The Journal of Korean Art and Archaeology

Gaya (42–562)1 competed with Silla (57 BCE–935 CE) for hundreds of years in the area that is now known as the Gyeongsang region of South Korea but eventually succumbed in the late sixth century. Silla’s unique historical perspective solidified over its process of unifying the Three Kingdoms and incorporating Baekje (18 BCE–660 CE) and Goguryeo (37 BCE–668 CE) in the late seventh century, which influenced its writing of history. Gaya’s history was resultingly disparaged because, from the victorious point of view, it had always been part of Silla. This is why Samguk Sagi (三國史記, History of the Three Kingdoms), the oldest canonical text on ancient Korean history, focuses mainly on Goguryeo, Baekje, and Silla. Gaya is rarely mentioned, and in the few instances when it is, it is described as a trivial polity that came to be annexed by various means in the early period of Silla’s history. This has meant that Gaya’s validity as a polity remained unacknowledged for over a thousand years. On the other hand, Gaya appears under the name “任那”, “Imna” in Korean and “Mimana” in Japanese in the Nihon shoki (日本書紀, The Chronicles of Japan), which relates the history of ancient Japan. Its descriptions seem to imply that Gaya’s territory was actually part of Yamato Japan and that an institution for its governance—“Imna Ilbon-bu” (任那日本府)—was maintained. It should be taken into account, however, that these passages on “Imna” reflect the historical perspective of aristocratic Baekje refugees who had fled to Japan following the demise of their kingdom. They were also products of Japanese attempts to compensate for their defeat at the hands of Silla, which had crushed the Japanese forces that came to the aid of the Baekje Revival Movement in the seventh century. Japan’s sense of superiority over Korea following its colonization of the peninsula in the twentieth century led heavily biased accounts on “Imna” to be accepted as straightforward history. Burdened by the tag of “Imna,” Gaya territory came to be regarded as a source of embarrassment for Koreans. In 1945, the military government, led by General Douglas MacArthur in Japan and the American Military Government in Korea, undertook the task of reviewing and editing Japanese and Korean history textbooks. The first phrases to be deleted were those concerning the legend of the ancient Japanese Empress Jingu’s conquest of Samhan and the “Imna Ilbon-bu” theory. These phrases were found to promote an aggressive nature on the part of the Japanese and their discrimination against Koreans. However, rather than the restoration of its history, only silence followed on Gaya. This was exacerbated by the fact that Korean historians possessed little information on the ancient polity. The situation changed markedly in the 1970s with the excavation of Gaya tombs by Korean archaeologists, however. The nature of the artifacts recovered from these tombs, especially their quality rivaling that of artifacts from Silla, came as a welcome surprise to all. Data produced by archaeological excavations have since provided a foundation for Gaya research and have been used to address the errors found in previous studies that had been based only on textual research. As a result, it has now become possible to present new knowledge of Gaya’s history and culture. Brief Summary of the Special Articles on “Gaya” “Gaya History and Culture” provides an overview of the papers I published over roughly 30 years since 1985. Gaya history can be divided into Early Gaya history (from the first century BCE to fourth century CE) and Late Gaya history (from the fifth to sixth century CE). Gaya existed as a confederation of ten to twenty statelets. Early Gaya history was centered on Gayaguk (加耶國), a polity based in Gimhae, while Late Gaya centered on Dae Gayaguk (大加耶國) in what is now Goryeong. This provides a basic framework for a new history of Gaya. Gaya history began with the influx of refugees from Wiman Joseon (194–108 BCE) into the Gyeongsang region in the first century BCE. By the second century, approximately ten Gaya statelets (國, guk) had come to be established around Gimhae and other areas. In the third century, Gayaguk in Gimhae emerged as the head of the twelve Byeonhan statelets and began to organize the Early Gaya Confederacy. In the fourth century, a confrontation took place between Gayaguk and Silla. Gayaguk originally gained the upper hand but suffered great losses when Goguryeo forces came to Silla’s defense. In the fifth century, Banpaguk (伴跛國) developed rapidly in Goryeong, which had originally been a more backward area. Later renamed “Dae Gayaguk” or “Garaguk (加羅國),” it played a central role in establishing the Late Gaya Confederacy. In the late fifth century, Dae Gaya presented an offering of tribute to the Southern Qi Dynasty (479–502) court. It even sent reinforcements to Silla when that kingdom faced a threat from Goguryeo. In this way, Dae Gaya made its presence known on the international stage. However, a number of statelets based in the eastern Jeolla region were lost in the wake of a conflict with Baekje in the early sixth century. Dae Gaya subsequently lost the statelets composing the southeastern part of its territory to Silla, which annexed them by abusing a marriage alliance between Silla and Dae Gaya. This further weakening of Dae Gaya created an opening for Baekje to attack, and the statelets that formed the southwestern part of Dae Gaya territory took this opportunity to escape from Dae Gaya’s influence. These circumstances resulted in the weakening of Dae Gaya’s hegemony. Amidst the division of the Late Gaya Confederacy into northern and southern factions in the mid-sixth century, great effort was put into resisting Baekje and Silla attacks and maintaining independence. However, its conquest by Silla was completed in 562 CE. The economy was well developed in Gaya territory, with fishing, farming, iron production, and trade being the most prominent activities. The development of iron deposits and favorable conditions for maritime transport led to the establishment of long-distance trade, which fueled Gaya’s cultural progress. Numerous Buddhist legends are set in Gaya territory, some of which have been used to suggest links with Indian Buddhism. However, these legends appear to date back only to the waning stages of Gaya history or even to the period after Gaya’s demise. Human sacrifices as components of burials were more prevalent in Gaya compared to other regions. This is considered to be associated with the fact that Gaya remained in a transitional state and never fully emerged as an ancient state. “Gaya Armor: The Culmination of Gaya Iron Crafting” was written by Kim Hyuk-joong, a curator at the Gimhae National Museum. Made by connecting long vertical plates, vertical plate armor first came to be used around the fourth century in southern region of Korean peninsula, when social tension increased due to warfare and the development of weaponry accompanied by improvements in iron production technology. The vertical plate armor appeared in the similar time in both Silla and Gaya. Gaya armor with its distinctive bird- and fern-shaped decorations is believed to have been worn to display social standing. Helmets made with vertical plates, which were the main type of helmet in this period, were common throughout the Three Kingdoms with little regional variation observed. Examples from Gaya were fitted with sun-shades or cheek-covers with fern-design, and these greatly influenced helmets worn on the Japanese Archipelago. The infantry played a central role in the military tactics of the period in which iron armor first came into use in Gaya. The presence of lamellar armor, horse armor, horse helmets, saddles, and stirrups in Gaya tombs dating to the fifth century onward seems to indicate the adoption of heavy cavalry to a certain extent. However, it is likely that Gaya’s heavy cavalry was not equal in number to that of Goguryeo or Baekje. Upon its transmission to the Japanese Archipelago, Gaya armor developed into “Wa-style armor” starting in the fifth century. Such Wa-style armor has been found in several parts of the Gaya region, leading to the proposal of the “Imna Ilbon-bu” theory. However, it has more recently come to be regarded as the result of exchanges between Gaya and Wa (倭) in Japan or of other interactions that took place over a long period of time. Kim’s paper presents a fairly objective overview of the nature of the research that has been undertaken on Gaya iron armor to date. Unfortunately, neither Gaya’s advanced iron production technology, a key element in Gaya’s iron culture, nor its iron farming tools, weapons, or horse gear were fully addressed in this paper. This is due to the fact that Gaya archaeological research has branched out in numerous channels, making it diffcult for a single researcher to possess a specialist’s knowledge and insights on all topics. The page limit for the articles also restricted a broader overview. “Developments in the Pottery Culture of Gaya,” by Lee Jeonggeun, Chief Curator at the Gimhae National Museum, examined issues such as the emergence of grayish-blue stoneware, the Gaya pottery production system, the mass-production and distribution of jars, the diversity of Gaya pottery, and the influence of Gaya pottery technology on Japanese stoneware. Grayish-blue stoneware began to be produced around the fourth century CE. There are two hypotheses on its origins: One is that it can be traced back to the two-lugged jars of the Northeast region of China. The other is that it was an independent invention that occurred following indigenous technological innovation. Gaya pottery is generally divided into an early (the fourth century) and a late phase (the fifth to the sixth century). In the early phase, Gimhae and Haman were the main centers of pottery production, and a key product was jars used as containers for liquids. During this phase, a system for the mass-production of pottery that took advantage of nesting two or three vessels in the kiln was well established in the Haman area. Such mass-produced jars were used in Gaya, Silla, Baekje, and Japan. The late phase witnessed the spread of grayish-blue stoneware production techniques from the Gimhae and Haman regions to other areas within Gaya territory, as well as to the neighboring kingdoms of Silla and Baekje. Following the diffusion of this technology throughout Gaya territory, vessels used for grave goods, such as mounted dishes, lids, and vessel stands, came to be produced in large quantities and jars lost their predominance. In addition, the diversity of Gaya pottery increased with the production of figurative vessels in the form of birds, houses, boat, wagons, mounted warriors, and more. Sue ware (須惠器, Jp. sueki), characterized by firing at high temperatures to produce hard, dense walls with good water retention, came to be produced in Japan from the fifth century. It is considered to represent a new technological system distinct from the existing pottery production methodology in Japan. Suemura in Osaka, the largest pottery production site of the era discovered so far in Japan, is a type site for early sue ware kilns. A strong influence of Gaya pottery can be observed at this site where ceramic artisans presumably crossed over to the Japanese Archipelago around 400 CE and created wares. Lee proposes that the Haman region supported an advanced pottery mass-production system in the fourth century and was the center of a large-scale distribution network for jars. This is significant because, until now, research on Allaguk (安羅國), an Early Gaya statelet that emerged in Haman from the third to the fourth century, had been based mainly on textual evidence due to the lack of burial sites and other forms of archaeological evidence. This study sheds light on how Allaguk could have developed into the second greatest power in Early Gaya. Professor Hong Bosik’s contribution to this volume was “The Ancient East Asian World and Gaya: Maritime Networks and Exchange.” Geumgwan Gaya was located around the natural harbor of the Gimhae Bay of the time, which had provided a major hub for goods flowing in from China and on to the Japanese Archipelago since the Samhan period. The importance of Old Gimahe Bay in ancient trade networks lasted into the fourth century. This is illustrated by the presence of artifacts from Goguryeo, Baekje, and Silla, Wa-type goods, items from China’s central plain region, and Xianbei-style artifacts at the burial complexes in Daeseong-dong and the Bonghwang-dong sites, both located in Gimhae. Advanced weapons and protective gear from Goguryeo and Xianbei appeared in the early fourth century in the Gimhae and Busan areas. This indicates that Geumgwan Gaya had established trading networks with groups to the north. Furthermore, Geumgwan Gaya artifacts dating to this period have been recovered from the southwestern coastal areas of the Korean Peninsula, as well as the Kinki region and northern Kyushu in the Japanese Archipelago. Allaguk in Haman and Garaguk (i.e., Dae Gaya) in Goryeong emerged as the cultural centers of Late Gaya. The Late Gaya statelets were generally quite active in adopting elements of Baekje culture, and associated sites have also yielded Silla and Goguryeo artifacts. Silla prestige goods dominate the Okjeon burial ground at Hapcheon, whereas Baekje prestige goods are common in the burial grounds of Wolsan-ri and Durak-ri in Namwon. This indicates that although the statelets of the Gaya Confederacy acknowledged the authority of the Dae Gaya ruler, their diplomatic authority had not been subsumed by Dae Gaya. Objects from Dae Gaya, Silla, and the Yeongsangang River system, as well as Wa items from Kyushu have been found in the areas of Uiryeong along the Namgang River drainage, Sancheong, and Goseong in the southern coastal region. The people interred in Wa-style tombs in these areas are presumed to have been individuals from Wa that settled in Gaya to serve as traders obtaining advanced technologies and goods from Gaya and conveying them to Wa. Hong identifies Gimhae and Busan as the centers of trade in the fourth century and Goryeong and Haman as the centers of trade from the fifth century onwards. It also recognizes Hapcheon, Namwon, and Goseong as possible sub-centers of trade, respectively, for the eastern, western, and southern regions of Gaya. This provides a case of archaeological evidence being applied to illustrate Gaya’s nature as a confederacy of statelets in great detail. The three articles examined above demonstrate how the archaeological discourse on Gaya culture has mainly been limited to the period of the fourth century and beyond. This is because the period prior to the third century has been commonly regarded as the “Samhan” era and distinct from the era dominated by the four ancient states of Goguryeo, Baekje, Silla, and Gaya. However, such a chronological framework is inappropriate for research on ancient history since, for example, the territorial boundaries of Samhan and Goguryeo are entirely exclusive, invalidating any temporal sequencing of Samhan followed by Goguryeo. The ways in which researchers have endeavored to construct and adopt a strict chronological framework are, of course, highly laudable. It is to be expected that the fourth to sixth centuries would be regarded as the key period for Gaya culture, and the manner in which the iron objects, pottery, and international exchanges of this period were examined in detail in the articles is to be applauded. Nevertheless, from the perspective of the author, whose research has focused on revealing the continuity of Gaya history, the fact that the area’s culture prior to the third century went unexamined leaves much to be desired. The nature of trade and the metal and pottery culture of Byeonhan, one of the three pillars of Samhan, should also be addressed in order to provide a richer and more multi-dimensional portrait of Gaya culture. Just as the development from Jinhan to Silla must be examined as an element of Silla history, the history of Byeonhan prior to the third century must be included to provide a proper overview of Early Gaya history. An Ordered, Diverse, and Advanced Culture What are the cultural characteristics of Gaya and the Three Kingdoms, the major players in ancient Korean history? Based in the vast lands of Manchuria and the northern reaches of the Korean Peninsula, Goguryeo grew into a center of trade in Northeast Asia. This led to the clear confidence that is reflected in Goguryeo’s grand and dynamic culture manifesting cultural elements from China and the steppe people of Central Asia. Baekje was heavily influenced by Nangnang (108 BCE–313 CE, Ch. Lelang) culture due to its geographical position in the northernmost portion of Mahan territory, but it also adopted and developed Goguryeo cultural elements such as stone-piled tombs. From the fourth century onward, influences from the aristocratic culture of China’s Southern Dynasties were steadily adopted, contributing to the elegant and sophisticated nature of Baekje culture. Silla’s culture emerged out of a convergence of the plain pottery culture indigenous to the southern regions of the peninsula and the Korean-type Bronze Dagger Culture of the northwestern region, which formed the common basis for Jinhan and Byeonhan culture. However, the wide-ranging social transformations that took place around the period of Silla’s emergence as an ancient state based on aid from Goguryeo led to a transition to a simpler, more utilitarian culture. There have been suggestions that Gaya was subject to the authority of Wa for hundreds of years, or alternatively, that it was controlled by Baekje for a period of time, or even that it had been annexed by Silla early in its history. It must be remembered that Gaya territory once covered approximately one-third of the southern portion of the peninsula, and its existence spanned seven hundred years from the first century BCE to the sixth century. During this time, Gaya fostered a culture that was distinct from that of Goguryeo, Baekje, or Silla. Although Gaya shares a common cultural foundation with Silla (i.e., Jinhan and Byeonhan culture), its geographic position was more fortuitous for the development of maritime activities. Gaya, therefore, played a major role in trade with Nangnang, which influenced Gaya culture to a greater extent compared to Silla. Chinese culture during the Han Dynasty (202 BCE–220 CE) focused on harmoniously merging the philosophy of the Hundred Schools of Thought (諸子百家) with the notion of Yin-Yang and Five Agents Theory (陰陽五行) within the framework of Confucianism. Having adopted this philosophical position, Nangnang material culture is characterized by a rational and ordered appearance. Following the successful adoption of this aspect of Nangnang, Gaya did not actively absorb other foreign cultural elements and instead independently developed its indigenous culture over a long period of time. This resulted in a culture that was orderly and conservative in nature. As discussed in the archaeological research described above on Gaya iron, pottery, and foreign objects, Gaya culture is diverse. The Nakdonggang River system flows through Gaya territory, which faces out towards the Korea Strait, allowing maritime trade to flourish. Therefore, not only did exchanges take place with neighboring Baekje and Silla, but diverse channels of exchange could also be established with Nangnang through which objects from China and the horse-riding peoples of Northeast China could be obtained. Furthermore, contact could be maintained with various Wa groups. In comparison, Baekje’s location along the southwestern coastline facilitated exchange with China’s Southern Dynasties. Silla’s position was isolated along the southeastern region of the peninsula, only facilitating exchanges with Goguryeo. Under these circumstances, Gaya’s cultural diversity can be understood in terms of the international nature of its trade networks. Gaya did not prioritize elevating its political standing, but rather was open to adopting in groups from various regions and trading in their goods. Gaya was renowned for its iron ingots, which were also used as a form of currency, and the confederacy’s international orientation is reflected by the fact that they actively traded with the peoples of Nangnang, Daebang (early 3rd century–314, Ch. Daifang), Han, Ye, and Wa, and supplied them all with iron. What was the basis of driving Gaya to become an active leader of maritime trade? The key factor was the markets of the Japanese Archipelago (which had a large population even at the time) located at the opposite end of the maritime route that began at the mouth of the Nakdonggang River. The Japanese had to pass through Gaya if they wished to overcome the isolation of their surroundings. In addition, the Japanese Archipelago only became self-sufficient in iron from the sixth century onwards. Given that iron was a key factor in the establishment of ancient civilizations, the Japanese had no other option but to import iron from Gaya. In addition to iron, Gaya provided the communities on the Japanese Archipelago with other needed advanced cultural elements. By exporting them, Gaya obtained great economic benefits. It is possible to maintain that the elements of ancient material culture such as metal weapons, armor, horse gear, sue ware, and gold and silver decorations were mostly introduced into the Japanese Archipelago via Gaya. In order to reach the profitable Japanese markets, Wiman Joseon in the second century BCE, Nangnang from the first century BCE to the third century CE, and Baekje in the fourth to fifth centuries CE all had to depend on Gaya intermediaries. This is the reason why, at least up until the late fifth century, Gaya was able to maintain its cultural superiority over Wa on the Japanese islands. However, with the commencement of iron production in the Kinki region of Japan in the sixth century and Baekje’s efforts to interact directly with the communities of the Japanese Archipelago for their own political purposes and exclude Gaya, a sudden decline in power occurred. One reason for this can be found in the fact that Gaya did not provide the philosophical culture, such as Confucianism or Buddhism, that communities of the Japanese Archipelago needed to stabilize their institutions. Gaya’s growth had been based on flourishing trade, and centralized political power was a secondary concern. It is this aspect of Gaya society that was the key factor in the confederacy’s ultimate fall to Silla.

Gaya (42–562) has long been denied its proper place in Korean history due to the “Three Kingdoms Period” conception that views ancient Korean history exclusively in terms of the Goguryeo (37 BCE–668 CE), Baekje (18 BCE–660 CE), and Silla (57 BCE–935 CE) Kingdoms. Furthermore, the view of “Imna Ilbon-bu” (任那日本府) portrays Gaya territory as actually having been under the control of forces from ancient Japan. However, evidence from the excavation of numerous Gaya tombs that has been taking place since the 1970s in order to allow land development has clearly established a unique and independent Gaya culture which in many ways rivaled that of Silla. Centered in what is now Gyeongsangnam-do Province at the southernmost reach of the Korean Peninsula, for most of its history, Gaya existed as a confederation of statelets and then took on the form of an early ancient state in its final stages. Gaya culture emerged early in the first century BCE and survived for nearly seven centuries (Kim Taesik 2002, 78–79). No systematic chronology of Gaya exists since the polity did not record its own history. The historical data on Gaya, therefore, consists only of fragmentary records extracted from the writings of neighboring states. These sources are as follows: Chinese sources, including “Account of the Eastern Barbarians” (東夷傳) in the “Book of Wei” (魏書) from Sanguozhi (三國志, Records of the Three Kingdoms) and “Account of Gara” (加羅國傳) and “Illustration of Envoys Presenting Tribute at the Liang Court” (梁職貢圖) from Nanqishu (南齊書, Book of Southern Qi); Korean sources, including Samguk sagi (三國史記, History of the Three Kingdoms), Samguk yusa (三國遺事, Memorabilia of the Three Kingdoms), the Stele of King Gwanggaeto, and Sinjeung dongguk yeoji seungnam (新增東國輿地勝覽, Revised and Expanded Edition of Survey on the Geography of Joseon); and Japanese sources, such as the Nihon shoki (日本書紀, The Chronicles of Japan). Among these sources, three are particularly essential to our understanding of Gaya history. Consisting of sixty-five chapters written in the annals-and-biographies format, Sanguozhi is a historical text covering the period of the Chinese Three Kingdoms (220–280) of Wei, Shu, and Wu that was compiled by a Western Jin Dynasty (265–316) official named Chen Shou (陳壽, 233–297). The rich and detailed passages concerning relations between the Samhan polities that appear in the “Account of the Eastern Barbarians” section of the “Book of Wei” make it the most important written source for understanding the nature of Gaya up to the third century. Spanning fifty chapters also written in the annals-andbiographies format, Samguk sagi, is a historical text covering Korea’s Three Kingdoms and Unified Silla periods that was compiled by the Goryeo Dynasty (918–1392) official Kim Busik (金富軾, 1075–1151) in 1145 after his retirement. The “Annals of Silla” (新羅本紀) is scattered with numerous references to Gaya that have been used to establish the conventional chronology for Gaya’s history. The “Book of Music” (樂志), “Book of Geography” (地理志), and “Biographies” (列傳) sections also contain fragmentary records related to Gaya. Nihon shoki is a historical text consisting of thirty chapters in a chronological format covering the ancient period of Japanese history. It was completed in 720 by Prince Toneri (676–735), a son of Empress Genshō (r. 715–724). Records from the reigns of Emperor Keitai (r. 507–531) and Emperor Kinmei (r. 539–571), which contain information obtained both directly and indirectly from the “Annals of Baekje” (百濟本記), provide plentiful accounts on the respective members of the Gaya Confederacy around the early to mid-sixth century. However, these records can, in places, be severely distorted by the Japanese and Baekje perspectives from which they were derived. The available archaeological data concerning Gaya has mostly been recovered from tomb sites in today’s Gyeongsangnam-do and Gyeongsangbuk-do Provinces in South Korea, where the bulk of Gaya territory was concentrated. Many of these Gaya tombs were elite burials, and the practice of interring objects used in daily life along with the deceased has made it possible to compile important information on Gaya’s history and culture. Gaya tomb sites of particular importance are the burial grounds of Yangdong-ri and Daeseong-dong in Gimhae, the Bokcheon-dong burial ground in Busan, the Marisan burial ground in Haman, and the Jisan-dong burial ground in Goryeong. Establishment and Development of Early Gaya First Century BCE to Second Century CE An agricultural society characterized by dolmens and pottery with undecorated surfaces flourished in the Nakdonggang River Basin of southern Korea during the Bronze Age and Early Iron Age. However, the geographical conditions of this region nestled in the southeastern corner of the peninsula were unfavorable to the introduction of more developed elements of civilization. As a result, the degree of social organization and the ability to craft metal objects were less developed there compared to other regions around the Korean Peninsula. Fortunately for the residents, an opportunity to overcome this geographic handicap came from the outside. Gojoseon located in the northwestern region of the peninsula fell in 108 BCE under attacks from an army dispatched by Emperor Wu (r. 141–87 BCE) of China’s Han Dynasty (206 BCE–220 CE). The chaos of war produced a large number of refugees, and a large number of them appear to have made their way to southeastern Korea via land or sea routes. Sites such as the Daho-ri burial ground in Changwon, which was established in the first century BCE, demonstrate the emergence of new power groups distinct from the indigenous groups of the region. The definitive tomb at this site, Dahori Tomb No. 1, is a wooden coffin tomb that yielded black and brownish pottery with undecorated surfaces, Gojoseon-style lacquerware, and metal objects such as Korean-type bronze daggers, bronze spearheads, bronze bells, bronze mirrors, iron daggers, and iron axes (Fig. 1). The pottery from this burial shows no significant changes from earlier examples, but the lacquerware and metal objects are novel examples not previously seen in this region and reflect the introduction of more evolved culture elements. Fig. 1. Artifacts from Daho-ri Tomb No. 1 in Changwon. L. 61.1 cm (Left Scabbard), L. 60.9 cm (Right Scabbard). National Museum of Korea The area that is now Gyeongsangnam-do Province witnessed an increase in the number of wooden coffin tombs from the first century BCE onwards. The pottery recovered from these wooden coffin tombs shows the transition from brownish wares with decorated surfaces to grayish wajil pottery (瓦質土器). The impetus for the manufacture of this wajil pottery was the adoption of new production technologies by local craftspeople. The Gaya foundation myth presented in the “Garak Gukgi” (駕洛國記) section of Samguk yusa relays how the ‘nine gan’(九干), the chiefs of the indigenous polities of the Gimhae region, selected as their king Suro—a man from an external lineage said to have descended from the heavens—and established the state. This myth can be viewed as depicting the complex interactions of cultural elements observed in the archaeological record. A transition in burial practices during the middle and late second century can be observed at the Yangdong-ri burial ground, as reflected by the use of wooden chamber tomb that feature greater internal space and an increase in the number of iron objects found among the grave goods. Yangdong-ri Tomb No. 162 is a typical example of an early-phase wooden cist burial (Fig. 2). Individuals interred in the Yangdong-ri burial ground are likely to have been the heads of Guyaguk (狗邪國)—later became Geumgwan Gaya—which controlled both economic wealth and political power in the area. Fig. 2. Yangdong-ri Tomb No. 162 in Gimhae Third Century The Gyeongsang region in the early third century was home to the twelve statelets (國, guk) of Jinhan including Saroguk (斯盧國, which later became Silla) and the Byeonhan twelve statelets including Gayaguk — also known as Guyaguk, Gaya Hanguk (加倻韓國), or Geumgwan Gaya. Archaeological evidence indicates that these guks were peer-level polities, and even the comparatively advanced statelets of Gayaguk and Saroguk had yet to gain hegemony over other statelets. Long-distance trade within northeast Asia at the time was carried out through the Nangnang (108 BCE–313 CE, Ch. Lelang) and Daebang (early 3rd century–314, Ch. Daifang) Commanderies, which served as intermediaries connecting China, the Korean Peninsula, and the Japanese Archipelago. In the southern part of the Korean Peninsula, Gayaguk was located at the mouth of the Nakdonggang River in what is now Gimhae and served as the most active intermediary in this trade in iron and other advanced items of civilization. This suggests that the twelve statelets of Byeonhan had integrated with Gayaguk at the political center and formed a coalition of Byeonhan statelets, otherwise known as the Early Gaya Confederacy (Kim Taesik 1993, 66) (Map 1). Map 1. Location of the Statelets Forming the Early Gaya Confederacy The confederacies of the southern part of the Korean Peninsula, namely the one controlled by the chief of Mahan, the one ruled over by the king of Jinhan, and Sinmiguk (新彌國), all actively conducted trade as independent entities. The political growth of these Samhan heads went hand in hand with the decline of the governing authority of the Western Jin Dynasty in China (Yoon Yonggu 2004, 138). In the late third century, the center of power in Gimhae was relocated from Yangdong-ri and its environs in Jucheon-myeon to the area around the Daeseong-dong burial ground, which is located in the center of present-day Gimhae. The situation at the time is well illustrated by Tomb No. 29 of the Daeseong-dong burial ground (Fig. 3). Compared to the tombs in the burial grounds found in the surrounding areas, the Daeseong-dong tombs are relatively larger in scale and contain a greater volume of grave goods. This is taken to illustrate a power imbalance that existed between the group associated with the Daeseong-dong burial ground and neighboring groups. Fig. 3. Daeseong-dong Tomb No. 29 in Gimhae The Jinhan cultural sphere with Saroguk as its political center and the Byeonhan cultural sphere, centered on Guyaguk, were originally quite similar, but a clear distinction in local pottery styles emerged in the late third century along the border between what is now Busan and Ulsan. This period also witnessed the deposition of large quantities of iron weapons as grave goods, an expansion of the social classes that could access such weapons, and the appearance of iron armor. This indicates that relations between the statelets of the Gyeongsang region had become strained in the late third century and warfare had increased, resulting in a consolidation of the Jinhan statelets around Saroguk and the Byeonhan statelets around Gayaguk, and the two polities’ respective development into political authorities (Kim Taesik. 2010, 296). Fourth Century The Nangnang and Daebang Commanderies were destroyed in the early fourth century by the Goguryeo Kingdom, a state based in the northern portion of the peninsula. This had significant implications for Gayaguk, which had developed in the southeastern part of the peninsula by facilitating longdistance trade with the Daebang Commandery. As the maritime network through which Gayaguk had obtained advanced culture weakened, the eight Posang statelets (浦上八國, eight polities located in coastal port regions) who no longer acknowledged Gayaguk hegemony mounted an attack. Gayaguk fended them off by successfully rallying the polities located in the hinterlands of its riverine trade routes (in the middle and upper reaches of the Nakdonggang River) and those along the southern reaches of the eastern coast. Gayaguk’s victory does not seem to have been complete, however: Starting in the early fourth century, two distinct ceramic styles for mounted pottery and other vessel types came to be established in the broader region, respectively to the east and west of the Masan Bay area (Ahn Jaeho and Song Gyehyun 1986, 53). This appears to indicate an east-west divide in the Gaya Confederacy that emerged out of the war of the eight Posang statelets. Following the erasure of Daebang from its existing trade network in the late fourth century, eastern Gaya, centered in modern Gimhae, had no option but to focus on trade with Wa (倭) in the Japanese Archipelago. Daeseong-dong Tombs Nos. 2, 13, and 23, all dating to the late fourth century, yielded bronze spiral-shaped shield ornaments of Japanese origin (Kyungsung University Museum 2000, 183) (Fig. 4). On the other hand, the Jinhan region grew increasingly closer to Goguryeo, as demonstrated by the appearance there of Goguryeo-style objects. Fig. 4. Bronze Spiral-shaped Shield Ornaments. (top and center) Excavated from Daeseong-dong Tomb No. 13 in Gimhae. Gaya, 4th Century. D. 12.0 cm. Jinju National Museum. (bottom) Excavated from Daeseong-dong Tomb No. 2 in Gimhae. Gaya, 4th Century. D. 6.1 cm. Jinju National Museum The tombs of the Bokcheon-dong burial ground, situated in between the Gimhae and Gyeongju areas, were characterized by a juxtaposition of elements from both Silla and Geumgwan Gaya culture until the middle of the fourth century. Silla-style artifacts disappear entirely, starting with tombs from the late fourth century, however, while Geumgwan Gaya-style artifacts continued to be interred (Bokcheon Museum 2004, 90). This has been taken to indicate an expansion of Geumgwan Gaya’s influence and power, which was supported by alliances with Baekje and Wa. In other words, it seems likely that Gaya was pressuring Silla with help from the Wa army. Around the same period, King Geunchogo (r. 346–375) of Baekje attacked Goguryeo and gained control over territories in what is now Hwanghae-do Province. The situation quickly reversed following ascension of King Gwanggaeto (r. 391–412) to the Goguryeo throne in 391. By 396, Goguryeo’s might had grown to the extent that it was able to seize 58 Baekje fortresses in that year alone. King Gwanggaeto also sent an army of fifty thousand infantry and cavalry troops to assist Silla in the year 400. Upon their arrival, the Goguryeo units chased the Wa forces from Silla territory and back to Imna Gara (任那加羅), referring to the Gimhae region, the center of the Gaya Confederacy. The fact that the Wa troops fled all the way from Gyeongju to the Gimhae area indicates that these Wa forces had been dependent upon Imna Gara’s support. When the pursuing troops of the Goguryeo-Silla alliance reached Imna Gara, several fortresses surrendered. This event eventually allowed Silla to overcome Gaya in the competition for hegemony over the Gyeongsang region. Another result is that Baekje’s trade network with Wa, which had utilized the Gaya regions as intermediary nodes, could no longer be maintained. The cessation of large-scale tomb construction in Gimhae in the early fifth century—as evidenced in the Daeseong-dong burial ground—can be seen as an indication of the sudden collapse of the Gaya polity in this area. Growth and Collapse of Late Gaya Fifth Century Gaya polities to the east of the Nakdonggang River, in areas such as modern-day Changnyeong and Busan, surrendered without resistance to the allied Goguryeo and Silla forces that had vanquished Imna Gara. Gaya polities located to the west of the Nakdonggang River, on the other hand, continued on with little disturbance. In particular, the polities in the Goryeong and Hapcheon areas in the mountainous interior of the Gyeongsang region, which had previously been a backwater region, began to slowly develop from the early fifth century onwards. The Banpaguk (伴跛國) polity in the Goryeong region accepted migrant artisans from the Gimhae area specializing in pottery and iron production and began to exploit the local iron resources of the Yaro (冶爐) area at the foot of Mt. Gayasan to develop its iron production system (Kim Taesik 1986, 28). From the middle of the fifth century, Banpaguk began to dominate trade with both Baekje and Wa. After changing its name to Dae Gaya, it organized the surrounding polities to once again embody a Gaya Confederacy. The fact that Dae Gaya’s mythological progenitor King Ijinasi was believed to be the son of the spirit of Mountain Gayasan and the elder brother of King Suro of Geumgwanguk (金官國), also called as South Garaguk (加羅國), seems to suggest that political forces based in Goryeong had become the leaders of the Late Gaya Confederacy. Dae Gaya expanded westwards in the late fifth century, crossing the Sobaek Mountain Range to annex the polities in Namwon and Imsil (in present-day Jeollabuk-do Province) and Yeosu, Suncheon, and Gwangyang (in present-day Jeollanam-do Province) and consolidate their territories. The Late Gaya Confederacy at its zenith consisted of 22 statelets organized around Dae Gaya in Goryeong (Map 2). Map 2. Location of the Statelets Forming the Late Gaya Confederacy In 479, the king of Dae Gaya—King Haji (r. unknown), also referred to as the King of Gara (加羅王)—received the title Bogukjanggun Bongukwang (輔國將軍本國王), meaning “General Defending the State, King of Gara” after presenting a tribute to the Southern Qi Dynasty (479–502) court. It is highly likely that the final resting place of King Haji is Tomb No. 44 in the Jisan-dong burial ground in Goryeong (Fig. 5). The largest tomb in this burial ground is Tomb No. 47 (presumed to be the Royal Tomb of King Geumrim) (Fig. 6) with a mound measuring 50 meters in diameter, demonstrating the grandeur of Dae Gaya royal tombs. Fig. 5. Jisan-dong Burial Ground in Goryeong Fig. 6. Jisan-dong Tombs Nos. 49, 48, and 47 in Goryeong (from left to right) Around the same time, a consensus emerged in Silla regarding the need to distance itself from Goguryeo influence, which ultimately backfired and resulted in the capture of seven Silla fortresses by the Goguryeo army in 481, including Homyeongseong Fortress in present-day Yeongdeok in Gyeongsangbuk-do Province. Goguryeo forces then marched on Mijilbu Fortress in now Pohang in Gyeongsangbuk-do Province, but Silla was able to successfully defend it with support from troops dispatched by Gaya and Baekje. The state of affairs on the Korean Peninsula in the late fifth century stabilized with the forces of Baekje, Silla, and Gaya allied to defend against southern expansion by Goguryeo. 510s In 512, in an effort to facilitate direct trade with Wa, King Muryeong (r. 501–523) of Baekje took control over the eastern Jeolla region statelets of Sangdari (上哆唎), Hadari (下哆唎), Sata (娑陀), and Moru (牟婁), which were formerly within the sphere of influence of Gaya. A large quantity of late fifth-early sixth century Goryeong-style pottery was found in burial ground at Wunpyeong-ri in Suncheon where believed to be associated with the Moru polity, whereas a transition to Baekje-style stone-lined tombs can be observed around early sixth century in the burial grounds of present-day Yeosu and Suncheon (Lee Donghee 2007, 103). This seems to suggest that these four locales in the southern Jeolla region were originally Gaya statelets that later became incorporated into Baekje territory. In 513, Baekje also made inroads into the Gaya statelet of Gimun (己汶, located in what is now Namwon, Imsil, and the southern portion of Jangsu) under the pretext of facilitating trade with Wa. Baekje thus appears to have emerged victorious in a territorial struggle with Dae Gaya and gained control of all of the eastern Jeolla region. As a result, the Sobaek Mountain Range became established as a natural boundary between Gaya and Baekje, and Dae Gaya lost access to the trade route along the Seomjingang River leading to the southern coast. In 515, Dae Gaya constructed fortresses at Jatan (子呑) and Daesa (帶沙), polities located respectively in presentday Jinju and Hadong, to defend against both Baekje and Wa. Fortresses were also constructed at Iyeolbi (爾列比) and Masubi (麻須比), located respectively in Burim-myeon in Uiryeong and Yongsan-myeon in Changnyeong, in order to pressure Silla. Dae Gaya’s territory at the time was centered around Goryeong in Gyeongsang-do Province. It extended westwards to reach the Sobaek Mountain Range and Mt. Jirisan and southwards to the Namgang River. This territory coincides broadly with the distribution of Goryeong-style pottery around the beginning of the sixth century (Park Cheonsu 2004, 236–237). It, therefore, appears that Gaya had managed to successfully consolidate its control over this territory and developed into an early ancient state by the 510s at the latest (Kwon Haksoo 2003, 86). The fact that Dae Gaya fortress construction did not take place south of the Namgang River, such as in Haman, Gimhae, or Goseong, seems to indicate that although the polities based in these areas were also members of the Gaya Confederacy, they had managed to remain relatively independent of Dae Gaya’s influence. 520s–530s In order to secure a route to the southern coast along the Nakdonggang River, Dae Gaya sought an alliance with Silla through marriage. King Inoe (r. unknown) of Dae Gaya proposed a marriage alliance in 522, in response to which Silla’s King Beopheung (r. 514–540) sent a sister of the Silla nobleman Ichan (伊湌, second highest official rank in Silla) Bijobu (比助夫) to seal the alliance. The Silla princess dispatched to Dae Gaya soon gave birth to a son, Crown Prince Wolgwang, who became the last king of Dae Gaya. This alliance fell apart a few years later, however, following King Beopheung’s original design, and signs of division appeared within the Gaya Confederacy. Taking advantage of this situation, Silla threatened Takgitanguk (喙己呑國), a polity located in present-day Yeongsan-myeon in Changnyeong, in 529 and forced its surrender. The southern polities of the Gaya Confederacy consequently lost faith in Dae Gaya and strengthened their allegiance to Allaguk (安羅國), a Gaya polity based in what is now Haman. Baekje attacked Allaguk in 531 in order to preempt an invasion of southern Gaya by Silla and took possession of Geoltakseong Fortress. Silla responded by annexing Geumgwanguk in Gimhae in 532, but still allowed its royal family preferential treatment. In 534, Baekje advanced into the northern region of Taksunguk (卓淳國), located in present-day Changwon, and built a fortress and stationed troops at Guryemora (久禮牟羅), located in present-day Chilwon-myeon in Haman. Geumgwanguk and its neighboring polities were subsequently annexed by Silla and reorganized into prefectures and counties of the Silla Kingdom. Allaguk and the Gaya polities located to its southwest, on the other hand, fell under the political influence of Baekje. Around 538, the king of Taksunguk summoned the Silla army in order to counter pressure from Baekje and willingly submitted to annexation by Silla. Silla went on to expel the Baekje forces stationed at Guryeosanseong Mountain Fortress. The northern Gaya polities under the influence of Dae Gaya adopted a pro-Baekje stance in response to Silla’s betrayal and to counter the independent attitudes of the southern Gaya polities. However, Baekje’s increasing authority led to a weakening of Dae Gaya control, and the degree of its consolidation lessened to that of a confederation of statelets. On the other hand, the southern Gaya polities came to form an autonomous confederation led by Allaguk. Allaguk was able to adopt a relatively independent attitude toward Baekje based on its close relationship with Silla and Wa and emerged as a central power within the Gaya Confederacy to rival Dae Gaya. This ultimately led to a north-south division within the confederation. 540s–560s In the middle of the sixth century, the Gaya Confederacy sought ways to maintain its independence while remaining prepared for invasions by Baekje and Silla. Despite their alliance against southward advances by Goguryeo, Baekje and Silla still competed against each other over Gaya territory. Faced with this situation, the Late Gaya Confederacy—although divided into the two factions led by Dae Gaya in Goryeong to the north and Allaguk in Haman to the south—formed a committee consisting of seven or eight officials (執事) to undertake diplomatic negotiations with both Baekje and Silla. However, Allaguk sent secret communications to Goguryeo in 548 and requested that its forces attack Baekje’s Doksanseong Mountain Fortress, in Yesan, Chungcheongnam-do Province. This battle was won by Baekje, and Allaguk lost the confidence of the members of the confederation when its betrayal was revealed. As a result, the Gaya Confederacy became subordinate to Baekje around 550, accepting Dae Gaya’s preference in the matter. Having increased his power through the subordination of Gaya, King Seong (r. 523–554) of Baekje established an alliance with Silla and recovered his kingdom’s former territory in the Hangang River basin in 551 in an attack on Goguryeo’s southern reaches. However, the 120-year-old Silla-Baekje alliance soon came to an end due to conflicts over the lower reaches of the Hangang River. In order to reclaim this region, Baekje’s King Seong led an army comprised of troops from Baekje, Gaya, and Wa to attack Silla. This allied force suffered heavy losses at a battle at Silla’s Gwansanseong Mountain Fortress in Okcheon, Chungcheongbuk-do Province, resulting in great military and political confusion in Baekje, which in turn made it more difficult for the Gaya Confederacy—now with a heavy reliance on Baekje—to maintain its independence. Silla raised a large army and conquered Dae Gaya in 562. Around this period, ten polities of the Gaya Confederacy fell to Silla: Allaguk, Saigiguk (斯二岐國), Daraguk (多羅國), Jolmaguk (卒麻國), Gojaguk (古自國), Jataguk (子他國), Sanbanhaguk (散半下國), Geolsonguk (乞飡國), and Imnyeguk (稔禮國). Gaya Culture Industry Gaya’s economy consisted primarily of fishing, farming, iron production, and long-distance exchange. Shell middens, which can shed light on everyday life, began to make an appearance in Busan, Changwon, Gimhae, and Yangsan around the beginning of the first millennium. Artifacts that have been recovered amidst the shells in these middens include spindle whorls, fishing-net sinkers, bone implements, pottery, iron hand knives, iron sickles, iron arrowheads, and fishhooks. Along with the oyster, clam, and conch shells, the middens have also yielded remains of food, including rice, barley, wheat, legumes, and millet (Donga University Museum 1981). The contents of these shell middens indicate that the communities of this period relied on farming and fishing. Iron farming tools such as iron hand knives, axe-heads, hoes, and weeding plows came to be used in what is now Gyeongsangnam-do Province after the first century BCE. Farming at the time was carried out in an intensive manner, with rice being sown in paddy fields irrigated through channels. Iron sickles and iron spades for irrigation work were deposited as grave goods in small Gaya tombs from the fourth to fifth centuries. In contrast, cast iron axe heads, U-shaped shovel blades, and iron rakes (with finds being limited to one or two per item) have only been discovered in medium- to large-sized tombs (Lee Hyunhye 1991, 57–68). A wide range of crops was grown in Gaya territory, including rice, common millet, barley, foxtail millet, wheat, legumes, and adzuki beans. In addition, the animal bones recovered from archaeological investigations have revealed the presence of domesticated dogs, swine, cattle, horses, and chickens, along with wild deer, roe deer, and boar. Bronze items, including daggers, spearheads, mirrors, and bells, began to be used by Gaya communities after the first century BCE, alongside iron weapons such as daggers, arrowheads, spearheads, and dagger-axes. The volume of iron objects deposited as grave goods increased dramatically from the late second century CE, accompanied by the appearance of longswords, plowshares, rakes, and horse bits, and a diversification of the existing iron axe heads. It was around this time that ‘flat iron axe heads’ (板狀鐵斧), which lack a sharpened blade and could not have served any utilitarian function, appeared. These axe heads were, in fact, iron ingots (Song Gyehyun 1995, 131–133) prepared in advance for the mass production of iron objects and were exported to Nangnang, Han, Ye, Wa, and other regions. Ritual implements (such as highly decorated iron spearheads or weapon-like saw knives), vertical plate armor, and iron horse armor were new types of iron objects that came to be used after the late third century. In the lower reaches of the Nakdonggang River, iron ingots standardized in terms of weight and dimensions (similar to coinage) came to be produced on a large scale after the late fourth century. The large iron ingots and cast iron spades that have been recovered from early fifth century tombs in Osaka and Hyogo, Nara, and Shiga Prefectures in Japan’s Kinki region originated in the lower reaches of the Nakdonggang River (Azuma Ushio 2002, 33). From the perspective of maritime trade, Gayaguk (based in Gimhae) was the center of a long-distance trade network that linked the Nangnang Commandery with the statelets along the western and southern coastal regions of the Korean Peninsula and Wa. Gaya exported iron to Nangnang, from which it imported advanced items of civilization. Iron and advanced items of civilization were sent to Wa, and human labor and military service were returned as payment (Suzuki Yasutami 2002, 15). Among the artifacts discovered at sites within Gaya territory, Wangmang coins (王莽錢), wuzhu coins (五銖錢), bronze tripods, bronze mirrors with inner flower designs (內行花文鏡), and bronze TLV mirrors reflect the trade with Nangnang. Artifacts associated with nomadic horse-riding cultures, such as bronze and iron cauldrons, tiger-shaped belt buckles, and swords with curved blades, have also been discovered at Gaya sites (Shin Gyeongcheol 2000, 59). These items were likely obtained through coastal maritime routes. Artifacts such as haji ware (土師器, Jp. hajiki) represented with double-rimmed pottery, broad bronze spearheads, bronze spiral-shaped shield ornaments, jade arrowheads, and spindle-whorl-shaped stone objects illustrate the exchanges taking place with Wa. Trade was not as active during the Late Gaya as it was during Early Gaya; nevertheless, trade with Wa was carried out after the middle of the fifth century by the polity based in Goryeong. Tombs of high-level individuals found throughout the Japanese Archipelago have yielded gilt-bronze crowns featuring large flame-shaped decorations, iron spearheads with octagonal sockets, s-shaped horse bits, gold belt buckles with openwork dragon patterns, earrings with hanging ornaments, dagger-shaped horse harness pendants, and horse armor, all of which are prestige goods associated with Dae Gaya (Park Cheonsu 1996, 85). Goryeong-style pottery, such as long-necked lidded jars, has been found throughout the Kyushu region of Japan and in the coastal areas of the Seto Inland Sea (Sadamori Hideo 1997, 174). Religion Regarding Gaya and Buddhism, it is recorded that the Pasa Stone Pagoda was brought to Gimhae from the Ayuta State of India by Queen Consort Heo in 48 CE (Fig. 7). It is believed, however, that the association with the Ayuta State was a later addition to the myth of Queen Consort Heo’s marriage appended in order to establish a link with Buddhism, very likely made around the time that Wanghusa Temple (wanghu means ‘queen consort’) was established during Silla’s Middle Period (654–780). In addition, the ‘Pasa Stone Pagoda’ myth relating that Queen Consort Heo’s ship carried a stone pagoda in order to ensure safe passage on the seas, as well as the story that this stone pagoda was the one later located at Hogyesa Temple (虎溪寺), were both Goryeo-era inventions. Fig. 7. Pasa Pagoda Situated in Front of the Tomb of Queen Consort Heo The royal genealogy of Dae Gaya features personages such as Queen Mother Jeonggyeon, the mother of King Ijinasi (the mythological founding father of Dae Gaya), and Crown Prince Wolgwang, who was born of the marriage alliance between the Gaya King Inwae and his Queen Consort from Silla. Queen Mother Jeonggyeon and Crown Prince Wolgwang are terms that appear in early scriptures of Hinayana Buddhism. Silla royalty and aristocracy from the early sixth century are said to have had names associated with Buddhism (Kim Cheoljoon 1990, 148) derived from myths associated with Siddhartha Gautama and reflecting the practices and status of Hinayana Buddhism. It is therefore likely that the use of Buddhist-influenced names by members of Dae Gaya’s royal family occurred around the final period of Gaya’s history after the marriage alliance between Gaya and Silla was established in 522. The stone chamber mural tomb of Goa-dong in Goryeong, which is believed to be the last Dae Gaya tomb ever constructed, features a tunnel-shaped ceiling structure similar to that of the brick chamber tomb of Songsan-ri in Gongju. The lotus blossom pattern decorating the ceiling shares stylistic elements with the motifs featured in the mural of the Neungsan-ri mural tomb in Buyeo (Jeon Hotae 1992, 171) (Fig. 8). It appears, therefore, that Buddhism was introduced into Gaya through Baekje. This demonstrates that Buddhism had been introduced to Dae Gaya at least by the last stage of the polity’s existence and that its original foundation myth was given a Buddhist bent by adding figures such as Queen Mother Jeonggyeon and Crown Prince Wolgwang in an attempt to shore up its eroding royal authority. Fig. 8. Lotus Pattern on the Ceiling of the Burial Chamber in the Goa-dong Mural Tomb in Goryeong Customs It is recorded in the “Account of the Eastern Barbarians” in Sanguozhi that the Samhan people placed a high value on beads, which they used to adorn their bodies, but did not treasure gold, silver, or silk. Their taste in luxury items was possibly limited by their shamanistic worldview. However, after the fifth century, precious metals such as gold, gilt-bronze, and silver came to be used to decorate not only personal ornaments such as crowns, necklaces, and rings, but also items such as longswords, horse gear, and coffins. It is, therefore, possible to assume that a wealthy aristocratic class that showed a preference for precious metals had come to be established in the Gaya region by the fifth to sixth centuries. Sacrificial practices in funerary contexts involved the killing and burial of humans or animals on behalf of the deceased. Numerous burials featuring evidence of sacrifices have been found throughout the Gaya region. The Daeseong-dong burial ground in Gimhae has yielded burials dating to the third and fourth centuries, which contained the remains of two to five sacrificed individuals. The Marisan burial ground in Haman has yielded fifth-century burials that contained the remains of from one to five sacrificed individuals (Kim Segi 1997, 109) (Fig. 9). In Goryeong’s Jisan-dong Tomb No. 44, which dates to the late fifth century, three centrally positioned large-scale stone cists surrounded by 32 smaller stone cists were found within a single burial mound (Fig. 10). Of the 32 smaller stone cists, which appear to have been sacrificial burials, 22 of them collectively yielded the remains of 24 human sacrifices (Kwon Haksoo 1992, 41) (Fig. 11). Fig. 9. Marisan Burial Ground in Haman Fig. 10. Layout of Jisan-dong Tomb No. 44 in Goryeong. Gaya, 6th century This is the largest Gaya tomb containing sacrificial burials. Fig. 11. Sacrificial Burial from Daeseong-dong Tomb No. 88 in Gimhae How should this feature of Gaya society be understood? Sacrificial practices appear to be associated with a worldview that regarded the king as the descendent of a heavenly god. The fact that royal authority at the time was dependent upon the semi-voluntary obedience of the populace must also be taken into consideration. Therefore, the widespread practice of human sacrifice in Gaya society can be considered a result of a transitional situation in which a centralized ruling authority had yet to be institutionalized, but the power of the heads of the Gaya Confederacy and its member statelets had been significantly strengthened.

White porcelain decorated with underglaze iron-brown (鐵畵白磁, cheolhwa baekja) is painted using an iron pigment the main component of which is ferric oxide (Fe2O3). This pigment, which was called seokganju (石間朱, red ocher) during the Joseon period (朝鮮, 1392–1910), is created by refining the pigment after combining it with certain components of clay and glaze, including feldspar, quartz, and limestone. Iron has long been used as a pigment in a variety of ways, as it is one of the decorative colorants most available in nature and readily usable. The official kilns in Gwangju, Gyeonggi-do Province that produced Joseon court wares continuously used iron oxide to decorate white porcelain. White porcelain decorated with underglaze iron-brown from the official court kilns have been much studied. Most investigations of this type of white porcelain have focused on the stylistic changes apparent among extant examples, particularly those from the seventeenth century. However, since white porcelain decorated with underglaze iron-brown was made throughout the entire period when the official kilns were in operation, limiting studies to seventeenth-century porcelain wares restricts a more comprehensive understanding of the characteristics of this type of ware. Moreover, recent discoveries of numerous porcelain wares, including white porcelain decorated with underglaze iron-brown, from excavations and archeological field surveys of Joseon white porcelain kilns have created a need for research into these newly discovered ceramics. This paper covers the period from 1466, when the Joseon royal court established a cluster of official court kilns in Gwangju, Gyeonggi-do Province as a branch office, known as the Bunwon (分院) of the Saongwon (司饔院, Office of Royal Cuisine), through 1883 when the official kilns were privatized. This timeframe is divided here into three periods based on changes in the characteristics of white porcelain decorated with underglaze iron-brown as identified through a study of historical sources as well as comparative analysis of porcelain excavated from kilns and surviving heirlooms (傳世品, jeonsepum). Phase I spans from the latter half of the fifteenth century to the sixteenth century; phase II runs from the seventeenth century to the first half of the eighteenth century; and phase III begins in the latter half of the eighteenth century and lasts into the nineteenth century. Referencing this timeframe, this study explains how the characteristics of white porcelain decorated with underglaze iron-brown from official kilns changed over time. This study also reexamines discussions on when white porcelain decorated with underglaze iron-brown began to be produced by analyzing historical records that describe factors which stimulated its production, such as the supply and demand of iron oxide pigment and the visual impact of foreign polychrome wares. Furthermore, it draws upon the outcomes of recent excavations and archeological field surveys to scrutinize Joseon people’s perception of white porcelain decorated with underglaze iron-brown by analyzing the types of wares unearthed and the distribution of their manufacturing locations. In so doing, this research will show how the nature of white porcelain decorated with underglaze iron-brown from official kilns developed and changed over time. The Manufacture of White Porcelain Decorated with Underglaze Iron-brown at Official Kilns and the Initiation of Its Production Background of the Manufacture of White Porcelain Decorated with Underglaze Iron-brown 1. Relationship between Iron and Cobalt Pigments The use of iron pigment for decorating white porcelain at the official kilns is closely connected to the procurement of cobalt pigment. From the early stages of their operation, the official court kilns produced white porcelain with underglaze cobalt-blue decoration (鐵畵白磁, cheonghwa baekja) as the finest ceramics available to the Joseon court. Records from the reigns of King Seongjong (成宗, r. 1469–1494) and King Jungjong (中宗, r. 1506–1544) in the Joseon wangjo sillok (朝鮮王朝實錄, Annals of the Joseon dynasty) repeatedly mention that the Joseon court imported cobalt from China to produce blue-and-white porcelain (Entries for the 11th day of the eighth month of the ninth year, Seongjong sillok, vol. 95; 23th day of the first month of the 19th year, Seongjong sillok, vol. 211; 28th day of the 12th month of the 36th year, Jungjong sillok, vol. 97). Yet, according to the Yongjae chonghwa (鐵畵白磁, Assorted writings of Yongjae), vol. 10, written by Seong Hyeon (成俔, 1439–1504) in the late fifteenth and early sixteenth centuries, “as cobalt (回靑, hoecheong) is rare and precious, it has become difficult to obtain the pigment even in China. . . . For this reason, Joseon has very few porcelain wares with paintings.” These records indicate that the procurement of Chinese cobalt became a challenge. During the first half of the seventeenth century, Joseon suffered financial difficulties stemming from constant warfare and recurring natural disasters, which led to the temporary closure of the official kilns. In China, during the same period, a Jurchen invasion and a peasant rebellion led by Li Zicheng (李自成, 1606–1645) disrupted porcelain production at the Jingdezhen (景德鎭) kilns in Jiangxi Province. The wars in both Joseon and China, not to mention Joseon’s foreign policy toward Ming and Qing China had an adverse impact on official trade. Accordingly, the procurement of Chinese ceramics and cobalt pigment became even more difficult in the seventeenth century, as shown in a record from the reign of King Gwanghaegun (光海君, r. 1608–1623): “The Saongwon reported that since none of the painted wine jars (畵樽, hwajun) used for court banquets for foreign envoys survived the wars, they attempted to make some by purchasing cobalt-blue pigment. However, there was no way to obtain the pigment. Thus, whenever court banquets were held, the office had no choice but to use faked versions. It seemed rather pitiful.” (Entry for the third day of the fourth month of the tenth year, Gwanghaegun ilgi, vol. 27) As indicated in this record, Joseon became unable to produce painted jars for holding wine or flowers at court banquets due to the difficulty of procuring cobalt, and had no choice but to use white porcelain adorned with fake blue-and-white decoration, known as gahwa (假畵, literally “fake painting”). Moreover, another record from 1634, the 12th year of the reign of King Injo (仁祖, r. 1623–1649) from the Seungjeongwon ilgi (承政院日記, Daily records of the royal secretariat) noted the following: “In the past, dragon jars used at the banquets for foreign envoys bore imitations of underglaze blue-and-white decoration [gahwa], and so when they were transported, the paints often flaked off. Thus, we [Saongwon] frequently entrusted money to official interpreters who went to Beijing and ordered them to purchase cobalt pigment; however, the interpreters failed to bring it back. As a last resort, we painted dragon jars with seokganju and fired them, which were then used when welcoming Chinese envoys. We would like to reuse them at future receptions for the Chinese envoys.” (Entry for the 18th day of the fifth month of the 12th year of King Injo’s reign, Seungjeongwon ilgi, fasc. 43) As shown in the above entry, when it became impossible to procure cobalt pigment to decorate dragon jars for use at the welcoming receptions for envoys, dragon jars began to be manufactured using seokganju, iron oxide pigment. This information, which is the earliest documentation of the use of seokganju in decorating white porcelain, indicates that the difficulties in producing blue-and-white porcelain resulted in a switch to white porcelain decorated with underglaze iron-brown. The scarcity of cobalt pigment continued to be mentioned until the early eighteenth century. 2. Supply and Demand of Iron Oxide Pigment and the Visual Effects of Decoration 1) Efficient Procurement of Iron Oxide Pigment Next, let us examine the ease of procuring iron oxide pigment as another context for the production of white porcelain decorated with underglaze iron-brown. Shards of white porcelain ware decorated with underglaze iron-brown from the Goryeo dynasty have been excavated at kiln sites in Gyeonggi-do Province, such as the klin sites in Bangsan-dong, Siheung, and the white porcelain kiln sites in Seo-ri, Yongin, which operated during the early Goryeo period. Moreover, celadon decorated with underglaze iron-brown were produced at the kilns in Jinsan-ri, Haenam, Jeollanam-do Province, which demonstrates the widespread use of underglaze iron-brown designs on both white porcelain and celadon wares. It is well known that iron oxide pigment was used to decorate buncheong ware (粉靑沙器, buncheong sagi) at the early Joseon kilns in Hakbong-ri on Mt. Gyeryongsan in the Gongju area. The continuous use of iron oxide to decorate ceramics from the Goryeo era onward may in part have been due to the pigment being easily obtainable. The Sejong sillok jiriji (世宗實錄 地理志, Geographical appendix to the annals of King Sejong, 1432), the definitive record reflecting Joseon in the first half of the fifteenth century, mentions the regions that produced iron oxide pigment, also known as juto (朱土, literally “red clay”). As juto was extracted in Gyeonggi-do, Chungcheong-do, Gyeongsang-do, Hwanghae-do, and Gangwon-do Provinces, the pigment was essentially available nationwide. The ease of obtaining iron oxide pigment naturally led to the use of iron to decorate white porcelain, and iron oxide replaced cobalt when the latter became unavailable. 2) Inspiration from Foreign Polychrome Wares The production of white porcelain decorated with underglaze iron-brown in the early and mid-Joseon period was closely related to the insufficient supply of cobalt pigment and the widespread availability of iron oxide pigment. In the late Joseon period, however, iron oxide pigment was deliberately used to maximize the visual effect of porcelain wares. Lavishly decorated polychrome wares were imported from China and Japan during the late Joseon period and transformed the tastes of the consumers. Writings by late Joseon scholars from the Bukhak school (北學派, School of Northern Learning) testify to the changes in the perception of porcelain among prospective consumers. In his book titled Seolsuoesa (雪岫外史, Seolsu’s miscellaneous notes), Yi Huigyeong (李喜經, 1764–1845) praised Chinese ceramics while criticizing the deficiencies in the manufacturing techniques used for producing their Joseon counterparts. He further criticized the policy of King Jeongjo (正祖, r. 1776–1800) that emphasized diligence and frugality, and asserted a need for systemic reform to promote consumption and develop commerce and industry. In a similar vein, the Imwon simnyukji (林園經濟志, Sixteen treaties on rural life and economy) by Seo Yugu (徐有榘, 1764–1845) describes the outstanding qualities of Chinese ceramics, as well as of three-tiered Japanese containers with a crackled glaze in the style of the Ge ware (哥窯, Ch. geyao), with gold cloisonné, or painted in gold and blue. Yi Gyugyeong (李圭景, 1788–1860) also wrote that Joseon porcelain was inferior to its Chinese and Japanese counterparts in his book Oju yeonmun jangjeon sango (五洲衍文長箋散稿, Random expatiations of Oju) and its appendix Oju seojong (五洲書種, Books on military technology). Several extant documents, including the aforementioned examples, reveal that some high-quality Chinese and Japanese porcelain wares were used in Joseon, and that these imported polychrome wares inspired a need for refining the production system of Joseon ceramics. Such changes in the aesthetic tastes of the consumers affected porcelain production at official court kilns, and porcelain began to be decorated using a variety of new styles and techniques. In the late Joseon period, iron oxide pigment was used to enhance the visual effects of porcelain in accordance with the trends that prevailed at that time. Beginning Phase of the Production of White Porcelain Decorated with Underglaze Iron-brown White porcelain decorated with underglaze iron-brown was created throughout the Joseon period; however, precisely when its production at official court kilns began has not yet been determined. Since no remaining early Joseon documents feature records on the production of white porcelain decorated with underglaze iron-brown and only a few examples with reliable dates have survived, it is not easy to ascertain the actual initiation of its production. The earliest dated example is White Porcelain Epitaph Plaque with Inscription of “Jeong Seon, Grand Supervisor at the Directorate for Palace Delicacies” in Underglaze Iron-brown, which was produced around 1468 and currently housed in the Haegang Ceramics Museum.1 Jeong Seon (鄭善, died 1468) was a Korean official who held the title of grand supervisor (太監, Ch. taijian) at the Directorate for Palace Delicacies (尙膳監, Ch. shangshanjian) in the Ming Chinese court. Although this is an example with a clear date, it cannot be concluded that white porcelain decorated with underglaze iron-brown was common at this time because this is an epitaph plaque. Nevertheless, some surviving documents indicate that the Joseon court made an effort to seek out a local cobalt pigment called tocheong (土靑, literally “local blue”) around the time when this epitaph plaque was made. For example, the Joseon wangjo sillok noted such efforts during the reigns of King Sejo (世祖, r. 1455–1468) and King Yejong (睿宗, r. 1468–1469) (Entries for the 24th day of the fifth month of the ninth year, Sejo sillok, vol. 30; third day of the seventh month of the ninth year, Sejo sillok, vol. 34; 13th day of the ninth month of the tenth year, Sejo sillok, vol. 34; fifth day of the tenth month of the first year, Yejong sillok, vol. 8). However, the production of blue-and-white porcelain using local blue pigment failed. Due to the failure to procure suitable local blue pigment, iron oxide pigment appears to have been used as a substitute for cobalt from the time when the official court kilns came into operation. The production of white porcelain decorated with underglaze iron-brown at official kilns is thought to be attributable to the influence of inlaid white porcelain, buncheong ware with underglaze iron-brown decoration, and Chinese porcelain. However, inlaid white porcelain and buncheong ware with painted iron decoration both present striking contrasts to white porcelain decorated with underglaze iron-brown in terms of manufacturing technique as well as the compositions and subject matters of the decoration. Moreover, white porcelain shards with underglaze black decoration in the style of Cizhou ware (磁州窯, Ch. cizhouyao) have been excavated at the Doma-ri kiln site No. 1 in Gwangju, Gyeonggi-do Province, which began operations around 1505 (Fig. 1). Recently similar types of shards have also been found in Cheongjin-dong, Jongno-gu, Seoul (Fig. 2). It may be argued that white porcelain with underglaze black decoration in the style of Cizhou ware influenced the emergence of underglaze iron-brown decoration in Joseon, but since the decorations on Joseon wares greatly differ from those on Chinese wares in terms of subject matter and composition, it seems that the Cizhou-type wares bore no direct influence on Joseon white porcelain decorated with underglaze iron-brown. Fig. 1. White Porcelain Jar Shards with Black and White Slip-painted Decoration. Late Yuan and early Ming, China. Excavated at the Doma-ri kiln site No. 1 in Gwangju, Gyeonggi-do Province. National Museum of Korea Fig. 2. White Porcelain Jar Shards with Black and White Slip-painted Dragon Design. Late Yuan and early Ming, China. Length: 19.6 cm (max). Excavated in Cheongjin-dong, Jongno-gu, Seoul. Hanul Research Institute of Cultural Heritage Rather, it is more likely that it was the blue-and-white porcelain produced at the official court kilns that prompted the widespread production and use of iron-painted designs on early Joseon porcelain wares. Both White Porcelain Jar with Plum Blossom Design in Underglaze Iron-brown and White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue from the early Joseon period in the collection of the National Museum of Korea show a thick, everted rim and a wide shoulder tapering to a narrow foot (Figs. 3 and 4). In addition they both bear similar designs of plum blossoms rendered in the “boneless” (沒骨, K. molgol; Ch. mogu) technique of ink and wash painting. Thus, white porcelain decorated with underglaze iron-brown seems to manifest a closer relation to contemporaneous official blue-and-white porcelain than to inlaid white porcelain, buncheong ware with underglaze iron-brown decoration, and Chinese Cizhou-type wares. Fig. 3. White Porcelain Jar with Plum Blossom Design in Underglaze Iron-brown. Joseon, 16th century. Height: 7.1 cm, Diameter: 3.9 cm (mouth); 5.6 cm (foot). National Museum of Korea Fig. 4. White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue. Joseon, 16th century. Height: 12.4 cm, Diameter: 7.3 cm (mouth). National Museum of Korea As discussed above, the production of white porcelain decorated with underglaze iron-brown at the official kilns was closely related to that of official blue-and-white porcelain. Since cobalt pigment was not produced domestically in the early and mid-Joseon periods, Joseon had no choice but to rely upon imports from China. Consequently, when Joseon’s diplomatic relations with China were strained and the country’s internal affairs disrupted trade, blue-and-white porcelain production suffered and cobalt had to be replaced with iron oxide, which was easily obtainable within Joseon. In the late Joseon period, imported polychrome wares sparked in consumers a penchant for more decorative wares. In response to this trend, official court kilns began to use iron oxide pigment as decoration to enhance the visual impact of their products and the kilns continued to do so until the end of the Joseon period. Excavation of White Porcelain Decorated with Underglaze Iron-brown from Official Kilns Official Kilns for White Porcelain Decorated with Underglaze Iron-brown and Relevant Findings Thus far, there is no known example of white porcelain decorated with underglaze iron-brown excavated from the sites of official court kilns that operated from the second half of the fifteenth to the first half of the sixteenth century. The earliest example unearthed was from the Beoncheon-ri kiln site No. 9, which is known to have been in operation around 1552, and a few examples have been found at early Joseon kiln sites from the second half of the sixteenth century, including Beoncheon-ri kiln sites Nos. 5 and 8, Hakdong-ri kiln site No. 2, and Mugap-ri kiln site No. 10 (Woo Minah 2011, 250). In terms of type, epitaph plaques have been the most commonly found, followed by foliated cups, jars, and animal-shaped ritual vessels, implying that white porcelain decorated with underglaze iron-brown was produced for special uses and occasions rather than for daily use. In general, underglaze iron decoration was applied to fine-quality white wares, and decorative motifs were limited to plum blossoms and Chinese characters. Approximately 180 official kiln sites from the fifteenth and sixteenth centuries have been investigated, and twelve of them (6.7 percent) have yielded white porcelain decorated with underglaze iron-brown (Map 1 ). Considering the small amount of this type of excavated porcelain, there must have been relatively few kilns producing white porcelain decorated with underglaze iron-brown during this time. Map 1. Distribution of the official kiln sites that yielded white porcelain decorated with underglaze iron-brown in Gwangju, Gyeonggi-do Province (Image edited by the author; Ewha Womans University 1993, P. 5) As for seventeenth-century kilns, a considerable number have yielded ceramics with underglaze iron-brown decoration, including those in Wangsan-ri in Yongin (1626–1627). Such kiln sites in Gwangju, Gyeonggi-do Province include Sangnim-ri (1618–1636), Seondong-ri (1640–1648), Songjeong-dong (1649–1654), Yusa-ri (1655–1664), Sindae-ri (1665–1676), Jiwol-ri (1677–1680), and Yujeong-ri (late seventeenth to early eighteenth century) (Woo Minah 2011, 251). Although the ancient sites of the Tanbeol-ri (1606–1612) and Hakdong-ri (1613–1617) kilns which operated prior to the Sangnim-ri kilns have yielded no white porcelain decorated with underglaze iron-brown, it can be assumed that white porcelain decorated with underglaze iron-brown was manufactured consistently from the latter half of the sixteenth century, and that its production increased around the time when the Sangnim-ri kilns were in operation. Seventeenth-century white porcelain decorated with underglaze iron-brown not only took the form of specialized vessels, such as jars, bottles, water droppers, lidded bowls, epitaph plaques, and ritual vessels, but also wares for daily use, including bowls, dishes, and cups. Moreover, compared to examples from the early Joseon period, those of the seventeenth century were embellished with more diverse designs, including dragon and cloud, plum blossom, bamboo, grape, chrysanthemum, orchid, and poetry motifs. In the latter half of the seventeenth century, wares for daily use were increasingly decorated in underglaze iron-brown, and simple designs such as flowers and plants became more frequently used. Among the 90 official court kilns from the seventeenth century that have so far been investigated, 46 have yielded white porcelain decorated with underglaze iron-brown, manufactured in both saggar- and stack-fired kilns. These comprise 51 percent of all seventeenth-century kilns that have been investigated to date. This reflects the increase in the production of white porcelain decorated with underglaze iron-brown during the seventeenth century. A small quantity of shards with iron-brown decoration were discovered at the eighteenth- and nineteenth-century kiln sites in Geumsa-ri (1721–1752) and Bunwon-ri (1752–1883). They have been identified as fragments of jars and bottles decorated with dragon and cloud, bamboo, or grass and flower designs. A shard of white porcelain with dragon and cloud design was unearthed at the Geumsa-ri kiln site, which suggests that products of the first half of the eighteenth century still exhibited characteristics of seventeenth-century royal porcelain. Excavations have thus far revealed a total of 20 kilns that operated in and after the eighteenth century, and white porcelain shards with underglaze iron-brown decoration have been found at five of them. This suggests that the production of white porcelain decorated with underglaze iron-brown declined compared to the seventeenth century. Even though a large number of kilns were in operation during the early Joseon period, only a few of them have been revealed to have manufactured white porcelain decorated with underglaze iron-brown. On the other hand, the discovery of white porcelain decorated with underglaze iron-brown at seventeenth-century official kiln sites indicates that its production expanded considerably at this time, although the quality of white porcelain decreased because of management problems at the official court kilns due to internal and external factors. This clearly confirms that iron oxide was the most widely used pigment for ornamenting official ceramics during this period. However, the production of white porcelain decorated with underglaze iron-brown declined once again in the eighteenth century as an amicable trade relationship with Qing China facilitated the procurement of cobalt pigment, and the production of blue-and-white porcelain accordingly increased. Characteristics of White Porcelain Decorated with Underglaze Iron-brown Excavated from Official and Regional Kiln Sites Ceramics unearthed from the Songjeong-dong kiln site No. 5 (1649–1659) mainly consist of bowls (20.54 percent), dishes (55.6 percent), and cups (6.02 percent). In other words, porcelain wares for daily use comprise 82.16 percent of the excavated items, while wares for special occasions, such as jars, bottles, and lidded bowls comprise only 17.84 percent of the total (Gyeonggi Ceramic Museum 2008, 75). Most of the Joseon kiln sites investigated present a similar configuration of ceramic types. Due to the difficulty of their manufacture, large-scale vessels seem to have been produced only in small quantities as an offering to the king. A total of 33 shards with underglaze iron-brown decoration were discovered at the Songjeong-dong kiln site. Most of these were remnants of jars and bottles, which were produced in small quantities, rather than bowls and dishes. Sindae-ri kiln site No. 18 yielded a greater number of daily wares decorated with iron oxide pigment, but as in the case of the Songjeong-dong kiln site No. 5 underglaze iron-brown decorations were found more frequently on jars and bottles than on wares for daily use (Fig. 5). These findings indicate that at the official court kilns, underglaze iron-brown decoration was predominantly used on special types of ware. Fig. 5. White Porcelain Shards Decorated with Underglaze Iron-brown. Joseon, second half of the 17th century. Height: 26.7 cm (bottle). Excavated at the Sindae-ri kiln site No. 18 in Gwangju, Gyeonggi-do Province. Gyeonggi Ceramic Museum (Gyeonggi Ceramic Museum 2008c, p. 10) Regional kilns, on the other hand, show different results compared to the official court kilns. The production of white porcelain decorated with underglaze iron-brown increased at regional kilns in the mid-seventeenth century. Excavations of regional kilns have revealed that white wares decorated with underglaze iron-brown were created in a number of regions. In contrast to the official court kilns, the regional kilns employed underglaze iron-brown decoration more frequently on wares for daily use, including bowls, dishes, and cups, rather than on special wares, such as jars and bottles (Central Institute of Cultural Heritage 2003 and 2004; Chungbuk National University Museum 1995; Chungcheong Research Institute of Archaeological Heritage 2002; Ewha Womans University Museum 2000; Gyeonggi Ceramic Museum 2009; Honam Cultural Property Research Center 2006; Korea Cultural Heritage Institute 2010). Moreover, decorations on official white porcelain wares differ from those on local kiln products. While the former was ornamented primarily with motifs reserved for the royal court, such as dragon and cloud, in addition to bamboo, plum blossom, and chrysanthemum, the latter were decorated chiefly with grass and flowers or simple dot designs (Fig. 6). Recently, shards of a jar with a dragon and cloud design were unearthed for the first time at a local kiln site in Hapan-ri, Gapyeong, Gyeonggi-do Province (Fig. 7). These jar shards exhibit crude strokes and a simple, carelessly depicted design compared to jars with dragon and cloud designs from the official court kilns. The ease of procurement of iron oxide pigment prompted the production of local iron-painted white porcelain ware, but the underglaze iron-brown decoration was applied in a different manner and style on royal and on local wares. Fig. 6. White Porcelain Bowl Shard and Cup with Grass Design in Underglaze Iron-brown. Joseon, second half of the 17th century. Bowl-Height: 11.2 cm, Diameter: 18.0 cm (mouth); 6.8 cm (foot). Cup-Height: 3.6 cm, Diameter: 14.4 cm (mouth); 6.5 cm (foot). Excavated at the Hapan-ri kiln site in Gapyeong, Gyeonggi-do Province. Gyeonggi Ceramic Museum (Gyeonggi Ceramic Museum 2008a, p. 30) Fig. 7. White Porcelain Jar Shards with Dragon and Cloud Design in Underglaze Iron-brown. Joseon, second half of the 17th century. Left–Height: 33.7 cm (max), Diameter: 13.7 cm (foot). Right–Height: 21.2 cm (max), Diameter: 15.8 cm (mouth). Excavated at the Hapan-ri kiln site in Gapyeong, Gyeonggi-do Province. Gyeonggi Ceramic Museum (Gyeonggi Ceramic Museum 2008a, p. 32) This section scrutinized aspects of the production of iron-painted white porcelain by briefly investigating the results of excavations of white porcelain ware decorated with underglaze iron-brown at both official court and regional kilns. While only a handful of kilns produced white porcelain decorated with underglaze iron-brown in the early Joseon period, the manufacture of iron-painted wares increased notably during the seventeenth century. In the eighteenth century, the resumption of ready imports of cobalt pigment led to a decline in the production of white porcelain decorated with underglaze iron-brown. In comparison with regional kilns, official court kilns issued relatively greater quantities of white porcelain decorated with underglaze iron-brown intended for special use, such as jars and bottles, and employed more varied motifs for decoration, including dragon and cloud, bamboo, plum blossom, grape, and chrysanthemum designs. Accordingly, official kilns produced fine quality wares with underglaze iron-brown decoration for court use on special occasions, such as banquets and ritual ceremonies. Types and Ornamentation of Official White Porcelain Decorated with Underglaze Iron-brown From the Late Fifteenth Century to the Sixteenth Century Only a small quantity of white porcelain decorated with underglaze iron-brown was manufactured during this period. In form, it was limited to jars, bottles, water droppers, funerary objects, epitaph plaques, ritual vessels, and dishes, and in decoration to plum blossom, bamboo, and simple linear designs. A case in point is White Porcelain Jar with Mountains and Fret Design in Underglaze Iron-brown and Cobalt-blue housed in Leeum, Samsung Museum of Art (Fig. 8). It takes its form after ancient ritual bronzeware. Excluding the rings on the shoulder, its form resembles that of White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue, also in the Leeum collection, which dates it to the second half of the fifteenth century (Fig. 9). Since no contemporaneous examples of iron-painted white porcelain in a similar style have been found, it is likely that iron oxide pigment was only temporarily used for making white porcelain ritual vessels. White Porcelain Bottle with Rope Design in Underglaze Iron-brown in the collection of the National Museum of Korea resembles White Porcelain Bottle with Plum, Bird, and Bamboo Design in Underglaze Cobalt-blue in a private collection, which is presumed to have been produced between the early to mid-sixteenth century (Fig. 10).2 Both bottles have an everted rim, narrow neck, and voluminous lower body that widens from the neck. A comparison between these two bottles proves that white porcelain decorated with underglaze iron-brown from the early Joseon period was created in similar shapes as blue-and-white porcelain from the same period. Fig. 8. White Porcelain Jar with Mountains and Fret Design in Underglaze Iron-brown and Cobalt-blue. Joseon, second half of the 15th century. Height: 27.8 cm. Diameter: 9.5cm (mouth); 11.8 cm (foot). Leeum, Samsung Museum of Art Fig. 9. White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue. Joseon, second half of the 15th century. Height: 41 cm, Diameter: 15.7 (mouth); 18.2 cm (foot). Leeum, Samsung Museum of Art Fig. 10. White Porcelain Bottle with Rope Design in Underglaze Iron-brown. Joseon, 16th century. Height: 31.4 cm, Diameter: 7.0 cm (mouth); 10.6 cm (foot). National Museum of Korea In addition to their forms, the decorations on blue-and-white porcelain impacted contemporaneous white porcelain decorated with underglaze iron-brown. As mentioned previously, White Porcelain Jar with Plum Blossom Design in Underglaze Iron-brown and White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue, both in the collection of the National Museum of Korea, share similarities in the placement of their plum designs that spread across the entire surface of the body, the use of “boneless” brushstrokes, and the painting style which depicts only the plum blossoms and branches without trunks (Figs. 3 and 4). Both jars have thick, everted mouths and wide shoulders that gradually taper to the foot. Such similarities indicate that during the early Joseon period, white porcelain decorated with underglaze iron-brown and white porcelain decorated with underglaze cobalt-blue were produced simultaneously in official court kilns and were used concurrently. Furthermore, extant examples of early Joseon white porcelain decorated with underglaze iron-brown are of a similar premium quality as the blue-and-white porcelain of the era, as evidenced by the quality of the clay, glaze, carved foot, and decoration. Seventeenth Century During the seventeenth century, white porcelain decorated with underglaze iron-brown was produced in the official court kilns in nearly all the then-current forms, including jars, bottles, lidded boxes, bowls, dishes, cups, water droppers, burial objects, ritual vessels, epitaph plaques, and barrel-shaped jars (janggun). They were also decorated in a range of motifs, such as dragon and cloud, bamboo, plum blossom, grape, chrysanthemum, flower, and Chinese characters. The traditions of the early Joseon period continued well into the first half of the seventeenth century, yet changes took place as well. For example, White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown in the collection of the National Museum of Korea follows the form of the early Joseon White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue discussed earlier (Figs. 9 and 11). Both jars have inverted mouths and broad shoulders that taper sharply toward a flaring foot. The traditions of the preceding period can also be seen in the decorations as well as in the forms. Plum and bamboo designs, common in the early Joseon period, continued to appear on white porcelain decorated with underglaze iron-brown from the first half of the seventeenth century. The only difference in the designs is found in their compositions. For instance, on White Porcelain Jar with Plum and Bamboo Design in Underglaze Cobalt-blue from the late fifteenth century plum and bamboo are painted together, whereas plum and bamboo are painted separately on different sides of the seventeenth-century White Porcelain Jar with Plum and Bamboo Design in Underglaze Iron-brown (Figs. 9 and 12). Fig. 11. White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown. Joseon, first half of the 17th century. Height: 36.2 cm, Diameter: 14.0 cm (mouth); 14.1 cm (foot). National Museum of Korea Fig. 12. White Porcelain Jar with Plum and Bamboo Design in Underglaze Iron-brown. Joseon, first half of the 17th century. Height: 36.9 cm, Diameter: 14.0 cm (mouth); 14.1 cm (foot). Leeum, Samsung Museum of Art In the latter half of the seventeenth century, the forms of white porcelain jars decorated with underglaze iron-brown changed once again. White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown in the National Museum of Korea represents a new shape with an upright mouth slightly inverted at the rim and voluminous shoulders that gradually taper to a somewhat flaring foot (Fig. 13). There are several surviving examples of this type of jar, suggesting that official white porcelain wares were made in a schematized style during the latter half of the seventeenth century. Fig. 13. White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown. Joseon, first half of the 17th century. Height: 35.7 cm. National Museum of Korea The second half of the seventeenth century witnessed an increase in the production of iron-painted white porcelain wares for daily use, such as bowls and dishes. Furthermore, their motifs expanded from plum and bamboo, dragon and cloud, to chrysanthemum, grape, and grass and flower designs. Despite this expansion of vessel types and designs, later productions repeat the same shapes and have simplified and schematized designs, which may be a result of the increased production of white porcelain decorated with underglaze iron-brown at the official kilns in the latter half of the seventeenth century. Such stylization of motifs can be observed in dragon and cloud designs (Table 1). The scales of the dragon on a jar from the first half of the seventeenth century are fan-shaped and carefully painted. On the other hand, those on a jar from the latter half of the seventeenth century are roughly depicted and the fan-shaped scales are finished with simple dots. Moreover, the clouds on the former jar are outlined and colored in, whereas only the contours of the clouds on the latter jar are drawn up. Considering that the design resembles that on a shard uncovered from the Sindae-ri kiln site No. 18, the latter jar seems to have been made around the time when the Sindae-ri kilns were in operation (Fig. 14). In the case of the plum and bamboo design, in the first half of the seventeenth century it was elaborately rendered in a realistic manner, whereas examples from the latter half of the seventeenth century show a simple composition of bamboo leaves and less complicated brushstrokes (Fig. 12). Such a simplified composition is well represented in a white porcelain bottle shard with iron-painted bamboo leaves excavated from the Sindae-ri kiln site No. 18 (Fig. 15). This site also yielded other iron-painted white porcelain shards decorated with simple grass and flower designs using only dots and lines. These wares seem to have been fired without saggars at the official kilns and were used at the royal court and government offices Table 1. Comparison of jars with dragon and cloud designs in underglaze iron-brown from the early and late seventeenth century Fig. 14. White Porcelain Jar Shard with Dragon and Cloud Design in Underglaze Iron-brown. Excavated at the Sindae-ri kiln site No. 18 in Gwangju, Gyeonggi-do Province. Joseon, second half of the 17th century. Height: 7.6 cm (max). Gyeonggi Ceramic Museum (Gyeonggi Ceramic Museum 2008c, p. 117) Fig. 15. White Porcelain Bottle Shards with Bamboo Design in Underglaze Iron-brown. Joseon, second half of the 17th century. Height: 26.7 cm (max). Excavated at the Sindae-ri kiln site No. 18 in Gwangju, Gyeonggi-do Province. Gyeonggi Ceramic Museum (Gyeonggi Ceramic Museum 2008c, p. 121) From the Eighteenth through the Nineteenth Century There is very little white porcelain with iron decoration from the eighteenth century that has survived till today. What remains is mainly jars, bottles, water droppers, paper-roll holders, and brush holders. A representative example from the first half of the eighteenth century is White Porcelain Bottle with Inscription of “Tea Bottle for the King” in Underglaze Iron-brown, probably made between 1710 and 1720.3 According to the Seungjeongwon ilgi, while King Yeongjo (英祖, r. 1724–1776) held the title of chief superintendent (都提調, dojejo) of the Bunwon official court kilns, during his term as crown prince, he commanded that an inscription of “tea bottle for the king” (進上茶甁) be written in seokganju on royal wares in order to prevent Bunwon officials from misappropriating official wares (Entry for the 21th day of the tenth month of the third year of King Yeongjo’s reign, Seungjeongwon ilgi, fasc. 648). Therefore, this bottle was possibly created either in the 1710s when King Yeongjo held the position of dojejo, or in the 1720s during the early years of his reign. It attests to the ongoing use of iron oxide pigment on royal white porcelain during the eighteenth century. There are a few examples of white porcelain jars with grape designs in underglaze iron-brown which may date to the first half of the eighteenth century. Formerly, grape designs had mainly been used to decorate dishes and flat bottles (扁甁, pyeonbyeong), but only rarely jars. It is noteworthy that in this period, the grape design began to be used as a chief motif for decorating ceramics and that iron oxide pigment was used to execute the design. White Porcelain Jar with Grape Design in Underglaze Iron-brown in the collection of the Ewha Womans University Museum is decorated with grapes on the upper half of the front and rear sides of the body, while the lower half is left blank (Fig. 16). The diagonally arranged composition of grapes is elegant, and the grape leaves, stems, and vines are painted in a delicate and elaborate manner. White porcelain jars with dragon and cloud designs were still being decorated using underglaze iron-brown during the eighteenth century, as evidenced by a small number of jar shards bearing dragon and cloud designs excavated from the Geumsa-ri kiln site. White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown, a tall jar in the National Museum of Korea, was probably made in the late eighteenth or early nineteenth century, considering that its shape and overall design reflect those of blue-and-white porcelain jars with dragon and cloud designs from the same period (Fig. 17). Fig. 16. White Porcelain Jar with Grape Design in Underglaze Iron-brown. Joseon, first half of the 18th century. Height: 53.3 cm, Diameter: 19.0 cm (mouth); 18.6 cm (foot). Ewha Womans University Museum Fig. 17. White Porcelain Jar with Dragon and Cloud Design in Underglaze Iron-brown. Joseon, 18th century. Height: 57.5 cm, Diameter: 21.3 cm (mouth). National Museum of Korea In the latter half of the eighteenth to the nineteenth century, white porcelain decorated with underglaze iron-brown was produced in various novel types and shapes, including cylindrical bottles, paper-roll holders, coins, and saucers. Most examples from this period were not decorated with iron oxide pigment alone, but in combination with cobalt and copper pigments. Furthermore, decorative techniques diversified and came to incorporate relief and openwork. New motifs such as potted plants or grass and insects began to be employed, and the technique of covering the entire body with iron oxide pigment, known as the cheolchae (鐵彩) technique, was more frequently used. This phenomenon could be linked to the aforementioned polychrome wares of China and Japan. White Porcelain Mold-impressed Bottle with Orchid, Chrysanthemum, Grass, and Insect Design in Underglaze Cobalt-blue, Iron-brown, and Copper-red (Fig. 18) in the Gansong Art Museum exhibits the contemporaneous trend of including several different decorative methods within a single piece, such as a relief technique with various colorants, including cobalt, copper, and iron. Another example from this period is White Porcelain Bottle Decorated with Underglaze Iron-brown (Fig. 19). Its unprecedented cylindrical body is entirely covered with iron oxide pigment. Thus, white porcelain decorated with underglaze iron-brown changed following the trend of the time of enhancing visual effects by incorporating multiple pigments and techniques within a single piece or by painting the entire body of the vessel with iron pigment. Fig. 18. White Porcelain Mold-impressed Bottle with Orchid, Chrysanthemum, Grass, and Insect Design in Underglaze Cobalt-blue, Iron-brown, and Copper-red. Joseon, late 18th century. Height: 42.3 cm, Diameter: 4.1 cm (mouth); 13.3 cm (foot). Gansong Art Museum Fig. 19. White Porcelain Bottle Decorated with Underglaze Iron-brown. Joseon, 19th century. Height: 27.0 cm, Diameter: 3.3 cm (mouth); 13.9 cm (foot). National Museum of Korea Periodic Division Based on the Variations of White Porcelain Decorated with Underglaze Iron-brown and Changes in Its Characteristics Phase I: From the Latter Half of the Fifteenth through the Sixteenth Century During this phase, official court kilns were established in Gwangju, Gyeonggi-do Province and iron oxide pigment was used only sparingly as a recourse to the instability of the supply of cobalt pigment. Since iron oxide pigment was easily obtainable in Joseon, it began to be employed temporarily as a substitute for cobalt pigment. Examinations of extant dated examples verify that iron oxide pigment was being used to decorate white porcelain around the time when official court kilns were established in Gwangju. As this period coincides with the attempts to locate a domestic source of cobalt pigment, known as tocheong, we can confirm that iron oxide pigment was used when the supply of cobalt pigment was problematic. The proportion of white porcelain wares with underglaze iron-brown designs uncovered at the kilns that yielded such wares suggests only a limited production of white porcelain decorated with underglaze iron-brown. The types of wares unearthed were also restricted to those made for special purposes, such as jars, bottles, epitaph plaques, ritual vessels, and funerary objects, and the motifs of their designs were limited to plum blossoms and Chinese characters. Produced in relatively small quantities, early Joseon white porcelain decorated with underglaze iron-brown shares similarities with contemporaneous blue-and-white porcelain in terms of form and design (i.e., motifs, composition, and techniques), which indicates that both types were produced at the official kilns. From these results, it can be concluded that in the early Joseon period, white porcelain decorated with underglaze iron-brown was produced on a limited scale as a substitute for blue-and-white ware for the royal court at times when the supply of cobalt pigment was unstable. Phase II: From the Seventeenth to the First Half of the Eighteenth Century Phase II ranges from the seventeenth century after the invasions by China and Japan to the first half of the eighteenth century when the Geumsa-ri kilns (1721–1752) were in operation. During this period, iron oxide functioned as the main decorative pigment at the official kilns. Although iron oxide was a replacement for cobalt, which could not be imported due to internal and external difficulties, it was principally used to decorate vessels for the royal court. In the seventeenth century, white porcelain decorated with underglaze iron-brown was used at major functions at the court and was therefore decorated with dragon and cloud designs symbolizing the royal household. The percentage of kiln sites where white porcelain decorated with underglaze iron-brown have been excavated and the types and designs of the ceramics unearthed reveal that the official kilns of the seventeenth century produced white porcelain decorated with underglaze iron-brown as their chief form of decorated ceramics. While white porcelain wares decorated with underglaze iron-brown from the first half of the seventeenth century still showed features of their early Joseon precedents, those from the latter half of the seventeenth century introduced innovations. Furthermore, in the latter half of the seventeenth century, painting in underglaze iron-brown gradually became the dominant decorative technique for vessels made for daily use such as bowls, dishes, and cups, and they incorporated a wide range of motifs such as grapes, chrysanthemums, and grass and flowers. Despite such diversification, however, underglaze iron-brown designs became simplified, cruder, and more stylized: grass and flower designs were rendered with merely one or two dots and lines, and supplementary patterns were omitted to highlight only the main motifs. This trend might have arisen out of changes in quality that came in conjunction with the more extensive production of ceramics using both saggar- and stack-firing and the expanded demand for porcelain decorated with underglaze iron-brown. When official kilns actively resumed the production of blue-and-white porcelain in the first half of the eighteenth century, the production of white porcelain decorated with underglaze iron-brown plunged. Nevertheless, since royal ceramics were still decorated with underglaze iron-brown, examples from the first half of the eighteenth century continued to retain features of their seventeenth-century precedents. Phase III: From the Latter Half of the Eighteenth through the Nineteenth Century During phase III, the kilns in Bunwon-ri were still in operation. A tendency toward extravagance and a preference for colorful porcelains from China and Japan inspired the production of a new type of decorative white porcelain. Starting in the late eighteenth century, landscapes, human figures, and supplementary motifs elaborately embellished the entire bodies with no background left undecorated. Moreover, several techniques were employed to adorn a single piece. For example, cobalt, copper, and iron pigments were used simultaneously, or the entire surface of a vessel was covered with one of these pigments. Diverse decorative techniques, such as relief and openwork, were also used. When iron and cobalt pigments were used together, the former generally supplemented the latter. Furthermore, the cheolchae technique emerged wherein iron oxide pigment covered the entire surface of ceramics. In the late Joseon dynasty, iron oxide pigment was used as a supplement to cobalt for enhancing visual effects, which in turn prompted the employment of new decorative techniques. Conclusion The production of white porcelain decorated with iron oxide continued uninterrupted from the time when the official kilns were established in Gwangju, Gyeonggi-do Province until the privatization of the official kilns toward the end of the nineteenth century. The characteristics and status of white porcelain decorated with underglaze iron-brown changed over the course of its history. Above all, the ease of procuring iron oxide pigment encouraged official kilns to produce white porcelain decorated with underglaze iron-brown, and along with blue-and-white porcelain it was crafted as the finest ware destined for use in the royal court since the early stages of the operation of the official kilns. A comparison between porcelain excavated from official kiln sites and examples unearthed at regional kiln sites reveals that official white porcelain decorated with underglaze iron-brown functioned not as daily tableware, but as specialized pieces used for ceremonies at the royal court and government offices. From the latter half of the fifteenth to the sixteenth century, due to the unreliable supply of cobalt pigment white porcelain decorated with underglaze iron-brown was produced on a limited scale as a substitute for blue-and-white porcelain used at the royal court. During the seventeenth and the first half of the eighteenth centuries, white porcelain decorated with underglaze iron-brown became the dominant form of premium-quality royal porcelain. However, due to changes in consumer preferences in the latter half of the eighteenth century, iron oxide was used alongside cobalt as a means to achieve a heightened visual effect, or it was used to coat the entire surface of a vessel. The use of iron oxide pigment continued from the early through the late Joseon period, presenting different characteristics and meanings in each phase. Since iron oxide pigment was readily available on the Korean Peninsula, white porcelain decorated with underglaze iron-brown was produced not only for the royal court, but also in different styles meeting the tastes of ordinary customers. Therefore, vessels for the royal court exhibit a refined appearance, while those destined for commoners reflect a more exuberant beauty. Such characteristics make Joseon white porcelain decorated with underglaze iron-brown unique and distinguishable from other Korean ceramics.

Writings on Korea are found in official Chinese historical records such as the Shiji (史記: Annals of History), the Hanshu (漢書: Han History), and the Weishu Dongyichuan (魏書 東夷傳: Record of the Foreigners in the East in the Wei History) in the Sanguozhi (三國志: Records of Three Kingdoms). These texts precede even the oldest remaining records in Korea, providing invaluable information about the politics, diplomacy, geography, culture, and customs of ancient Korea.1 In particular, these Chinese records on Korean customs and dress make repeated reference to Goguryeo (37 BC-668) people with feathers on their official headdresses, which indicate that this Goguryeo custom was well known in contemporary China. The Goguryeo headdress with feathers is first mentioned in the Weishu Dongyichuan, where the chapter on Goguryeo says, “... [the people] wear a headdress named jeolpunggeon (折風巾), which is pointed on top, and has feathers on the sides... there are differences according to one's social status ...”2 Further, the Zhoushu (風書: Zhou History) states, “... men wear long-sleeved jackets and wide trousers, with white leather belts and yellow leather shoes. Their headdress is called golso (骨蘇). It is mostly made of purple silk with intricate gold and silver decorations. The high officials show their status by wearing two feathers in their headdress ...”3 Similar records are also found in the Beishi (北史: History of Northern Dynasties), the Suishu (隋書: Sui History), the Jiu Tangshu (舊唐書: Tang History) and the Xin Tangshu (新唐書: New Tang History).4 Based on such Chinese records, the section of Saekbok (色服: dress codes and color protocols) in the chapter Japji (雜志: Miscellanea) of the Samguksagi (三國史記: History of the Three Kingdoms) by Gim Busik (金富軾), also contains the same description.5 As to Goguryeo's jougwan, Yi Yongbeom (李龍範) already suggested in 1956 that the custom of decorating one's headdress with feathers is related to the animism of Siberian nomadic tribes who worshipped birds,6 which has been widely accepted in the study of the history of dress and ornaments ever since.7 This paper shall not elaborate in detail the widely-known depictions of Goguryeo people wearing jougwan in the tomb murals of Ssang-yeongchong (雙盈塚) and Muyongchong (舞踊塚) (Plates 1 and 2). In the case of the Baekje kingdom, the Zhoushu and the Beishi describe Baekje (18 BC-660) dress as being similar to that of Goguryeo, although in Baekje the feathers were put on the headdress in the presence of the king at official meetings called johoe (朝會: court ceremonies), baerye (拜禮: ritual bows), or jesa (祭祀: ancestral rites); and while playing musical instruments, but not during military events.8 In Silla, too, the influence of the jougwan is found in gold crowns, which, once again, can be traced back to Goguryeo customs. ( Plate 1 ) A man wearing headdress crowned with two feathers Goguryeo, 6th century, wall painting Ssang-yeongchong (雙楹塚), Pyeongyang ( Plate 2 ) A horse rider wearing headdress crowned with two feathers in Ssang-yeongchong Goguryeo, 6th century, wall painting fragment National Museum of Korea, Seoul The mural painting, Reception of Foreign Envoys depicted on the entrance wall in the tomb of Li Xian (also known as Prince Zhanghuai [章懷太子], 651-684) of the Tang dynasty (618-907), has long drawn academic attention with particular respect to one of the foreign emissaries, who wears two feathers on his headdress, as being of probable Korean origin (Plate 3).9 Li Xian, the second son of Emperor Gaozong (高宗, r. 649-683) and Empress Wu Zetian (則天武后, r. 690-705), died in 684 and was reburied in 706 after Empress Wu's death. Princess Yongtai (永太公主, d. 701) was reburied in the same year and Li Chongrun (李重潤: Crown Prince Yide [懿德太子], 682-701) in 711. All three had died young during the empress's usurpation of the throne. Although a jougwan-styled official headdress had long been considered a characteristic of Goguryeo dress, the emissary in the tomb painting was assumed to have come from Unified Silla as the state of Goguryeo had collapsed in 668, and by the time of Prince Zhanghuai's death, Silla had unified the peninsula. However, it remains unclear whether the mural is based on the actual historical presence of a Silla emissary at his funeral. ( Plate 3 ) A Goguryeo envoy in Reception of Foreign Envoys Chinese, Tang, circa 706, wall painting Tomb of Prince Li Xian (Prince Zhanghuai), near Xi'an Another example of a Korean emissaries wearing a similar jougwan was found in the mural of Foreign Emissaries (使節圖, Shijietu) in the western wall of Room 1 at the ruined site of a palace in the Afrasiab plateau near Samarkand, Uzbekistan (Plates 4-1 and 4-2). This site is the remains of the ancient kingdom of Sogdiana, known in China as Kangguo (康國), which during the Sui and Tang dynasties flourished as a commercial center on the Silk Road. It is recorded that the Tang emperor Gaozong, during the Yonghui (永徽) era (650-655), appointed the Sogdian king Varxuman (拂呼縵) as governor of Kangju (康居都督),10 and there is an inscription in Sogdian language on the tomb mural that confirms this: “four geese to King Varxuman of the Unasi tribe.”11 Therefore, it would be safe to assume that the mural was completed between the late seventh century and the year 712, when the kingdom was plundered in the wake of the Arab invasion. This mural, thought to have been painted in the mid-seventh century within the private quarters of King Varxuman, contains what appears to be the reception of foreign emissaries from Chaganian of west Tokharia, Turks, Gaochang (高昌) in Central Asia (西域) or China, and India, as well as scenes of religious ceremonies and the paradise of gods to celebrate the glories of the king and to reinforce the king's authority. ( Plate 4-1 ) Outlined envoys in Foreign Emissaries, detail ( Plate 4-2 ) Goguryeo envoys wearing feathered headdresses in Foreign Emissaries Late 7th century-early 8th century, mural Western wall of Room 1, Afrasiab, Samarkand, Uzbekistan A report on this mural from the former Soviet Union interprets two figures in the mural, wearing round-collared jackets, circle-headed swords, and headdresses with two feathers, as Korean. Like the speculations surrounding Prince Zhanghuai's tomb mural, Kim Won-yong considered the depicted figures as likely to be from Silla, but did not exclude the possibility of them being envoys from Goguryeo.12 Anazawa Wako suggests that the two emissaries wearing jougwan were most likely of Goguryeo origin.13 Based on the characteristics of the Tang Chinese dress depicted in the mural, Anazawa dates the mural to the mid-seventh century coinciding with the rise of Empress Wu Zetian. It was a time when Goguryeo maintained close diplomatic relations with western Chinese states, ever since it dispatched an envoy to Tuque (突厥: Turks) during the Sui (隋) dynasty (581-618). Along similar lines, Roh Taedon (盧太敦) has suggested that the two figures probably were emissaries from Goguryeo who, facing an imminent war with Tang China at the time, were desperate to find military allies in continental Asia.14 It is difficult to conclude at this point whether the figure wearing jougwan in the tomb of Prince Zhanghuai was in fact a visiting emissary from Goguryeo or whether his headdress was depicted as a typical characteristic of Koreans as perceived by contemporary Chinese. However, considering the historical context of the time, contemporary historical records on Goguryeo dress, and the depiction and inscription of people in jougwan in Chinese paintings to be discussed later, it becomes clear that the people portrayed donning jougwan were from Goguryeo rather than Silla.15 II GOGURYEO ENVOYS WEARING JOUGWAN IN DUNHUANG WALL PAINTINGS The wall paintings of the Dunhuang caves are a rich source of the few surviving paintings of the Tang dynasty. Besides the diverse representations of Buddhist iconography, they provide a rare insight into the religious life of the time as well as the development of pictorial art, such as landscapes and portrait paintings. The depictions of the various figures appearing in narrative Buddhist stories in addition to illustrations of male and female donors provide invaluable historical information on costumes and customs of the time. Among the many wall paintings of Dunhuang that contain figures wearing jougwan, this paper focuses on three in particular. The first is Cave 220, depicting the paradise of Amitabha on the south wall, the paradise of Bhajsajyaguru on the north, and three Buddhas accompanied by Bodhisattvas on the east, above the entrance.16 Inscriptions on the northern wall and below the central Buddha depicted on the east wall indicate that the cave was completed in the year 642, the sixteenth year of the Zhenguan (貞觀) reign (627-649), and was the votive cave of the Zhai family.17 On either side of the entrance wall, there is a mural that features a scene from the Vimalakirti Sutra in which the Bodhisattva of wisdom, Manjusri (文殊菩薩, Wenshupusa), visits a Buddhist layman Vimalakirti (維摩居士, Weimo jushi), who has fallen ill, and both have entered into a discussion of the purity of the Buddha land and ‘non-duality' in the teachings of the Buddha. Just below the representation of Manjusri on the left, there is a depiction of the audience, including the Tang emperor and his officials. On the right, along with the representation of Vimalakirti, foreign emissaries in various different attires are portrayed, among whom one is wearing a jougwan. Preceded by a small person carrying an offering on a tray over his head, two people in short trousers carrying lotus flowers, and a person in a long Chinese costume, is a person wearing a jougwan decorated with two long feathers (Plate 5). Although further details of the headdress are not clearly recognizable, the features of this person in a blue jacket and wide trousers with his hands gathered together are similar to those of the emissary identified wearing jougwan in the tomb mural of Prince Zhanghuai. ( Plate 5 ) Foreign emissaries attending the debate between Vimalakirti and Manjusri Chinese, Tang 642, wall painting Eastern wall of Dunhuang Cave 220, Gansu According to the Vimalakirti Sutra, the visit of Manjusri Bodhisattva is said to have been accompanied by many sovereign kings. The mural depicts the Tang emperor flanked by foreign emissaries, who probably represent the sovereigns or envoys of Tang China's peripheral states. Therefore, the Tang emperor and foreign emissaries depicted in the wall paintings should be understood to have gathered to seek an audience to the conversation between Vimalakirti and Manjusri and to listen to the debate on the teachings of the Sutra and to reaffirm faith in Buddhism. The Zhenguan reign (627-649) under Emperor Taizong is when Tang China built the largest empire to date with the conquest of a number of states in western China, thus succeeding in identifying itself as the center of the universe. Zhang Yanyuan (張彦遠) in his Lidai minghuaji (歷代名畵記: Record of Famous Paintings of Successive Dynasties, 847) records that Taizong commissioned the court painter Yan Liben to paint portraits of visiting foreign emissaries.18 The fact that the wall paintings of Cave 220 show much more detailed illustrations of foreign emissaries than earlier murals of Dunhuang could be understood in the context of this record. Compared to the wall paintings of Caves 420, 314, and 380 from the Sui dynasty that depict scenes from the Vimalakirti Sutra, the wall paintings of Cave 220, which are representative of the early Tang period, exhibit a much more sophisticated composition and finer, more dynamic illustrations of the figures.19 In particular, the depiction of the emperor in his red garb with the imperial crown shows similarity in dress and crown with the portrait of Emperor Wendi (文帝, r. 581-604) of the Sui dynasty by Yan Liben, now housed in the Museum of Fine Arts, Boston. Therefore, the portraits in Dunhuang murals, which are contemporary with Yan Liben, are an invaluable source for the study of portrait paintings of that era.20 The Xuanhehuapu (宣和畵譜: Catalog of Paintings during the Xuanhe Era) records that Yan Liben worked on two portraits of Vimalakirti,21 which suggests that the Dunhuang murals were probably influenced by Yan's paintings, and that the new style of painting apparent in these murals emerged at that time. Another example of the foreign emissary wearing jougwan is found in the wall painting of Cave 335 in Dunhuang (Plate 6). This cave postdates Cave 220 as identified by the inscription “Chuigong (垂拱) second year” (686) on the east wall.22 The scene of Manjusri visiting Vimalakirti from the Vimalakirti Sutra covers the whole of the north wall of this cave and is known to be the largest in scale among such scenes illustrated in the Dunhuang caves. ( Plate 6 ) Foreign envoys on the northern wall of Dunhuang Cave 335 Chinese, Tang 686, wall painting Here, two people with feathers on their headdresses are portrayed just below Vimalakirti on the right, above the other visiting sovereigns. The two figures appear to be in conversation with one another. Their headdresses with two feathers are fastened by a wide headband run over both ears and tied in a bow under the chin. Although the painting is dark and it is difficult to recognize the details, these figures certainly wear the similar shape of jougwan depicted on the mural painting in the tomb of Prince Zhanghuai in Xi'an. Duan Wenjie (段文杰), formerly Director of the Dunhuang Research Institute (敦煌文物硏究所) has already noted the illustration of a person with jougwan in the wall painting of Vimalakirti and Manjusri in Dunhuang Cave 220,23 and Gwon Yeong-pil (權寧弼) has also written about the two people wearing jougwan in the mural of Cave 335.24 Compared to the scene depicted in Cave 220, the scene in Cave 335 has a more complicated, albeit similar, composition, and is more stylized in detail. This raises the possibility that there was an original version of this subject matter before the mid-seventh century on which the painting of Cave 335 was based, or that it was modeled on the painting in Cave 220 with some added variations. As the painting in Cave 335 is dated to 686, many years after the collapse of Goguryeo, it is conceivable that the Korean emissaries depicted were from Silla. However, if the painting is indeed based on the earlier model in Cave 220, it would make sense that the Koreans in cave 335 wearing Goguryeo-style garb represent emissaries from Goguryeo. As these foreign guests in the wall paintings were depicted as representatives of neighboring states attending the conversation between the retinues of Bodhisattva Manjusri and Vimalakirti along with the Tang emperor, it is possible to conjecture that the Koreans among them were portrayed in Goguryeo outfits based on the typical image of Koreans held by the Chinese, even sometime after the collapse of the Goguryeo state. This hypothesis carries more weight when we consider two factors: first, it is unlikely that Silla would feature in Tang cave paintings as one of the peripheral states of Tang; second, there is no known Chinese record of Silla people wearing jougwan, whereas the jougwan had been regularly recorded in ancient Chinese texts as a feature of Goguryeo dress. The Korean emissary in the tomb mural of Prince Zhanghuai can also be understood in the same way. As there is no evidence that the mural depicts an actual historical event, it is more likely that the emissary was portrayed according to the typical image of Koreans held by the Chinese, rather than based on a clear distinction of outfits between Goguryeo and Silla. Lastly, I have added another example of Goguryeo envoys wearing the same jougwan on the north wall of Cave 332, dated to 698 (Plate 7).25 They are present in the same scene painted in a near-identical composition and style, as seen on the north wall of Cave 335. This particular depiction has only been reproduced in a brief sketch and is only partially visible behind the big Buddha triad on the same north wall.26 However, I noted their presence in cave 332 during a workshop in Dunhuang in the summer of 2003 and reconfirmed my views during my recent visit to Dunhuang in June, 2008. ( Plate 7 ) Foreign envoys, from the painting on the northern wall of Dunhuang Cave 332 Chinese, Tang 698, wall painting After Dunhuang was occupied by the Tibetan state, Tufan (吐蕃), in the late eighth century, paintings of Vimalakirti Sutra continued to be commissioned. In those murals, the king of Tufan is depicted in his traditional garb but Goguryeo figures wearing jougwan are not found among the foreign emissaries.27 However, a painting of the Vimalakirti Sutra on silk, now a part of the Stein Collection in the British Museum, shows the Tibetan king at the front of the foreign emissaries and a Goguryeo person appears to be wearing jougwan at the left side of the painting. The two feathers on the headdress are arranged in a similar way as those illustrated on the wall in Cave 220 (Plate 8, see also Plate 5 for comparison).28 Apparently, the motif of foreign envoys produced in the seventh century seems to have continued into the late eighth century. ( Plate 8 ) Foreign Envoys, detail of the Vimalakirti Sutra Chinese, Tang, late 8th Century, ink and color on silk British Museum (Stein Collection 57 Ch.00350) Image from Roderick Whitfield, The Art of Central Asia: The Stein Collection in the British Museum, vol. I, Tokyo: Kodansha International, 1982, pl. 20. III GOGURYEO ENVOY DEPICTED ON SCROLL PAINTINGS IN THE NATIONAL PALACE MUSEUM, TAIPEI Two scroll paintings in the National Palace Museum of Taipei lend further credence to the suggestion that the Korean emissaries wearing jougwan in the Dunhuang wall paintings were of Goguryeo origin. One is Emperor Taizong Meeting with Foreign Envoys by Yan Liben of Tang (唐閻立本王會圖, Tang Yan Liben Wanghuitu, hereafter “Wanghuitu”), which portrays twenty-six foreign emissaries in color on silk along with the names of their respective countries. The other scroll, Foreign Envoys Paying Tribute to Emperor Yuandi of Liang State copied by Gu Deqian of Southern Tang (南唐顧德謙摹梁元帝蕃客入朝圖, NanTang Gu Deqian mo Liang Yuandi Fanke ruchaotu, hereafter “Fanke ruchaotu”), which is an ink drawing copied by Gu Deqian, depicts thirty-one foreign emissaries.29 These two works include emissaries from Goguryeo, Baekje, and Silla, among whom only the emissary from Goguryeo is wearing jougwan.30 In Wanghuitu, a Goguryeo emissary appears in a red jacket, wide green trousers with red trim, and a headdress decorated with feathers secured to the head with a band (Plate 9). Fanke ruchaotu also depicts a Goguryeo emissary, painted only in ink, wearing a headdress with two feathers (Plate 10). These two pictures are significant in that both contain an inscription that defines the emissaries wearing jougwan as being from Goryeo (高麗 which means Goguryeo [高句麗] then). ( Plate 9 ) Foreign emissaries from Emperor Taizong Meeting with Foreign Envoys painted by Yan Liben of Tang Chinese, Tang, 8th century or later copy, color on silk National Palace Museum, Taipei, Taiwan ( Plate 10 ) Foreign emissaries including a Goguryeo envoy, from Foreign Envoys Paying Tribute to the Emperor Yuandi of Liang Dynasty, copied by Gu Deqian of Southern Tang Chinese, Five Dynasties, 10th century, ink on paper National Palace Museum, Taipei, Taiwan Although it is difficult to date these paintings with precision, they are undoubtedly copies of the original Painting of Tribute Bearers to the Liang State (梁職貢圖, Liang zhigong tu), such as the one in the Nanjing Museum. Copies of artwork can in some cases prove to be a valuable source of information for the missing originals. Furthermore, these scroll paintings give a glimpse of the international nature of Tang culture of the time. Wanghuitu is dated to the late seventh century or is possibly a later copy, while Fanke ruchaotu dates to the tenth century. The fact that foreign emissaries of similar attire appear in many Dunhuang wall paintings highlights that the tradition of making copies was in practice for quite some time. Furthermore, the reason that the Dunhuang wall paintings contain only Goguryeo emissaries wearing jougwan demonstrates that this was the typical Chinese image of Koreans even after the collapse of the Goguryeo state. IV GOGURYEO PEOPLE ON THE COVER OF A SIX-LOBED SILVER BOX, AND TWO SARIRA RELIQUARIES The two scroll paintings in Taipei strongly indicate that the foreign emissaries wearing jougwan depicted in the wall paintings of Prince Zhanghuai's tomb, the Dunhuang caves, and the Sogdian palace were also from Goguryeo. In a similar vein, further details of Goguryeo people can be seen in other forms of Tang Chinese art. In 1979, a small silver box was excavated in Xi'an, near the grounds of Jiaotong University (文通大學). The silver box, now housed in the Bureau of Cultural Relics and Gardens (文物園林局) in Xi'an, was introduced in 1984 as Duguan qigeguo liuban yinhe (部管七箇國六瓣銀盒: Six-lobed silver box representing seven neighboring states under Chinese governance).31 It has also been displayed in the exhibition, Hidden Treasures of Chang'an in Japan in 1992 (Plate 11-1).32 Inside the lobe on the lid of this silver box are portrayed people wearing jougwan with an inscription 高麗國 next to them, providing essential information of the nationality of the people with feathers on their headdress (Plate 11-2).33 ( Plate 11-1 ) Six-lobed silver box representing seven neighboring states under Chinese governance Chinese, Tang, late 7th to early 8th centuries Discovered from Baoyingsi temple site in Xi'an in 1979 H: 5 cm, W: 7.5 cm, Weight: 121 g Office of Culture and Garden in Xi'an ( Plate 11-2 ) Five Goguryeo people wearing hat crowned with two feathers and inscription of “Goryeo State,” 高麗國, detail Photograph by Felix Schoeber The lid of the silver box is in the form of six oval-shaped petals with a pointed hexagonal compartment in the center. Against a background of eojamun (魚子文: punched ring patterns), each compartment has embossed images of neighboring states identified by names such as Poluomenguo (婆羅門國), Tufanguo (土蕃國), Shuleguo (疏勒國), Gaoliguo (高麗國), Baituo􀆶guo (白拓􀆶國), and Niaomanguo (鳥蠻國) with Kunlunwangguo (崑崙王國) at the center. In the seventh century, these neighboring states were under the political influence of the Tang empire, as confirmed by the inscription of 部管七箇國 (duguan qigeguo meaning seven states under governance) in the middle of the lid, and 將來 (jianglai meaning brought [as tribute]). The inscriptions conform to historical facts of territorial expansion undertaken by Tang and the practice of establishing regional governing posts in the neighboring states of the great Tang empire. Under the reign of Emperor Taizong and Gaozong, the Tang empire occupied the largest territory in its history and established its governing bodies, such as Anxi Duhufu (安西都護府) in Kashgar (疏勒) in the farthest west, Yingzhou Dudufu (營州都督府) in the northeast, and Andong Duhufu (安東都護府) in Pyeongyang, the capital city of Goguryeo, which was later moved to Liaodong (遼東). Turning to the details of the seven compartments on the lid of the silver box, one finds the following. In the hexagonal compartment at the centre, along with the inscription duguan qigeguo and Kunlunwangguo (崑崙王國) on its right side, six people in total are embossed including one riding an elephant in the center followed by another person holding a parasol high over him as well as the mahout sitting on the elephant's head. They are wearing short dresses that only cover their waists, suggesting that they come from a warm region. Kunlun refers to a tribe of people with dark skin color, as recorded in the Nanmanzhuan (南蠻傳: Record of Barbarians in the South) in the Jiu Tangshu: “the people with curly hair and dark skin in the south of Linyi (林邑) are generally called Kunlun.”34 The term seems loosely to refer to the Malays who lived scattered around the south Pacific and at the time possibly formed a small kingdom in a region between present day Sri Lanka (獅子國) and Canton (廣州).35 Inside the oblong compartment facing the elephant in the centre are five people wearing headdresses decorated with two long rabbit ear-shaped feathers – apparently jougwan – along with the inscription 高麗國. With a person slightly bigger in size on the left and the other four people facing him, all five people are clearly wearing jougwan of Goguryeo. As already mentioned above, the Goguryeo outfit is typically described in old Chinese records as a long-sleeved jacket with wide trousers, a leather belt, leather shoes, and a headdress decorated with feathers as seen on this silver box. In the petal to the right, the name of a state (白拓口國) is inscribed next to two people. An old man is sitting on a large, rectangular-shaped woven straw mat to the left, with a boy to the right presenting an offering to him. The name of the state has not been identified as the third character is missing.36 In the next petal to the right, under the inscription 鳥蠻國 are two people with their hair in topknots standing with backpacks on the left, and on the right three people seem to be receiving guests while standing. Niaoman (鳥蠻) is related to a contemporaneous tribe called Nanzhao (南詔), referring to a state, which existed in the northwestern region of Vietnam or the eastern region of Myanmar.37 The next petal has an inscription of 婆羅門國, which seems to refer to today's India. On the left stands a monk in his robe, holding a cane, and on the right are two people seemingly having a conversation with a long bottle between them. The bottle has a narrow neck and a round body with something long sticking out of its mouth. The scene seems to depict a certain religious ceremony. The next petal to the right illustrates 土蕃國, present-day Tibet. Two people are herding a yak, which is dashing forward at a gallop. This seems to symbolize the wildness of the Tibetan plateau. Tufan (土蕃) maintained a special relationship with Tang since 641, in the reign of Emperor Taizong, when a Tang princess Wencheng married the Tibetan emperor, and introduced Buddhism to that country. The last petal is engraved 疏勒國, today's Kashgar, where Anxi Duhufu was placed in the wake of Taizong's western expedition. Two people are standing on the right holding long swords and, on the left, is one person holding a sword and another, a bow, illustrating the militaristic ethos of the west Chinese tribes. This silver box was previously thought to have been used to contain perfumery or medicine on the assumption that it was found in the grounds of Xingqing Palace (興慶宮). However, as the excavation site was recently confirmed as the site of Baoyingsi (宝[寶]應寺: Baoying Temple) founded in 767 and which once stood in front of the palace, the decoration on the box is now assumed to be depicting the distribution of sariras.38 Among the artifacts displayed during the exhibition Hidden Treasures of Chang'an in Japan was yet another example of Goguryeo people wearing jougwan. It is a reliquary for sariras made of marble, found in 1990 in the historic site of Fachi-si (法池寺), a Tang Temple located in Caiguaicun (蔡拐村), Lantian (藍田), Shaanxi Province (陕西省). This marble casket, 33 cm long on all sides, portrays two people wearing feathers on their headdresses, sitting with other foreign guests (Plate 12-1).39 The scene depicts the distribution of sariras (舍利分配圖). In front of a person with a crown sitting inside a building with a rounded roof, an old man with a topknot and short trousers is portrayed as if he is taking something out of a big jar, probably a relic container. In front are three people sitting on each side, among whom, on the left, are two people wearing jougwan next to a short man who is probably from western China or Kunlun. On the right are two people wearing hats and an old man with a beard. These visiting foreigners seem to be sitting there to receive sariras taken from the jar. Further, a series of scenes such as the carrying of sariras on an elaborately decorated vessel (Plate 12-2), the burial of sariras in deep mountains, and the reception of sariras in a remote fortress are carved in low relief on the other three sides of this reliquary. ( Plate 12-1 ) Reliquary for sarira Chinese, Tang late 7-8th century, marble Unearthed at the temple site of Fachi-si, Caiguancun, Lantian, Shaanxi Province in 1990 H: 32 cm, W: 32 cm, Depth: 32 cm Cultural Relics Center of Lantian, Shaanxi ( Plate 12-2 ) Placement of sariras into a Jar, one side of the reliquary The style of the relief carving on the casket indicates that it dates from the early Tang and the foreign guests depicted on this marble reliquary share similar features with the figures illustrated in the silver box discussed above.40 Therefore, it is possible to conclude that the people wearing jougwan were also from Goguryeo who were included among the guests to receive sariras from the neighboring states of Tang. This assumption is further supported by the fact that an outer stone casket for a sarira reliquary in the Sumitomo collection, Senoku Hakukokan (泉屋博古館), Kyoto, known to have been brought from Shandong area in China, also shows three emissaries from Goguryeo in its depiction on all four sides of foreign envoys receiving sariras (Plate 13). The inscription on one side of the casket reads “Qianyuan Xiaoyi Huangdi baguo deng” (乾元孝義皇帝八國王等: Qianyuan reign of Emperor Xiaoyi and Kings of Eight States), alluding to the scene related to the distribution of the sariras to eight nations after the cremation of the Buddha. As the Qianyuan reign covers the period between 758 and 760, the date of this casket is of the mid-eighth century.41 ( Plate 13 ) Outer casket for sarira reliquary Chinese, Tang, 758-760, limestone H: 65.6 cm, W: 45.6 cm, Depth: 61.4 cm Sumitomo Collection Senoku Hakukokan, Kyoto, Japan V THE EXISTENCE OF LESSER GOGURYEO It is written in Tang records that the Goguryeo refugees stayed in the Liaodong region until the early ninth century and were governed under the lineage of Bojangwang (寳藏王: Silla king, r. 677-682) who had been appointed by the Tang Emperor. It is now time to look into the historic context in more detail. A number of Chinese historical texts record the fact that after Goguryeo's collapse in 668, refugees settled in the Liaodong region, where the state persisted in another form in the area under the protection of Tang. While the Jiu Tangshu only records the history of Goguryeo until the time of Empress Wu Zetian,42 the Xin Tangshu states that a reconstituted Goryeo (meaning Goguryeo) existed in Liaodong as a peripheral state of Tang until it was absorbed into Balhae (渤海, 698-927; Chinese: Bohai), which was founded by a Goguryeo descendant, Daejoyeong (大祚榮, r. 699-719).43 The Samguksagi records the same facts in the chapter on Goguryeo.44 From this viewpoint, when it comes to Tang artists' propensity to draw images of Koreans paying tribute to Tang in the manner of Goguryeo people wearing jougwan, the following argument can be laid out. In the wake of the collapse of Goguryeo in 668, refugees were scattered over various areas. Unified Silla absorbed some of them, while captured officials and members of the Goguryeo royal family lived in Chang'an (present Xi'an). Others joined the Tang military to protect the western frontiers or were dispersed across China. Some people became a part of Balhae as it established itself into a new state while others joined ethnic tribes in Liaodong and nomadic Mongolia. There were also some who went to Japan. This paper is primarily interested in the refugees in Liaodong.45 After the fall of Goguryeo, Tang established the satellite regime of Andong Duhufu in Pyeongyang, with ambitions to rule the old territory of Goguryeo as well as Silla. But Silla put up a fierce resistance both by land and sea. When Goguryeo refugees in Liaodong sought to re-establish their state, Silla and Tang gathered forces to crush the movement. However, after their joint victory, Silla and Tang came into conflict again and this led Tang to relocate permanently its satellite governing post from Pyeongyang to Liaodong in 676. After that date, Goguryeo refugees in Liaodong were granted a broader range of autonomy, although they remained as Tang subjects and under Tang control. In order to strengthen the governance of the Goguryeo refugees in Liaodong, Tang appointed in 677 Gaozang (高藏: Bojangwang, the last king of fallen Goguryeo) as the governor of Liaodong as well as the king of the refugees. Tang also had Goguryeo refugees who had previously been sent to inner Tang territory in 669 relocated to the area. Bojangwang's jurisdiction fell well within the control of Andong Duhufu, a title given to the head of a peripheral tribe rather than an official of the Tang empire. Therefore, Bojangwang was a symbolic head of Goguryeo refugees within the hierarchy of Andong Duhufu. Despite Tang's efforts to maintain its stronghold in Liaodong, it proved to be a difficult task. Tang appointed Go Bowon (高寶元, dates of birth and death unknown), Bojangwang's grandson, new king of the peripheral state in 686 in the wake of the discovery of Bojangwang's collusion with the Mohe (靺鞨) tribe. In 698 Tang appointed Go Bowon zhongcheng guowang (忠誠國王: loyal king) with a view toward maintaining control over the Goguryeo migrants in Liaodong, but to no avail. However, Tang China was still able to maintain indirect control of Liaodong through granting the Goguryeo's royal family exile in the capital, Chang'an. In 699, Tang downgraded Andong Duhufu to Andong Dudufu (都督府) and appointed Bojangwang's son, Go Deokmu (咼德武, dates of birth and death unknown), as governor. Hino Kaizaburo (日野開三郞) considers this point in time as the start of the state of Lesser Goguryeo (小高句麗) that was led by Goguryeo's royal family.46 It is generally considered that the new state could have survived due to the fact that the Liaodong region played the role as a buffer between Tang and Balhae, which was established in 698, following a series of military conflicts in the region.47 Moreover, Tang selected another colonial government of Xiongjin (熊津) Dudufu in Jian'an (建安) of Liaodong to govern the refugees of Baekje, appointing Prince Buyeoryung (夫餘隆) of Baekje as the Governor King Daifang (熊津都督帶方郡王). But Prince Buyeoryung died in 682 while in exile in Goguryeo.48 Moreover, it is recorded that in 725 the King of Goguryeo as well as Governor King Daifang of Baekje were present in the religious ceremony of Emperor Xuanzong (玄宗) as officials of peripheral localities (內臣之蕃, neichen zhifan).49 This suggests that Goguryeo was considered as a locality of the Tang empire inhabited by ethnic minorities. In contrast, the same text records Tuque, Japan, and Silla as peripheral states paying tribute (朝獻之國, chaoxian zhiguo), indicating that they were autonomous states, albeit under the political influence of Tang. Goguryeo's tributary mission toward the end of the Yuanhe (元和 806-821) reign is recorded in the Xin Tangshu,50 which the Samguksagi correspondingly records as having taken place in the year 818.51 This suggests that the state of Lesser Goguryeo was established in Liaodong and that it had a semi-autonomous status, paying tribute to Tang until the early-ninth century. Lesser Goguryeo was, however, absorbed into Balhae during the reign of its tenth king, Daeinsu (大仁秀, r. 813-840). In summary, Tang seems to have paid respect to the royal family of Goguryeo by appointing Go Bowon as King of Goguryeo in 686, and Go Deokmu as the governor representing Goguryeo refugees. But Tang in fact kept them both in Chang'an in order to maintain indirect control of Liaodong. Roh Taedon's analysis summarizes well the peripheral status of Lesser Goguryeo: the members of the royal family in exile were used to keep up the imperial appearances of the Tang dynasty and reinforce its authority. In other words, they were asked to appear in official ceremonies to manifest externally to the kings of vassal states the power and benevolence of imperial China, and internally to show Tang subjects the imperial dominance and charismatic influence of the moral Son of Heaven (天子, tianzi) over its numerous neighboring states.52 VI CONCLUSION This paper has considered Goguryeo's presence from the late seventh century to the early-ninth century as a peripheral vassal state of Tang. Some of the remaining refugees of Goguryeo served the Tang military and contributed to the Tang expedition as far as to Sogdiana as led by the General Gao Xianzhi (Go Seonji, 高仙芝, d. 755), who was himself a Korean from Goguryeo. On the other hand, the Goguryeo people also played an important political role in Liaodong, maintaining a complex diplomatic relationship with the neighboring states of Qidan (契丹), Tuque, and Mohe, and establishing a new state of Balhae at the end of the seventh century. Returning to the discussion on the portraits of people wearing jougwan, this paper proposed that the wall paintings in Dunhuang Caves 220, 335, and 332 are likely to have depicted emissaries of Goguryeo, a powerful neighboring state, in the presence of the Tang emperor. Before its fall, Goguryeo was the strongest of the Three Kingdoms and had fought against the Chinese armies of Sui Yangdi and Tang Taizong and had defeated them. Therefore it seems to be natural that even after Goguryeo fell in 668, the Tang Chinese still continued to keep in their minds the Goguryeo envoy with feathered hat as a stereotypical image of Koreans. It is therefore likely that the tomb murals or wall paintings were either based on a model of well-established images at the time or were copied after the event, rather than the artists' keen understanding of history or political knowledge regarding the neighboring states. The fact that even after the collapse of Goguryeo, the remaining royal descendants were appointed to rule Liaodong in the form of a peripheral state of the Tang empire strongly suggests that the artists of the time were likely to have depicted Goguryeo people as emissaries from neighboring states. In particular, the silver box of Xi'an and the two scrolls Wanghuitu by Yan Liben and Fanke ruchaotu copied by Gu Deqian in Taipei clearly state in the inscription that the people in jougwan are from Goguryeo. This recognition in the appearance of Goguryeo people with the characteristic feature of the feathered hat, also applies to the identification of the figures with jougwan at the scene of the distribution of sariras to neighboring kings carved on the stone reliquary in Senoku Hakukokan.