blacksmith forge, underground storage facilities, and a mill structure, with the mortar and parts of the pestle in situ. These facilities yield insight into some of the activities undertaken by the soldiers stationed at the forts. In particular, the presence of iron farming tools suggests that the soldiers must have engaged in agricultural pursuits during times of peace. Fig. 7. Facilities located within the Goguryeo forts of Mt. Acha (① building, ② ondol heating facilities ③ water storage facilities, ④ water drainage facilities, ⑤ mill, ⑥ forge, ⑦ firing structure). (Institute for Archaeology and Environment of Korea University [② and ③], Seoul National University Museum [①, and ④-⑦]). Artifacts from the Fortification Sites Excavations carried out at the fortification sites have yielded numerous artifacts, comprising mainly pottery vessels and iron implements. The pottery can be divided into 24 types, including vessels for storage, cooking, serving and transport. Goguryeo pottery excavated from the sites around Mt. Acha typically dates to the early sixth century. The vessels were handmade using the coiling method, after which they were placed on a slow-turning potter’s wheel to smooth the surfaces. The pottery is typically brown, gray, or black in color (Fig. 8). Fig. 8. Goguryeo pottery from the Mt. Acha forts. (Seoul National University Museum). Large quantities of iron implements—including weapons, agricultural implements, and everyday tools and vessels—have also been discovered at Goguryeo fortification sites. The iron weapons can be divided into offensive weapons (e.g., swords, spears, axes, bows and arrows, etc.) and defensive weapons (e.g., armor, helmets, etc.). Excavations at Gueui-dong Fort, where an estimated ten soldiers were stationed, have uncovered around 3,000 arrows, for example, along with two swords, ten spears, and four battle-axes. This discovery reveals how the Goguryeo army was organized and allows us to estimate the number of weapons it may have possessed. As for defensive weapons, Mt. Acha Fort 4 yielded one iron helmet, while a suit of upper body lamellar armor was excavated at Mudeung-ri Fort 2 (Figs. 9 and 10). Fig. 9. Iron helmet from Mt. Acha Fort 4. (Seoul National University Museum). Fig. 10. Iron lamellar armor from Mudeung-ri Fort 2. (Seoul National University Museum). Goguryeo fortification sites have also yielded numerous iron agricultural implements and tools, such as ploughshares, shovels, spades, scythes, axes, and chisels, and various vessels, including iron jars and cooking pots. In addition, a limited quantity of iron horse-gear (e.g., bridle bits, bit stoppers, stirrups, and buckles) has been discovered, suggesting that (at least) the Goguryeo army officers stationed at the forts may have used horses to traverse the mountainous terrain. Numerous roof tiles, including some embellished with lotus patterns, were unearthed at Horogoru Fortress and Hongryeon Peak Fort 1 (Fig. 11), representing the first Goguryeo roof tiles to have been found in southern Korea. Given that Goguryeo roof tiles and roof-end tiles were used only for such public structures as palaces, temples, and administrative buildings, the presence of such roof tiles illustrates the significance of these defensive facilities. A ceramic drum sporting the two-character inscription “sanggo” (相鼓), or “small drum,” was also excavated at Horogoru Fortress; a similar ceramic drum was also found at Mt. Cheonbo Fort 2. In addition, the discovery of a gold earring at Namseonggol Mountain Fortress signals the importance of the site. Fig. 11. Roof-end tile decorated with lotus motif from Hongryeon Peak Fort 1. (Seoul National University Museum). Chronology of the Fortification Sites Given the period’s historical context, the Goguryeo fortification sites in southern Korea likely were first established around the late fourth century, just when Goguryeo intensified its policy of southern expansion. However, the morphological and technological characteristics of the pottery found at these sites indicate that the sites’ time of establishment and use differed according to region and type of fortification. Amongst the globular jars from the Han River region—the most intensively studied area—those from the forts of Mt. Acha and its environs date to the early sixth century. On the other hand, a long-necked jar with four handles from Mongchon Fortress dates to the late fifth century, based on its morphological attributes. These chronological interpretations are supported by the discovery at Hongryeon Peak Fort 2 of a pottery vessel bearing the inscription 庚子 (gyeong ja), a date that corresponds to 520 in the Gregorian calendar (Fig. 12). Based on this interpretation, the Goguryeo forces that conquered the Baekje capital in 475 would have remained at Mongchon Fortress until around 500. The forts located north of the Han River around Mt. Acha were established around 500 and were occupied until 551. Fig. 12. 庚子(gyeongja) inscription found on ceramic vessel from Hongryeon Peak Fort 2. (Institute for Archaeology and Environment of Korea University). Based on this chronological framework, the pottery from the fortification sites in southern Korea can be divided into two types based on their morphological and technical characteristics. The first type is distinguished by its black surface color, decoration of dots or concentric lines, and its base clay with sandy inclusions. The second type, which lacks decoration, has a brownish surface color and a fine base clay with no inclusions. According to chronological studies on the Goguryeo pottery of the Han River region, the first type dates from the mid-fifth to the early sixth century, while the second type dates from the early to mid-sixth century. Both types of pottery have been found at the fortification sites of the Imjin and Han River regions. In addition, some of the sites contain pottery from an even later date, suggesting that they were occupied from the mid-fifth until as late as the midseventh century, at least in some cases. The fortification sites of the Yangju Basin are limited in number, making it difficult to ascertain the dates when they were occupied. Nevertheless, pottery from Mt. Cheonbo Fort 2 indicates that they were built after the mid-fifth century. At the same time, fortification sites of the Geum River region, such as Namseong-gol Mountain Fortress, have yielded only pottery belonging to the first type, which might imply that they were occupied only for a limited period of time. In particular, based on the nature of the pottery and the historical context of the time, the Geum River sites seem to have been active from sometime after 475 to around 500. Goguryeo Tombs of Southern Korea Tomb Location and Arrangement At present, a total of 35 Goguryeo tombs have been identified in southern Korea. They have been found in the vicinity of large rivers (Imjin-Hantan Rivers, the upper reaches of the Bukhan River, and the upper reaches of the Namhan River), as well as at several inland locations (Seongnam, Yongin, Hwaseong). The inland tombs are found on flatland sites situated near rivers. Therefore, riverside locations can be identified as a common feature of these Goguryeo tombs. The tombs are found in isolation or in small groups of two to three. However, a single row of five tombs was discovered at the Dujeong-ri burial ground in Chungju, while the Gangnae-ri burial ground in Yeoncheon included a total of nine tombs, comprising three rows of three tombs each. Goguryeo tombs are also characterized by their placement on low flatlands and their construction in rows. Although only a limited number of tombs were constructed over a large area in southern Korea, the present evidence indicates that the tombs were located and arranged in regular patterns (Fig. 13). Fig. 13. Distribution of Goguryeo tombs. Tomb Structure and Construction Method Goguryeo tombs were covered with square earthen mounds (Fig. 14). The burial chamber was located either above ground or, more frequently, at a semi-subterranean level. The former type was built directly upon a prepared ground surface, but in some cases, an L-shaped pit was cut into an inclined ground surface prior to tomb construction. If the base soil was uneven, a fine clay mixture would be used to smooth it out. For the semi-subterranean burial chambers, stone walls were raised in the burial pit, and the space between the earth and the stone was filled with clay (Fig. 15). Fig. 14. Yeoncheon Sindap-ri tomb during excavation. (Korea Land Museum). Fig. 15. Plan of the Yeoncheon Sindap-ri tomb. (Korea Land Museum). The types of stones that were used differed according to the size and shape of the burial chamber, but flat stone slabs were generally used as the basic building material. The walls of the burial chamber were typically built by stacking rectangular slabs, with smaller stones used to fill the space between the slabs. Some of the chambers also had plastered walls. It was not always possible to conclusively identify structural elements of the tombs, but in some cases ceilings with parallel triangular corbels were confirmed (section highlighted in Fig. 15). Also known as a “laternendecke” or “lantern” ceiling, this structure was formed by stacking stone slabs atop the walls of the burial chamber, and then placing a large stone slab at each corner of the wall, forming four quadrants that constituted the ceiling. A passageway led to the burial chamber, which generally connected to the short wall on the right side of the burial chamber. The passageway had a flat ceiling comprising several stone slabs, and the section where the passageway met the walls of the burial chamber was covered with a large, rectangular piece of worked stone, which acted as a lintel for the doorway. The f loors of the tombs were often specially treated with either plaster or fire; in a few instances, the tomb floors were either partially or entirely paved with stones or slabs of cut stone. Some of the burial chambers contained stone platforms onto which the coffin of the deceased would have been placed, but most of the burial chambers with stone floors did not have a platform. The platforms were usually positioned on the left side of the burial chamber, but a tomb from the Cheojeon-ri burial ground was found to have two platforms, one on each side of the burial chamber. The platforms, which generally were around ten cm in height, were made by stacking either river stones or cut stones. In terms of dimensions, Goguryeo tombs of southern Korea can be divided into those with small burial chambers (length of long wall less than 2.2 m) and medium burial chambers (length of long wall greater than 2.2 m). In terms of the floor plan, the burial chambers are typically rectangular structures; some are longer than others, with the long walls more than 1.6 times longer than the short walls (Fig. 16). Fig. 16. Correlation between burial chamber length and the ratio of the chamber length to width of Goguryeo tombs. - Artifacts from the Tombs The Goguryeo tombs of southern Korea have not yielded many artifacts; moreover, the artifacts that have been unearthed do not show a great deal of variety. The most commonly found artifacts are items associated with the burial itself, such as coffin nails and rings. Other artifacts include pottery, jewelry, and other articles of personal adornment (e.g., gold and glass beads, silver bracelets, silver rings), and such iron implements as hairpins, knives, and rings. As mentioned, coffin nails represent the majority of the artifacts, with a total of 189 nails recovered from nine tombs. Some of the nails have round heads, some have square heads, and some are L-shaped, with no distinguishable head. The roundheaded nails are most common, while the square-headed and headless nails are few in number. Coffin rings have also been identified at Gangnae-ri Tombs 2, 4, 7, 8 and Bojeong-ri Tombs 1, 2. Each coffin ring consists of a base plate, connecting ring, and ring-shaped handle. The base plates of the coffin rings from the Bojeong-ri burial ground are circular, but those from the Gangnae-ri burial ground are flower-shaped and more decorative. Seven pottery vessels have been found in the tombs, belonging to four functional types: bottle, deep bowl, globular jar, and jar with elongated body. Four of the vessels are globular jars; each of the other three types is represented by a single artifact. In general, globular jars can be classified into various subtypes, while those found in the Goguryeo tombs represent just two types. Three of the discovered jars are C-type (with a long globular body and short protruding neck), while the fourth is B-type (with a slightly flattened body and short protruding neck). The B-type globular jar, found in Dujeong-ri Tomb 2, is similar to one from Gosan-dong Tomb 11 in Pyeongyang. However, the former has a slightly longer body and is widest at the center of the body, which would date it to the mid-fifth century, which is slightly earlier than the latter (Fig. 17). Fig. 17. Chronological scheme of globular jars from Goguryeo tombs. The C-type globular jars from Bojeong-dong Tomb 2 and Cheonggye-ri Tomb Chamber 1 are almost identical to those from Mongchon Fortress. The two former jars were decorated with the same pattern as many of the globular jars from Mongchon Fortress. Thus, based on the chronological scheme established for Mongchon Fortress, the jars can be dated to the late fifth century (c. 475 - 500), the same as the Mongchon jars. Another globular jar was found in Gangnae-ri Tomb 8, but it differs from the other two C-type globular jars in that the widest point of its body is located at the center of the vessel. Based on this difference, the C-type jar from Gangnae-ri is earlier than the other two, and can be dated to the mid-fifth century. Tombs 2 and 8 of the Gangnae-ri burial ground yielded silver bracelets and associated gold and glass beads. A silver ring was also found at Dujeong-ri Tomb 4. The gold beads were hollow and shaped like abacus beads, with holes through the center or sides, so that they could be threaded or suspended (Fig. 18). Fig. 18. Personal ornaments from the Goguryeo tombs of southern Korea. (Korea Institute of Heritage). Chronology of the Goguryeo Tombs of Southern Korea The Goguryeo tombs of southern Korea can be classified into small- and medium-sized tombs. They can also be classified into rectangular and elongated rectangular tombs, with the latter having long walls that are more than 1.6 times longer than the short walls. As can be seen in Figure 14, the nine tombs from the Gangnae-ri burial ground and the five tombs from the Dujeong-ri burial ground all belong to the elongated rectangular tomb type, as does Bangdong-ri Tomb 2. Of these elongated rectangular tombs, Dujeong-ri Tomb 2 and Gangnae-ri Tomb 8 have yielded C-type globular jars that can be dated, based on their shape, to the mid-fifth century. On the other hand, Bojeong-dong Tomb 2 and Cheongae-ri Tomb 1, representing the rectangular tomb type, have yielded C-type globular jars dating to the late fifth century (Fig. 12). It is also known that the burial chamber of Goguryeo stone tombs with horizontal entrances, which were initially rectangular, generally became more square-shaped over time. Based on this knowledge and the differences in the grave goods, the tombs with elongated rectangular burial chambers can be dated to the mid-fifth century, and those with rectangular burial chambers can be dated to the late fifth century. Various elements of the burial process seem to have been closely related, including the shape of the burial chamber, floor treatment, and stone plat-forms. Specifically, elongated rectangular burial chambers frequently had treated floors, but generally did not have stone platforms. Rectangular burial chambers, on the other hand, usually had stone platforms, but not treated floors. The type of floor treatment also differed according to the shape of the burial chamber. In all of the elongated rectangular burial chambers, the floors were treated with fire, but in the few instances of rectangular burial chambers with treated floors, the methods of the treatment varied. To summarize, in the mid-fifth century, coffins were often placed directly on the burial chamber floor, which had been treated with fire; in the late fifth century, stone platforms were used and the deceased was often placed directly (i.e., without a coffin) onto the stone platform. Southern Expansion of Goguryeo in the Fifth to Sixth Century According to such historical records as Samguk Sagi (History of the Three Kingdoms), the Goguryeo army advanced into the Chunggu region via the upper reaches of the Bukhan and Namhan Rivers in the late fourth century. In 396, King Gwanggaeto the Great (374–413; r. 391–413) attacked the Baekje capital of Hanseong, which resulted in the surrender of King Ashin. It is said that King Gwanggaeto the Great then returned home victorious, having annexed 52 walled towns and 700 villages north of the Han River. In 475, Goguryeo’s King Jangsu attacked Hanseong once again, killing King Gaero and forcing Baekje to move its capital south to Ungjin. According to Samguk Sagi, King Jangsu and the Goguryeo army returned home after Hanseong fell. Because the historical situation was not clearly recorded, there has been much debate over who controlled the Han River region, where the Baekje capital was originally located, after 475. Fortunately, the discovery of Goguryeo remains at Mongchon Fortress has shed some light on this matter. The archaeological evidence clearly shows that some Goguryeo soldiers remained at Mongchon Fortress after King Jangsu returned home with his army. Furthermore, sites such as Daemo Mountain Fortress, Namseonggol Mountain Fortress, and the site in Wolpyeong-dong all date to the late fifth century—similarly to Mongchon Fortress—which suggests that Goguryeo continued its southern expansion into the Geum River region, with Mongchon Fortress as its base. In addition, the irregular shape of the ground-level building at Mongchon Fortress, along with the discovery of ceramic ritual vessels, suggests that the commanding officer at the site was a person of considerable rank. Based on such evidence, the fortress seems to have been a central component of the state’s southern expansion policy. Moreover, excavations at Namseonggol Mountain Fortress indicate that it also may have functioned as a base fortress, given its scale and the contents of the site. Thus far, excavations have yet to be carried out at Daemo Mountain Fortress, so it is not known whether it also might have served a similar function. Therefore, Goguryeo’s attempts to expand its borders south of the Han River appear to have been carried out from these base fortresses. Even though the Goguryeo tombs of southern Korea are distributed over a wide area, they demonstrate a high degree of standardization in terms of their structure. The artifacts found within the tombs indicate that they were constructed by a Goguryeo population over a relatively short period of time, from the mid- to the late fifth century. Thus far, only a few tombs have yielded artifacts that can provide information about the deceased; Gangnae-ri Tombs 2 and 8 contained gold beads, silver bracelets, and glass beads; a silver ring was found at Dujeong-ri Tomb 4. The tombs containing these accessories are relatively large in size, suggesting that individuals of high status were buried in the Goguryeo tombs of southern Korea. In addition to fortifed sites and tombs, Goguryeo settlements and other sites yielding Goguryeo pottery have also recently been discovered. Most such sites were later reused by Silla populations, however, so it is difficult to identify with certitude archaeo-logical features or cultural strata that can be clearly assigned to Goguryeo communities. Even so, Goguryeo pottery is present at these sites, and settlement sites containing similarly shaped and decorated vessels can be dated to the late fifth century (Fig. 19). Fig. 19. Goguryeo pottery from settlement sites in southern Korea. Notably, these settlement sites have a characteristic pattern of distribution: the Udu-dong site in Chuncheon is located near the Mancheon-ri burial ground; the sites of Cheoljeong-ri and Yeongnae-ri in Hongcheon are located near the Cheoljeong-ri and Yeongnae-ri burial grounds; the Tappyeong-ri site in Chungju is located near both the Dujeong-ri burial ground and the Jungwon Goguryeo Monument; the Mabuk-dong site in Yongin is located near the Bojeong-ri burial ground. In addition, a great number of fortification sites have been found in the vicinity of the Sindap-ri and Gangnae-ri burial grounds, along with the Joowol-ri site in Paju. Based on this distribution pattern, the tombs at these burial grounds seem likely to have contained the remains of Goguryeo people who lived in the nearby settlements or fortification sites. Therefore, it seems reasonable to assume that the individuals buried in the Goguryeo tombs were not high-ranking officers who had fallen in battle, but elite members of Goguryeo communities that had long been established in the region. In conjunction with the distribution of Goguryeo fortification sites, this interpretation indicates that, although King Jangsu returned to Pyeongyang Fortress in 475 after attacking Hanseong, Goguryeo maintained control over the captured areas. How far did Goguryeo expand its southern boundary following the fall of Hanseong, and how was this southern expansion strategy carried out? Archaeological remains from Daemo Mountain Fortress (in Cheongwon) and the Wolpyeong-dong site (in Daejeon) demonstrate that the Goguryeo territory reached its largest point in the fifth century. The southern boundary fluctuated over time, but extended as far south as the Daejeon area of the Geum River region, which was deep into Baekje territory. Goguryeo’s method of controlling their annexed regions south of the Han River also has been the subject of much debate. The archaeological evidence suggests that the territory acquired by King Jangsu was governed directly, a strategy that contrasts with the earlier methods employed by King Gwanggaeto the Great. As for the detailed methods of governance, several scenarios can be proposed. The territory may have been reorganized as a habu (下部), an administrative unit that is mentioned in the Jungwon Goguryeo Monument. Although the traditional regional administrative system of Goguryeo (i.e., the gun-hyeon 郡縣 system) may not have been applied to this territory, the habu may have also been further divided into smaller units that were referred to as gun and hyeon, with the fortress acting as the administrative center for these regional units (Yi Hoyeong 1984, 5-8). Notably, the Han River region is also known to have been a Goguryeo gun at the time that it was reclaimed by Baekje forces in 551 (No Taedon 2005, 185-6). Then what happened to Hanseong after it fell to Goguryeo forces? It has been suggested that the forts of Mt. Acha and its environs were constructed to defend South Pyeongyang (Nampyeongyang), which was located in the area of the Jungnang River. These facilities are thought to have been established in relation to King Munjamyeong’s southern expedition of 495 (Choe Jangyeol 2001). It is believed that South Pyeongyang — which had been under the administrative control of Jangsu Mountain Fortress, located in Sinwon-gun, Hwanghae Province, until the midfourth century—was relocated to the Han River region after 475 (Son Yeongjong 1990, 175-8). According to this view, Goguryeo stationed its forces at Mongchon Fortress after the fall of Hanseong in 475, but continued to refer to the Baekje capital as “Hanseong.” Then, after 495, South Pyeongyang was established on the flatlands immediately to the north of the Han River. This scenario coincides with the author’s interpretation that the Goguryeo forces stationed at Mongchon Fortress after 475 were later withdrawn to the forts of Mt. Acha on the northern banks of the Han River. Evidence indicates that local officials were sent to the Han River region in order to implement Goguryeo’s territorial control over the region. The gun units were administered by susa (守事), who were officers of the daehyeong (大兄) rank in the Goguryeo state. As the term “susa” also appears in the text of the Jungwon Goguryeo Monument, it seems likely that the Han River region comprised an administrative unit similar to the gun (No Taedon 2005, 187), which in turn suggests the presence of a local official, such as a susa. As yet, there is little direct evidence for this possibility, but the discovery at Mt. Acha Fort 4 of a pottery piece bearing the inscription of “hubudo○hyeong” (後卩都○兄) may shed some light on the matter (Fig. 20, left). In this inscription, “hubu” (後卩) can be taken to mean “hubu” (後部, one of the five centrally ruled districts of Goguryeo) and “do○hyeong” (都○兄) seems to be the name of a person or rank. A similar example can be found in a stone inscription from Pyeongyang Fortress, reading “丙戌十二月中漢城下後卩小兄文達節自此西北行涉之” (Fig. 20, right). The inscription dates to the byeong-sul year, or 566, so it postdates the example from Mt. Acha Fort 4. As the two inscriptions are relatively close chronologically, and given that Hanseong in the Han River region was relocated to the Sinwon region of Hwanghae Province in 551 (Son Yeongjong 1990, 177), the hubu of the Mt. Acha Fort 4 inscription can be taken to mean the hubu district of Hanseong. If that is the case, then it would be reasonable to assume that the Han River region at the time was also divided into several districts, as was the capital region of Goguryeo. The Mt. Acha Fort 4 inscription does not contain any rank names, so it is difficult to know the precise rank of the administrative officer that was dispatched. However, it is highly likely that officers of various ranks were dispatched to the region. Fig. 20. (Left) Pottery with hubu inscription from Mt. Acha Fort 4. (Seoul National University Museum). (Right) Stone inscription dating to the byeongsul year from Pyeongyang Fortress. Joseon Yujeokyumul Dogam 3 (조선유적유물도감 3). (Pyeongyang: Joseon Yujeokyumul Dogam Compilation Committee, 1993, p. 95, Fig. 142). Conclusion At present, the earliest Goguryeo artifact to have been unearthed in southern Korea is the globular jar from the site of Juwol-ri in Paju, which dates to the late fourth or very early fifth century. Stone chamber tombs with horizontal entrances and elongated rectangular burial chambers appear from the mid-fifth century onward. The construction of these tombs (e.g., from the Gangnae-ri and Dujeong-ri burial grounds or Bangdong-ri Tomb 2) can be understood in relation to Goguryheo’s advancement into and annexation of the Chungju region, via the upper reaches of the Bukhan and Namhan Rivers, which took place in the late fourth century. On the other hand, the appearance of stone chamber tombs with horizontal entrances and rectangular chambers can be understood in relation to the advancement into and annexation of the areas of Jinchon, Cheongwon, and Daejeon by Goguryeo forces stationed at Mongchon Fortress, following the fall of Hanseong in 475. The fact that Goguryeo settlements have regularly been discovered in the vicinity of the tombs indicates that Goguryeo intensively and continuously maintained control over the captured territories for a substantial period of time. Archaeological remains of Goguryeo activity at Mongchon Fortress, Daemo Mountain Fortress, Namseonggol Mountain Fortress, the Wolpyeong-dong site, Eundae-ri Mountain Fortress, and Dangpo Fortress all date to the late fifth century, and can also be associated with Goguryeo’s attempts to maintain control over the Jinchon, Cheongwon and Daejeon areas. Finally, in the sixth century, Goguryeo forts came to be established on Mt. Acha and its environs, north of the Han River, and most of the forts of the Yangju Basin and the Imjin-Hantan River region also appear to date to this period.

Hwagyesa Temple (華溪寺, Fig. 1), located on Mt. Samgak in Seoul, is best known today as the home of the International Seon Center (國際禪院), where international visitors can study, practice, and experience Korean Buddhism. Behind this modern building are the traditional temple structures, most of which were built in the late nineteenth century; they include Daeungjeon Hall (大雄殿, Hall of Mahavira); Myeongbujeon Hall (冥府澱, Hall of Judgment); Samseonggak Hall (三聖閣, Hall of the Three Sages); and Bohwaru Tower (寶華樓, Tower of the Precious Flower). In 1974, Gwaneumjeon Hall was destroyed by fire, and Cheonburobaekseongjeon (千佛五百聖殿, Hall of 1000 Buddhas and 500 Arhats), with its enshrined Arhat statues, was built on the site (Hong Yunsik 1988, 160). Thus, although Gwaneumjeon disappeared nearly 40 years ago, the painting of Ksitigarbha (Korean, Jijang Bosal, 地藏菩薩) originally enshrined there, along with its accompanying inscription, has been preserved and now is in the collection of the National Museum of Korea (NMK).1 This painting provides a rare opportunity to examine the historical background of the construction of Gwaneumjeon, including the goals of its patrons. Fig. 1. Daeungjeon of Hwagyesa Temple, Seoul. (Author’s photograph). Ksitigarbha Triad (Fig. 2) shows the Ksitigarbha Bodhisattva at the center, accompanied by his two attendants, the monk Domyeong (道明尊者) on the right and the demon king Mudok (無毒鬼王) on the left. Seated on a lotus throne, the bodhisattva holds a transparent jewel with both hands, while Domyeong holds a monk’s staff, another conventional attribute of the bodhisattva. Frequently featured as the main icon of Buddhist paintings during the Joseon Dynasty (1392-1910), the Ksitigarbha Bodhisattva was typically shown surrounded by other deities in a complex compositional arrangement. By contrast, the NMK Ksitigarbha Triad has a rather simple composition—a triad featuring just the bodhisattva and his two attendants in a relatively small pictorial space. At the bottom of the painting is an inscription, or hwagi (畵記), which records that the painting was created in 1876 for the Gwaneunjeon of Hwagyesa Temple on Mt. Samgak. The inscription also states that Sanggung Kim Cheonjinhwa (尙宮 金天 眞華, dates unknown) was the patron, or ingwon siju (引勸施主), who raised the funds to pay for the commission. During the late Joseon period, Sanggung, an official title for a lady-in-waiting at the court, were generally assigned to preside over the inner affairs of the palace, and were often major patrons of Buddhist activities, including the renovation of temple buildings and the production of Buddhist icons. Notably, Sanggung Kim chose to refer to herself in the inscription by her ordained name, rather than her given name, demonstrating that she, like many other women of the Joseon court in the late nineteenth century, was a devout Buddhist patron. The inscription also lists Queen Dowager Hong (王大妃 洪氏, 1831-1903) and some deceased members of the Namyang Hong clan, further indicating the patronage’s royal associations. Fig. 2. Ksitigarbha Triad from Gwaneumjeon of Hwagyesa Temple, Seoul. 1876. Colors on silk, 131.0 x 70.3 cm. (National Museum of Korea). In order to examine the monk-painters who were active in the metropolitan area during the late nineteenth century, this paper examines in detail the Ksitigarbha Triad from Hwagyesa Temple. In particular, the painting’s inscription evinces the close relationship between the royal court and Hwagyesa Temple, which has relevance for a number of meritorious activities that took place at the temple late in the nineteenth century, including the production of Buddhist icons and the construction of temple buildings. Court Patronage and Buddhist Icons of Gwaneumjeon With the founding of the Joseon Dynasty in 1392, the national capital was moved from Gaeseong (開城) to Hanyang (漢陽, present-day Seoul). As a result, many Buddhist temples located in and around the new capital were elevated to the status of wonchal (願刹), or royal prayer halls, and they received the patronage of the royal court and nobility. Hwagyesa Temple was maintained in this manner; in the mid-Joseon period, royal patronage increased substantially, helping Hwagyesa to evolve into one of the most illustrious temples in the area by the late-Joseon period. Samgaksan Hwagyesa yakji (三角山華溪寺略誌, Abridged Record of Hwagyesa Temple on Mt. Samgak; hereafter Record of Hwagyesa), compiled in 1938, traces the history of the temple back to the founding of Bodeogam Hermitage (普德庵), the forerunner of Hwagyesa Temple, by the monk Tanmun (坦文, 901-975) in the early years of the Goryeo Dynasty (918-1392) (Gwon Jongsik 1938, 1). According to this source, Hwagyesa Temple underwent major renovations three times during the Joseon Dynasty, all under royal patronage. First, during the reign of King Jungjong (中宗, r. 1506-1544), the hermitage was relocated to Hwagye-dong, through the agency of Seopyeonggun Yi Gong (西平君 李公, dates unknown), at which time it was renamed Hwagyesa Temple (Figs. 3 and 4). Ravaged by fire in 1618, the temple was entirely rebuilt under the sponsorship of the descendants of Prince Deokheung (德興大君, 1530-1559). By the mid-nineteenth century, the temple had fallen into disrepair, so in 1866, two monks named Yongseon (龍船) and Beomun (梵雲) sought help from Heungseon Daewongun 興宣大院君 (1820-1898, better known in the West as Prince Yi Haeung), who was then ruling Joseon as regent until his son King Gojong (高宗, r. 1863-1907) came of age. Following the lead of Heungseon Daewongun, several women of the royal court—ranging from Grand Queen Dowager Jo (大王大妃 趙氏, 1808-1890) to high-ranking court ladies—contributed to the reconstruction of Hwagyesa Temple in the following years. Under the aegis of these influential palace women, the halls were newly built and the necessary paintings and sculptures were produced and installed. Thus, Hwagyesa Temple underwent three major renovations during course of the Joseon Dynasty. Fig. 3. Cover of Samgaksan Hwagyesa yakji (三角山華溪寺略誌) (Kyeongseong: Samgaksan Hwagyesa, 1938). (National Assembly Library of Korea). Fig. 4. Bodeogam Hermitage, the former site of Hwagyesa Temple. Samgaksan Hwagyesa yakji. (National Assembly Library of Korea). The construction of Gwaneumjeon marked a new phase in the history of patronage by royal women at Hwagyesa Temple. According to the 1876 document “Hwagyesa Gwaneumjeon jungchang danhwakgi” (華溪寺觀音殿重刱丹雘記, “Record of Reconstructing and Refurbishing Gwaneumjeon at Hwagyesa Temple”), the royal court gave the temple an icon of the Avalokitesvara Bodhisattva (Korean, Gwaneum Bosal, 觀音菩薩) called Daebi su taeng or “Embroidered Icon of Great Compassion” (大悲繡幀, here-after Embroidered Avalokitesvara Bodhisattva), which was enshrined in the winter of 1875 (Gwon Jongsik 1938, 6). However, as the document reports, the building where the icon was enshrined was very old and small, “to the extent of making the congregation too embarrassed to view the icon.” Thus, the following year, a larger building was constructed to house the icon. The bestowal of the Embroidered Avalokitesvara Bodhisattva and the reconstruction of Gwaneumjeon marked the first time in the history of Hwagyesa Temple that the temple received royal patronage. The construction project is particularly significant, since it was followed by a continuing series of similar projects funded at least in part by the royal court. More details can be found in the opening section of a text from 1900 entitled “Hwagyesa Myeongbujeon bullyang seo” (華溪寺冥府殿佛糧序, “Preface to [the Donation of] Grains to Myeongbujeon of Hwagyesa Temple”). According to the text, “Although Hwagyesa Temple was founded long ago, its only building was Daeungjeon. In 1875, a sacred image of Avalokitesvara, adorned with a hundred jewels, was embroidered, and shortly thereafter a pavilion was built in which to enshrine it. This is a very rare instance in the history of temples in this country…” The record further states that Myeongbujeon was built under the patronage of court ladies and that Great Queen Dowager Jo donated bullyang (佛糧, farmland) in order to maintain the temple (Gwon Jongsik 1938, 8-10). As the enshrinement hall for the special embroidered bodhisattva icon that had been donated by the royal court, Gwaneumjeon seems to have been a representative structure connecting Hwagyesa Temple and the royal court. Unfortunately, the hall burned down in 1974, before its historical significance had been adequately investigated. Thus far, there are no known photographs of the building, so its appearance can only be reconstructed based on written records from the early twentieth century. The Record of Hwagyesa relates that Gwaneumjeon was located to the right of Daeungjeon, and that it was rather small, with a floor space of 5 pyeong (about 16.6m2). A plaque bearing calligraphy by Nam Sik (南軾, 1803-1878) was displayed on the front of the building, while four long tablets of wood featuring verse couplets describing the exquisite scenery of the bodhisattva’s abode were hung on the anterior pillars of the pavilion. Another document entitled Bongeunsa bonmalsa jaesan mongnok (奉恩寺本末寺材産 目錄, List of Properties of the Bongeunsa Temple and Its Subtemples, hereafter Properties of Bongeunsa), which was commissioned by the Japanese Government-General of Korea in 1933, lists the statues and paintings housed in Gwaneumjeon. Interestingly, Table 2 indicates that, in addition to the aforementioned Embroidered Avalokitesvara Bodhisattva, Gwaneumjeon received another embroidered icon of the Avalokitesvara Bodhisattva from the royal court in 1910. Within the Joseon royal palace was an “embroidery room”, or subang (繡房), where female members of the royal court could request the production of various Buddhist textiles, including such items as monk’s robes, Buddhist altar coverings (佛卓褓, bultak bo), and covers for Buddhist scriptures (Fig. 5). Of the items produced in the embroidery room, the embroidered icons bestowed upon Hwagyesa Temple are very distinctive, in that they were objects of worship. Considering that the royal court bestowed such special icons not once, but twice, Gwaneumjeon must have been considered very important at the time. Neither of the embroidered icons of Avalokitesvara Bodhisattva has survived, but a black and white image of the 1875 icon was published in the Record of Hwagyesa (Fig. 6). The photo shows the Thousand-Armed Avalokitesvara Bodhisattva (with 42 hands), riding a lion atop a rocky mountain symbolizing Mt. Potalaka, the bodhisattva’s abode. A woodblock carved with an image of Avalokitesvara Bodhisattva (Fig. 7) was made by the temple in order to commemorate the icons bestowed by the court. Table 1: List of properties housed in Gwaneumjeon, Hwagyesa Temple from Properties of Bongeunsa. Name Quantity Medium Form Size Art Amitabha Buddha 1 Gilt stone image Seated Height: ~30.3 cm Unknown Amitabha Buddha 1 Gilt wooden image Seated Height: ~53 cm Unknown Acolyte 2 Wooden image with pigments Seated Height: ~21.2 cm Unknown Arhat 1 Wooden image with pigments Seated Height: ~3.6 cm Unknown Painting of Avalokitesvara 2 Paper Hanging Scroll Height: ~116.7 cmWidth: ~72.7 cm Palace woman Painting of the Nine Grades 1 Paper Hanging Scroll Height: ~157.6 cmWidth: ~203 cm Geumhwadang錦華堂 Painting of Ksitigarbha 1 Paper Hanging Scroll Height: ~133.3 cmWidth: ~142.4 cm Hwasandang華山堂 Painting of Guardian Deities 1 Paper Hanging Scroll Height: ~136.3 cmWidth: ~157.6 cm Hyegwadang慧果堂 Table 2: List of important properties housed in Gwaneumjeon from Record of Hwagyesa. Name Quantity Form Size Date Artist Amitabha Buddha 1 Height: ~30.3cm Width: ~15.2 cm Unknown Acolyte 2 Seated image Height: ~212.1 cmWidth: ~45.5 cm Unknown Arhat with bamboo branch 1 Height: ~36.4 cmWidth: ~18.2 cm Unknown Painting of Avalokitesvara 1 Embroidery Gwangseo 1, eulhae year (1875) Palace woman Painting of Avalokitesvara 2 Embroidery Yunghui 4 (1910) Palace woman Painting of Guardian Deities 1 Colored Fifth month of Gwangseo 2, byeongja year (1876) Geumhwadang錦華堂 Painting of Ksitigarbha and the Ten Kings of Hell 1 Colored Fifth month of Gwangseo 2, byeongja year (1876) Hwasandang華山堂 Painting of the Nine Grades 1 Colored Tenth month of Gwangseo 12 (1886) Hyegwadang慧果堂 Fig. 5. Textile covering for the main altar from Daeungjeon, Tongdosa Temple, Yangsan, South Gyeongsang Province. 1890. Embroidered silk, 124.1 x 924.0 cm. (Research Institute of Buddhist Cultural Heritage). Fig. 6. Embroidered Avalokitesvara Bodhisattva from Gwaneumjeon, Hwagyesa Temple, Seoul (current whereabouts unknown). 1875. Embroidered silk, 113.0 x 80.0 cm. Samgaksan Hwagyesa yakji. (National Assembly Library of Korea). Fig. 7. Avalokitesvara Bodhisattva from Hwagyesa Temple, Seoul. Late nineteenth to early twentieth century. Woodblock, 113.0 x 78.0 cm. (Research Institute of Buddhist Cultural Heritage). The Ksitigarbha Triad now in the collection of the National Museum of Korea is listed in the Record of Hwagyesa as Painting of Ksitigarbha, and is also listed in the Properties of Bongeunsa as Painting of Ksitigarbha and the Ten Kings. The Record of Hwagyesa also lists another Painting of Ksitigarbha and the Ten Kings (地藏及十王幀), which would seem to refer to a single painting depicting the bodhisattva with the ten kings of hell. However, Myeongbujeon houses a painting with a unique composition, showing the ten kings without Ksitigarbha Bodhisattva (Fig. 8). The Record of Hwagyesa lists this work as Painting of the Ten Kings (十王幀). It is rare to find a painting of a Ksitigarbha triad enshrined in Gwaneumjeon, which is a hall devoted to the Avalokitesvara Bodhisattva. Even more interestingly, the paintings of the Ksitigarbha triad and the ten kings, both enshrined in Gwaneumjeon, were produced by the same workshop, including the jeungmyeong (證明, the monk who ensures that Buddhist activities follow proper procedures); hwaju (化主, project supervisor and fundraiser); geumeo (金魚, monk-painter); ingwon siju (patron or fundraiser); and so forth. However, patronage stands as one major difference between the two paintings. While the Painting of the Ten Kings in Myeongbujeon was sponsored by Lady Hong and Lady Kim (presumably women of the yangban class), the Ksitigarbha Triad from Gwaneumjeon was dedicated by Sanggung Kim Cheonjinhwa for the wellbeing of Queen Dowager Hong and the peaceful repose of members of the Hong clan from Namyang. Gwaneumjeon was the private prayer hall of the royal court, and its main icon—the Ksitigarbha Triad—seems to have been commissioned by Queen Dowager Hong through the agency of Sanggung Kim Cheonjinhwa. The last section of this paper will further examine the characteristics of royal women’s patronage of Hwagyesa Temple. Fig. 8. Ten Kings of Hell from Gwaneumjeon, Hwagyesa Temple, Seoul. 1876. Colors on silk, 121.0 x 77.5 cm. (Research Institute of Buddhist Cultural Heritage). Hwagyesa Temple and the Head Monk-Painter of the Ksitigarbha Triad Gwaneumjeon was rebuilt in 1876 to enshrine the Embroidered Avalokitesvara Bodhisattva, which is said to have measured 120.0 centimeters in height. Produced in 1876, the Ksitigarbha Triad (Fig. 2) measures 79.3 centimeters in height and 131.0 centimeters in width; its size and proportions indicate that it was probably made in accordance with the dimensions of Gwaneumjeon. As mentioned, the painting has a simple composition, featuring the Ksitigarbha triad without any other figures. The monk-artist who made this painting opted for the triad composition in order to emphasize the subject matter and iconography, but also to accommodate the dimensions of the space where it would be enshrined. The Ksitigarbha Bodhisattva, the painting’s main deity (Fig. 9), sits with his legs crossed on a lotus throne that emerges from a lotus flower. Wearing a black mesh headpiece, the bodhisattva holds a jewel with both hands at chest level. Similar depictions of the bodhisattva appear in paintings of the Ksitigarbha Bodhisattva and the Ten Kings at Heungcheonsa Temple in Seoul (dated 1867, Fig. 10); Jeongsusa Temple on Ganghwado Island (dated 1878); and Bongnimsa Temple in Hwaseong (dated 1883), all of which were produced near the capital in the late nineteenth century. The Ksitigarbha Triad of Hwagyesa Temple shares close affinities with these other paintings, in terms of the colors and the overall styles of the main figures, illustrating how the same subject matter was depicted in the region. There are also notable differences between the Ksitigarbha Triad and the Heungcheonsa painting (Fig. 10); in the former, the figures are slender; the bodhisattva wears a black mesh hat against a gold aureole; and the figures of the monk Domyeong and the demon king Mudok are larger and nearer to the bodhisattva. These differences in expression and composition likely can be attributed to the individual styles of the monk-painters. Fig. 9. Ksitigarbha Bodhisattva, detail of Fig. 2. Fig. 10. Ksitigarbha Bodhisattva and the Ten Kings of Hell from Daeungjeon Hall, Heungcheonsa Temple, Seoul. 1867. Colors on silk, 177.0 x 240.0 cm. (Research Institute of Buddhist Cultural Heritage). The inscription at the lower left of the Ksitigarbha Triad lists the names of the monks involved in producing the painting, including the lead artist Hwasan Jaegeun (華山在根), Hyego Bonggam (慧杲奉鑑), Deoksun (德順), and Sango (尙悟), a novice. In the late Joseon period, Buddhist paintings were produced exclusively by monks, who held various titles, including hwawon (畵員), geumeo (金魚), yongmyeon (龍眠), or pyeonsu (片手). The terms “geumeo” and “aaa” refer to the head painter who is in charge of the entire production, while the assistant painters were typically known as “pyeonsu.” The first part of the inscription indicates that the Ksitigarbha Triad was painted under the leadership of “Hwasan,” which was the sobriquet of the monk-painter Jaegeun. Haedong buljo wollyu (海東佛祖源流, Origin of Buddhist Patriarchs in the East of the Sea), a book tracing the dharma lineage of Korean monks, relates that Hwasan Jaegeun was a disciple of Choam Kiju (草庵基珠), who was from the dharma linage of Yulbong Cheonggo (栗峰靑杲, 1738-1823). Notably, Choam Kiju is one of several monks who undertook the project to renovate Hwagyesa Temple at the end of the nineteenth century. Hwagyesa Temple currently houses several paintings that were produced at the end of the nineteenth century, including Painting of Amitabha Buddha’s Assembly (1875) in Daeungjeon; Painting of the Ten Kings in Gwaneumjeon (1876); Painting of Ksitigarbha Bodhisattva and the Ten Kings in Myeongbujeon (1878), and Gwaebul Painting (1886) (Figs. 11 and 12). Aside from Gwaebul Painting, the temple’s other Buddhist paintings from the 1870s were all completed under the leadership of Hwasan Jaegeun, with Choam Kiju acting as a jeungmyeong, (證明, one who verifies Buddhist procedures). The master-disciple relationship between the two monks might have enabled Hwasan Jaegeun to take full charge of making Buddhist paintings to be enshrined in Hwagyesa Temple during this period. Fig. 11. Preaching of Amitabha Buddha from Daeungjeon, Hwagyesa Temple, Seoul. 1870. Colors on silk, 251.0 x 225.6 cm. (Research Institute of Buddhist Cultural Heritage). Fig. 12. Ksitigarbha Bodhisattva from Myeongbujeon, Hwagyesa Temple. 1878. Colors on silk, 189.7 x 249.5 cm. (Research Institute of Buddhist Cultural Heritage). Fig. 13. Ksitigarbha Bodhisattva from Myeongbujeon, Buryeongsa Temple, Uljin, North Gyeongsang Province. 1880. Colors on silk, 258.0 x 273.0 cm. (Research Institute of Buddhist Cultural Heritage). Fig. 14. Ksitigarbha Bodhisattva and the Ten Kings of Hell from Daeungjeon, Heungguksa Temple, Namyangju, Gyeonggi Province. 1868. Colors on silk, 169.5 x 198.5 cm. (Research Institute of Buddhist Cultural Heritage). Hwasan Jaegeun’s name appears on several extant paintings that were made before he became the head painter at Hwagyesa Temple in the late 1870s. The earliest such appearance is in the inscription on the 1870 painting Guardian Deities, at Yeongsusa Temple in Jincheon, Gangwon Province, the production of which was led by the monk-painter Sangwol (尙月). As very few paintings by Sangwol have survived, it is difficult to compare his style to that of Hwasan Jaegeun’s later works and thus to assess the degree of Sangwol’s influence on Hwasan Jaegeun. Extant works suggest that Hwasan Jaegeun formulated his own style while participating in a number of painting productions led by renowned monk-painters active around the capital in the late 1870s. For example, he served as an assistant painter in the making of Buddhist paintings at Heungguksa Temple in Namyangju, at Baengnyeonsa Temple in Seoul, and at Bogwangsa Temple in Paju; the teams executing those paintings were led by Keumgok Yeonghwan (金谷永煥), Kyeongseon Eungseok (慶 船應釋), Bangu Jingo (放牛珎杲), and Manpa Donjo (萬波頓照), respectively, from 1868 to 1874. Hwasan Jaegeun’s name also appears in the inscriptions on several paintings from subtemples at Yujeomsa Temple in Goseong, Gangwon Province, including the main altar painting at Heulseongam Hermitage and paintings of Bodhidharma and Guardian Deities at Banyaam Hermitage. Keumgok Yeonghwan and Kyeongseon Eungseok were leading monk-painters who were active in the region around Seoul and Gyeonggi Province in the late nineteenth century (Ahn Kui-sook 1994, 70-72; Chang Hee-jeong 2003, 124-158; An Kui-sook and Choi Sunil 2008). Extant Buddhist paintings indicate that Hwasan Jaegeun collaborated with these painters and followed the painting style that was prevalent in the region of Seoul and Gyeonggi Province, but that he eventually developed his own style. Some scholars have suggested that Hwasan Jaegeun may have learned the style of the Geumgangsan School from Gangwon Province under Manpa Jeongik (萬波定翼) (Han Munyeong 2000). The only known extant work painted by both Manpa Jeongik and Hwasan Jaegeun is Painting of Ksitigarbha Bodhisattva in Myeongbujeon at Buryeongsa Temple (Fig. 13). Interestingly, and somewhat confusingly, Hwasan Jaegeun also seems to have worked closely with a monk named Manpa Donjo, who shared the same sobriquet as Manpa Jeongik. Several works by Hwasan Jaeegun and Manpa Donjo survive, including Painting of Ksitigarbha (1868) in Daeungjeon at Heungguksa Temple in Namyangju; Painting of Ksitigarbha and Painting of Messengers (1872) in Myeongbujeon at Bogwangsa Temple in Paju; and Painting of Guardian Deities in Wontongjeon at Jeokseoksa Temple on Ganghwado Island (Fig. 14). Manpa Donjo and Manpa Jeongik were contemporaries, but they worked in different regions. Based on extant works, Manpa Donjo was active from 1868 to 1891 in the area around the capital, encompassing Seoul, Namyangju, Paju, Ganghwado Island, Anseong, and Suwon. Manpa Jeongik, on the other hand, was active mainly from 1877 to 1898 and worked primarily in Seoul, Goseong, and Pyeongchang in Gangwon Province, and Uljin in South Gyeongsang Province. This evidence indicates that Manpa Donjo and Manpa Jeongik were in fact two different painters with the same sobriquet. Examination of Hwasan Jaegeun’s oeuvre indicates that Manpa Donjo exerted more influence on him than did Manpa Jeongik. The close ties between Hwagyesa Temple and Hwasan Jaegeun are further evinced by a pair of documents. First, “Gyeonggi-do Bukhan Samgaksan Hwagyesa Daeungbojeon junggeon gimun” (京畿道北漢三角山華溪寺大雄寶殿重建記文, “Record of Reconstructing Daeungbojeon of Hwagyesa Temple on Mt. Samgak, North of the Han River in Gyeonggi Province”), compiled in the Record of Hwagyesa, lists the names of monks who were living in Hwagyesa Temple in 1870, when Daeungbojeon was being rebuilt (Gwon Jongsik 1938, 4-5). Hwasan Jaegeun is listed first among twelve novice monks of Hwagyesa Temple. He is referred to by the term “sami,” or “novice,” which refers to a man between the ages of 14 and 20 who has taken the ten precepts (十戒). Thus, it would seem that by the 1870s, Hwasan Jaegeun had taken the tonsure at Hwagyesa Temple and was participating in meritorious activities led by his dharma teacher, Choam Kiju. The second document is Buryeongsa sichang gi (佛影寺始創記, Record of the Founding of Buryeongsa Temple), a manuscript from the late nineteenth century, which records the establishment and renovation of Buryeongsa Temple in Uljin, South Gyeongsang Province. According to this text, 26 monk-painters contributed to the production of 15 Buddhist paintings and to the re-gilding and re-painting of 30 Buddhist sculptures (Center for Korean Studies References 1983, 367-368). Hwasan Jaegeun is listed as one of two head painters directing the project, along with Seokong Cheolyu (石翁喆侑) of Singyesa Temple on Mt. Geumgang in Gangwon Province. More specifically, the text lists “Hwasan Jaegeun of Hwagyesa Temple on Mt. Samgak in Gyeonggi,” clearly demonstrating that Hwasan Jaegeun was closely associated with Hwagyesa Temple by that time. Reconstruction and Patronage of Hwagyesa Temple in the Late Nineteenth Century Record of Hwagyesa documents the location of various paintings at the temple. According to this record, Painting of Ksitigarbha (1875) by Seongam (聖庵) was enshrined in Daeungjeon; Hwasan Jaegeun’s Ksitigarbha Triad (1876) was in Gwaneumjeon; Hwasan Jaegeun’s Painting of Ksitigarbha Bodhisattva (1878) was hung in Myeongbujeon. “Seongam” refers to the monk-painter Seongam Seungseon (聖庵勝宜 or 性庵勝宜), whose Painting of the Ten Kings (1878) still hangs in Myeongbujeon at Hwagyesa Temple. All of these paintings were commissioned by Heungseon Daewongun (i.e., Prince Yi Haeung), and women of the royal court, who commissioned paintings of virtually identical subject matter for each prayer hall in the temple hall within a period of four years. As mentioned, these same patrons also spearheaded the renovation of Hwagyesa Temple after 1870. Heungseon Daewongun’s patronage of Buddhist affairs seems to have begun with the reconstruction of Yonggungsa Temple on Ganghwado Island in 1864 (Center for Korean Studies References 1978, 121-122). Other patrons contributing to that project included Grand Queen Dowager Jo, Queen Dowager Hong, and Gyeongbin Kim (慶嬪 金氏, 1832-1907), with Sanggung Kim Cheonjinhwa serving as the fundraiser (Center for Korean Studies References 1978, 121). Notably, these are the same major donors that later would lead the reconstruction of Hwagyesa Temple. After the reconstruction of Daeungjeon at Hwagyesa Temple, Heungseon Daewongun no longer participated as a major patron for Buddhist activities. His patronage of Buddhist temples occurred between 1863 and 1873, the period during which he served as regent for his son, who had not yet come of age to rule. In the footsteps of Heungseon Daewongun, various queens, royal concubines, and court women came to prominence as major patrons of the reconstruction of Hwagyesa Temple. In particular, the bestowal of the Embroidered Avalokitesvara Bodhisattva in 1875 seems to have marked a turning point in the patronage pattern, as women of the royal court assumed control of the patronage from that time on. This embroidered icon is believed to have been donated to the temple by Empress Myeongseong (明成皇后, 1851-1895) to celebrate the birth of her son, Crown Prince Yi Cheok, who later became Emperor Sunjong (純宗, r. 1907-1910) (Choi Wan-soo 1994, 330; Institute of Temple Culture 1994, 174). Empress Myeongseong was known as a prominent and generous patron of Buddhist activities, who funded a variety of Buddhist endeavors in the hopes that she would be rewarded with a son. For example, she sponsored the carving of a seated Avalokitesvara Bodhisattva on the cliff at Hakdoam Hermitage in 1872; the production of an illuminated manuscript of the Lotus Sutra in gold pigment at Tongdosa Temple in 1880; the construction of monks’ quarters and other buildings at Yeonjuam Hermitage on Mt. Gwanak in 1888; and the reconstruction of monks’ quarters at Bomunsa Temple in 1888 (Lee Kyunghwa 2002). However, none of the inscriptions or other textual records preserved at Hwagyesa Temple mention Empress Myeongseong, suggesting that she had no connection with the temple. Indeed, all available evidence indicate that it was Grand Queen Dowager Jo, Queen Dowager Hong, and Gyeongbin Kim—rather than Empress Myeongseong—who lavishly sponsored Hwagyesa Temple at the end of the nineteenth century. Grand Queen Dowager Jo became Crown Princess in 1819, when she married Crown Prince Hyomyeong 孝明世子, son of King Sunjo (純祖, r. 1800-1834). She became Queen Dowager in 1834, when her son ascended the throne as King Heonjong (憲宗, r. 1834-1849), and she was granted the title of Grand Queen Dowager in 1857, when King Cheoljong (哲宗, r. 1849-1863) succeeded to the throne. Thus, Grand Queen Dowager Jo was the mother-in-law of both Queen Dowager Hong (queen consort of King Heonjong) and Gyeongbin Kim (royal concubine of King Heonjong). As mentioned, Grand Queen Dowager Jo and Queen Dowager Hong sponsored both the creation of the Ksitigarbha Triad for Gwaneumjeon of Hwagyesa Temple and the reconstruction of Yonggungsa Temple in Ganghwado Island. Queen Dowager Hong also donated grains to Gwaneumjeon in 1883, further supporting her close relation to the Hwagyesa Temple. In addition, after Queen Dowager Hong died in 1903, a brass water jar was given to Hwagyesa Temple in 1904 according to her will (Gwon Jongsik 1938, 23). Again, Queen Dowager Hong and the deceased members of the Namyang Hong clan appear as the major beneficiaries of merit accrued from the dedication of the Ksitigarbha Triad, which was made in 1875 when Gwaneumjeon was built. All of this evidence indicates that Queen Dowager Hong, along with the mother and other royal consorts of King Heonjong, played a significant role in bestowing the Embroidered Avalokitesvara Bodhisattva and financing the reconstruction of Gwaneumjeon. Royal patronage of Hwagyesa Temple continued after the construction of Gwaneumjeon. In 1878, by royal decree, sculptures of Ksitigarbha Bodhisattva and the ten kings of hell (Fig. 8 and Fig. 12) were transferred from Gangseosa Temple (江西寺) in Hwanghae Province to Myeongbujeon of Hwagyesa Temple (Gwon Jongsik 1938, 7). Also, in addition to supporting the construction of Gwaneumjeon and Myeongbujeon at Hwagyesa Temple, the royal court also bestowed farmland to the two halls in order to supply grains for their maintenance. The most important patron of Myeongbujeon was Grand Queen Dowager Jo. According to “Hwagyesa Myeongbujeon bullyang seo” (華溪寺冥府殿佛糧序, Preface to the [Donation of] Grains to Myeongbujeon of Hwagyesa Temple), written in 1900, Grand Queen Dowager Jo granted farmland to the hall in 1880 (Gwon Jongsik 1938, 8-9).2 Along with Queen Dowager Hong and Grand Queen Dowager Jo, Sanggung Kim Cheonjinhwa raised funds for the production of the Ksitigarbha Triad and also played a significant role as a patron of Hwagyesa Temple. Extant sources indicate that her patronage of Hwagyesa Temple had already begun in the 1860s, long before the reconstruction of Gwaneumjeon in 1876. For example, Sanggung Kim Cheonjinhwa was one of several court women who sponsored the production of Painting of the Seven Stars (Fig. 15, now housed in Hyeondeungsa Temple), which was originally made for Hwagyesa Temple in 1861. The group patronage by the royal family and court women continued in the late nineteenth century, as exemplified by the third major reconstruction of Hwagyesa Temple. Sanggung Kim may have acted as emissary between the monks of Hwagyesa Temple and the women of the royal court, who were not free to leave the palace and were banned from visiting Buddhist temples by the official Neo-Confucian policy of the Joseon state. Notably, of the 37 court women who sponsored the transfer, repair, and repainting of the sculptures of Ksitigarbha Bodhisattva and the Ten Kings of the Underworld from Gangseosa Temple, she is the only one whose name was recorded in the dedicatory inscription (Gwon Jongsik 1938, 8). Fig. 15. Seven Stars from Samseonggak Hall, Hyeondeungsa Temple, Gyeonggi Province. Colors on silk, 167.7 x 187.7 cm. (Research Institute of Buddhist Cultural Heritage). Women with the title of “Sanggung” were among the most active contributors to Buddhist temples during the late Joseon period, donating either individually or in groups. Sanggung Yi Seongae (李性愛), for example, personally sponsored the production of the Gwaebul Painting (1735) of Bongseonsa Temple (奉先寺) for the repose of King Sukjong’ s royal concubine Yeongbin Kim (寧嬪 金氏, 1669-1735), whom she must have served. On behalf of the royal court, a court lady sponsored another Gwaebul Painting (1862) from Cheonggyesa Temple (淸溪寺). When such women acted as proxies for the royal court, they usually were listed as the “ingwon siju”, or donor, in the painting inscription, as in the case of the Gwaebul Painting (1803) from Jikjisa Temple (直指寺). Detailed information about the royal patrons of this painting can be found in a separate record, entitled Jikjisa Gwaebul jungsu si chugwonmun (直指寺掛佛重修時祝願文, Dedicatory Prayer Written at the Time of Repairing the Gwaebul Painting of Jikjisa Temple), which was inscribed on a wooden plaque. This record suggests that Queen Jeongsun (貞純王后, 1745-1805), a queen consort of King Yeongjo (英祖, r. 1724–1776), dedicated this painting in hopes of having good health and of bearing a son to become crown prince. Extant evidence indicates that Sanggung Kim may have been acting as proxy for the women of the royal court for activities at Hwagyesa Temple for as many as 15 years before Gwaneumjeon was built. Standing on Mt. Samgak, north of the capital, Hwagyesa Temple received patronage from various members of the royal court, including Prince Seopyeong, Prince Deokheung, and Heungseon Daewongun throughout the mid- to late Joseon period. With the patronage of Grand Queen Dowager Jo, Queen Dowager Hong, and Gyeongbin Kim, the temple thrived and became known as a “gungjeol”, or palace temple, because of its close relationship with the royal court in the late nineteenth century. Conclusion Through an examination of the Ksitigarbha Triad from Gwaneumjeon of Hwagyesa Temple, this paper explored individual painting styles of monk-painters active in the Seoul and Gyeonggi region during the late nineteenth century. Relevant textual sources were analyzed to provide details about the major patrons of Hwagyesa Temple. In the late nineteenth century, Hwagyesa Temple underwent a major reconstruction project, centered around the construction of Gwaneumjeon, which was built to house the Embroidered Avalokitesvara Bodhisattva, which had been donated by the royal court. Notably, the donation of this icon marked a turning point in the pattern of royal patronage at the temple, wherein women of the royal court—as represented by Grand Queen Dowager Jo, Queen Dowager Hong, and Sanggung Kim Cheonjinhwa—became the main benefactors, replacing their male counterparts, such as Seopyeonggun Yi Gong, Prince Deokheung, and Heungseon Daewongun. Examination of the inscription on the Ksitigarbha Triad, which was originally housed in Gwaneumjeon, reveals that both Grand Queen Dowager Jo and Queen Dowager Hong actively supported Hwagyesa Temple. This study also investigated the role of court women in the royal patronage of Buddhist temples during the late Joseon period, focusing primarily on the case of Sanggung Kim Cheonjinhwa, who seems to have acted as a vital emissary between Hwagyesa Temple and the royal court. The Ksitigarbha Triad of Gwaneumjeon at Hwagyesa Temple allows us to examine the royal patronage of Buddhist temples, as well as the work of the monk-painter Hwasan Jaegeun. Research into textual sources and Buddhist paintings of the late nineteenth century, preserved at Hwagyesa Temple, revealed that Hwasan Jaegeun, who was active in the area of the capital in the late nineteenth century, was a monk-painter trained at Hwagyesa Temple.