Keyhole-shaped Tombs in the Yŏngsan River Basin: A Reflection of Paekche-Yamato Relations in the Late Fifth.Early Sixth Century

Keyhole-shaped tombs were originally thought to only exist on the Japanese archipelago and to mark the hegemony of the early Japanese Empire. The discovery of keyhole-shaped tombs on the southwestern Korean peninsula in the Yŏngsan River basin reignited the debate on the nature of the relationship between early "Korean" and "Japanese" polities between the fourth–seventh centuries CE. Questions about the identity of those buried in the "Korean" keyhole-shaped tombs became a hot topic: Were they Paekche? Japanese or Wa? Or Mahan? These tombs only appeared for a single generation in the late fifth–early sixth century in a region that was quickly adopting influences from Paekche and the Japanese archipelago until the region was absorbed by Paekche by the early sixth century. Based on the limited textual evidence, I argue that Paekche's move south in 475 and the instability of the Yamato court in the late fifth and early sixth centuries created opportunities for the local elites in the Yŏngsan River basin to make connections with both Paekche, Yamato, and other regional elites in northern Kyushu to create an eclectic and multicultural environment that gave rise not only to keyhole-shaped tumuli but also to other burial innovations in the Yŏngsan River basin.

With the rise of nationalism came the claiming of historical monuments and archaeological sites as physical manifestations of national identity. In East Asia, historical ownership of material culture is further reinforced by long historio-graphical traditions that emphasize political and cultural continuity to an ancient past. Like the Egyptian pyramids, one prominent example of this intersection of nationalism, archaeology, and historical texts in East Asia is the keyhole-shaped tumuli (zenpō kōenfun 前方後円墳)¹ in Japan. Ranging from several dozens to several hundreds of meters in length, these massive mounded tombs seem ubiquitous on the Japanese archipelago. The Daisen Kofun 大仙陵古墳 in present-day Osaka, Japan is the largest and most famous example, measuring a maximum of 840 meters long, 654 meters wide, and 35.8 meters high with the central mounded tomb surrounded by moats (see Figure 1).²

Figure 1.

The Daisen Kofun in Sakai City, Osaka, Japan³

Japanese nationalist historians claim that the largest of these massive mounded tombs or kofun 古墳 (K. kobun) were the resting places of "emperors"⁴ described in the earliest extant Japanese historical sources, the Kojiki 古事記 (compiled 712 CE⁵) and the Nihon shoki 日本書紀 (compiled 720). Therefore, the appearance of keyhole-shaped tumuli (ca. mid-third century) marked the beginning of "Japan" and the Japanese nation. In other words, keyhole-shaped tumuli archaeologically symbolize the beginning of an "unbroken Japanese imperial line" and the "integration of the [Japanese] nation."⁶ As highly conspicuous physical representations of Japanese national identity seen through historical texts, keyhole-shaped tumuli were considered to be a uniquely "Japanese" burial system and presumed only to exist on the Japanese archipelago.

However, South Korean scholar Kang In'gu shattered this understanding when he announced that he had confirmed the presence of keyhole-shaped tumuli on the Korean peninsula, more specifically in the Yŏngsan River basin in South Chŏlla Province.⁷ On the one hand, Korean scholars had successfully ignored or refuted Japanese nationalist claims that an early Japan had conquered and controlled the southern part of the Korean peninsula from the fourth–sixth century⁸, due to a lack of archaeological evidence, but these tombs threatened to overturn that view. Likewise, Kang's assertion that keyhole-shaped tumuli originated in Korea challenged the Japanese nationalist narrative that they symbolized the autonomous formation of a native Japanese state. Since then, scholars on both sides of the Korea Strait have produced a wide range of opinions on this topic due to its importance in understanding early Korean-Japanese relations and its potential impact on national identity narratives. The central question became, who were buried in these keyhole-shaped tumuli on the Korean peninsula?

This article seeks to understand the geopolitical situation of Paekche and Yamato, as seen through historical sources, that may have led to construction of the keyhole-shaped tumuli in the Yŏngsan River basin (hereafter "YSR keyhole-shaped tumuli"). Although there are no written records directly pertaining to the Yŏngsan River basin during the period when the keyhole-shaped tumuli appear (late fifth–early sixth century), a survey of the geopolitical situation of Paekche and Yamato may help historically contextualize any political or social changes in the Yŏngsan River basin and identify external events that may have contributed to the construction of the YSR keyhole-shaped tumuli.

BRIEF ARCHAEOLOGICAL BACKGROUND

Although the YSR keyhole-shaped tumuli attract the most attention, the Yŏngsan River basin was already considered an unusual region due to the use of jar coffin burials there since the late third century. This contrasts with the use of wooden or stone burials found in the rest of the Korean peninsula. The material culture of the Yŏngsan River basin was so different from regions associated with Paekche, Silla, Koguryŏ, or even Kaya that Korean archaeologist Hwang Yonghon described it as a "cultural island."⁹ Some scholars even argued that the Yŏngsan River basin was the last bastion of remnant Mahan¹⁰ that were not immediately annexed by Paekche early in its expansion.¹¹ As far as the author is aware, no prestige goods from Paekche, Kaya, or the Japanese archipelago have been found in elite tombs jar burials in the Yŏngsan River basin prior to the fifth century, instead preferring its own local prestige goods and ritual pottery.

This changed in the late fifth–early sixth century with the appearance of Paekche gilt-bronze crowns and shoes, decorative horse trappings as well as beads and swords from the Japanese archipelago. Another major innovation starting in the late fifth century was the introduction of horizontal corridor stone chambers. Instead of placing a coffin into a pit and covering it up, a horizontal corridor stone chamber tomb had a stone chamber constructed in the center, where the coffin and ritual objects could be placed, that was linked to an opening at the side of the tomb via a short corridor. The chamber and corridor would be large enough for someone to enter the tomb and place additional coffins or perform later ceremonies inside the tomb after the coffin was placed inside. The unfortunate side effect of this new tomb style was it made robbing the tomb far easier. This innovation on the Korean peninsula and the Japanese archipelago began in Paekche but apparently spread to the Yamato core region in present-day Nara on the Japanese archipelago and spread westward to Kyushu before finally reaching the Yŏngsan River basin.¹² It is quite a circuitous route to take, considering the Yŏngsan River was immediately south of Paekche, but it speaks to the relationships that the Yŏngsan River had with people in northern Kyushu. The shapes of tombs also began to change. The earlier jar coffin mounds were originally U-shaped or semicircular in shape, but the late fifth century saw a diversification of tomb shapes, including triangular-shaped tombs, and, of course, the keyhole-shaped tombs.

At present, there are fourteen confirmed YSR keyhole-shaped tumuli, and they were all constructed between the late fifth and early sixth century,¹³ meaning that this tomb style only lasted only one or two generations.¹⁴ This is in sharp contrast to the tens of thousands of keyhole-shaped tumuli found on the Japanese archipelago with constructions dates spanning the third to sixth century. Additionally, the YSR keyhole-shaped tumuli are peripherally interspersed among the far more common jar-coffin tumuli, the political center presumed to being the massive cemetery complex in Pannam-myŏn near Naju in South Chŏlla Province.

What could have prompted this flurry of burial style changes and influx of prestige goods from Paekche and the Japanese archipelago? The answer may partially lie in the historical records of Paekche and Yamato.

PAEKCHE AND YAMATO IN THE LATE FIFTH AND EARLY SIXTH CENTURY

To situate the late fifth and early sixth century historically for Paekche and Yamato, see Table 1. For Paekche, this period starts from the reign of King Kaero (r. 455–475) through King Sŏng (r. 523–554). For Yamato, this corresponds to the reigns of Yūryaku (r. 456–479) through Keitai (r. 507–535).¹⁵

Table 1.

Paekche and Yamato Rulers from the Late Fifth–Early Sixth Century

As mentioned above, virtually no known records directly concerning activities in the Yŏngsan River basin exist. However, the increasing presence of burial prestige goods from Paekche and the Japanese archipelago suggests increased contact and developing relations. The Yŏngsan River basin's placement along the maritime trade route between Paekche and Yamato also makes it sensitive to events that might affect relations between Paekche and Yamato. To better understand the circumstances of both, we must turn to historical sources such as the earliest extant native Korean history the Samguk sagi (comp. 1145) and the Nihon shoki. But even those sources are limited and conflicting. The "Paekche Annals" of the Samguk sagi is remarkably silent on Paekche's relations with the Yamato court between 428 and 653. Jonathan Best notes that the compilers of the Samguk sagi had very little material on Paekche and probably did not have access to Japanese source materials.¹⁸ Therefore, the richest materials still available on Paekche-Yamato relations, albeit scanty, are from the Nihon shoki. Other than shifts in the capital, the "Paekche Annals" barely mentions Paekche's southern frontier and says nothing about its southward expansion. For information on most of this, we also have to rely on the Nihon shoki for indirect clues at best, for even the Nihon shoki mentions nothing directly. Interestingly, most of our knowledge about Paekche's frontier and regional administration comes from Chinese dynastic sources, and even they are also extremely limited. Even with these limitations, the available historical sources can provide a rough sketch of important political and military events that may have influenced the Yŏngsan River basin.

THE FALL OF HANSŎNG AND THE MOVE TO UNGJIN (475)

It is a curious historical coincidence that the construction of keyhole-shaped tumuli in the Yŏngsan River basin begins around the fall of Paekche's original capital of Hansŏng (present-day Seoul) in 475, one of the most traumatic events in Paekche history.¹⁹ Koguryŏ had been constantly threatening Paekche from the north since at least the fourth century and even assaulted Paekche's capital once before.²⁰ The year 475, however, was the first time that Koguryŏ took the Paekche capital, killed its king Kaero, and completely removed the Paekche ruling elite from the Han River basin. With most of the royal family captured or killed, the surviving remnants re-established a new capital to the south at Ungjin (present-day Kongju). This led to conflicts between the displaced Hansŏng elite and the local elites of the new capital. Militant factionalism consumed the Paekche court, which led to a swift succession of short-lived rulers until the reign of King Tongsŏng (r. 479–501), who had more success in consolidating his rule but ultimately was assassinated within the context of infighting among the Paekche elite. According to the Samguk sagi, Paekche only began to recover during the reign of Muryŏng in the early sixth century. In other words, Paekche was military weak and somewhat politically unstable for most of the late fifth century.

The construction of the YSR keyhole-shaped tumuli, as well as the other materials changes happening in the Yŏngsan River basin, sometime after Paekche's move south to Ungjin suggests a possible connection between the two. The southern relocation of the Paekche capital and the loss of its territory in the Han River basin would have undoubtedly changed the geopolitics of the southwestern Korean peninsula. Ungjin, on the Kŭm River, originally marked the southern edge of Paekche territory, meaning that Paekche's center had now shifted to its southernmost limit. Confronting the possibility that Koguryŏ would have continued further south, the remnant Paekche elite would have sought refuge as far south on territory it solidly controlled without leaving it altogether. The fortress at Ungjin would have made a logical choice in that regard. Paekche had not made any inroads into the Yŏngsan River basin prior to the late fifth century. With Paekche's interests now turning south, this would have created more opportunities and incentives for Paekche to cultivate relationships in the Yŏngsan River basin, especially in its weakened condition soon after its move to Ungjin. Later, as Paekche gained in military and economic strength, it could have converted those relationships with Yŏngsan River basin polities into annexation. In this regard, this change in geopolitics presented both an opportunity and a threat to the polities in the Yŏngsan River basin. Paekche offered advanced technology, culture, and political legitimacy, while the Yŏngsan River basin offered human as well as agricultural resources from its large swaths of arable land. At the same time, increased Paekche influence also threatened the autonomy of the Yŏngsan River basin polities, which may have encouraged local polities opposed to Paekche to seek other places for support, such as polities on the Japanese archipelago.

YAMATO INSTABILITY

The Japanese archipelago was far from a singular unified empire as described in the Nihon shoki and was more a loosely connected network of polities with Yamato (located in present-day Nara, Japan) at the center. Although the Nihon shoki claims places, such as Tsukushi (present-day northern Kyushu), as provinces under direct control of the Yamato court, it was merely an anachronistic projection of Yamato's eighth-century political situation into the past. In reality, the Yamato court's hold on Kyushu up until the early sixth century was based on a series of constantly negotiated relationships (fictitious or actual) with regional authorities,²¹ who had their own autonomy and own relations with other regions, such as the Yŏngsan River basin. This was especially true in the late fifth century after the death of Yūryaku in 479.

The Yamato court was plagued with succession issues, competing factions, a rash of "rebellions" of powerful lineages, as the Yamato court tried to continue it process of consolidating rule over the Japanese archipelago. Things became so dramatic that one of the imperial princes was roasted alive.²² For almost the next three decades there was a succession of several rulers with relatively short reigns until Keitai in 507. Even Keitai, according to the Nihon shoki, was an outsider of the Yamato core and originally came from Tamba Province 丹波国 (part of present-day Kyoto Prefecture and Hyōgo Prefecture). In other words, powerful local authorities outside the Yamato core region were competing for power at the Yamato court.

This instability at the Yamato court weakened its influence on frontier local authorities, such as those on Kyushu, who were part of the Yamato network of kinship ties. This weakening may have incentivized local authorities, such as those in Kyushu, or those outside the Yamato court to further strengthen their own ties with other regions, such as the Yŏngsan River basin. We can see evidence of this interaction with the appearance of material culture from the Yŏngsan River in northern Kyushu along the Ariake Sea as well as Kyushu prestige goods and Kyushu-style horizontal corridor stone chambers appearing in the Yŏngsan River basin tombs in the late fifth and early sixth century.²³ As with Paekche, Yamato found itself in a weakened state for most of the late fifth century and only began to reassert itself in the early sixth century.

This idea of powerful independent Kyushu elites is supported in the Nihon shoki with examples such as the story of Iwai's Rebellion. Kyushu had the geographic advantage of occupying the middle of the valuable trade route between the Korean peninsula and the Yamato court. For the most part, the Nihon shoki portrays Tsukushi in northern Kyushu as being a cooperative partner in its dealings with the southern Korean peninsula. This, however, was not the case when a powerful local Kyushu leader, Iwai of Tsukushi, resisted Yamato's attempt to support the Kaya region in 527.²⁴ This shows that the regional authorities in Kyushu were strong enough to keep an expeditionary force from Yamato at bay if they wanted to and had considerable control over the sea routes between the Korean peninsula and the Yamato court. Second, the Yamato court called Tsukushi "the territory of the Western Barbarians" or seijū 西戎, which could also translate to the "western wilderness," and is probably used here to denigrate Tsukushi, but it also points out that Tsukushi is Yamato's western borderland, which it obviously did not have full control over. It is possible that the prior instability at the Yamato court after Yūryaku's reign encouraged elites in Kyushu to rely less on Yamato and to grow more independent, resulting in this military conflict. It is also telling that the Yamato ruler "allowed" the local elites to rule Kyushu after putting down Iwai's rebellion, leaving only Honshu for himself. Even at this point, Yamato continued to rely on these relationships to enforce its influence. As Iwai refused to acknowledge the reciprocal responsibilities of the relationship, he was eventually replaced by someone who would.

As for this account's relevance to the Yŏngsan River basin, it indirectly shows that the Kyushu region during the construction period of the YSR keyhole-shaped tumuli had the autonomy to form its own relationships with groups on the Korean peninsula, as Iwai had also tried to bribe Silla officials for his own purposes. Instead of seeing the keyhole-shaped tombs in the Yŏngsan River basin as extensions of Yamato power, they could merely be reflecting ties with other local elites in northern Kyushu during the late fifth–early sixth centuries.

LOCAL AUTHORITIES WITH PAEKCHE AND YAMATO AFFILIATION

In the milieu of a disrupted Paekche and Yamato, it appears that local authorities in between the Paekche-Yamato trading network took advantage of the situation to further their own political and economic interests. However, Paekche also quickly started engaging with its southern neighbors and began expanding its influence. We see evidence of this with the discovery of Paekche prestige goods, such as Paekche gilt-bronze crowns and shoes, found in most of the local authority tombs of the Yŏngsan River basin, such as Sinch'on-ni Tomb 9 (present-day Pannam-myŏn in Naju, South Chŏlla Province). At the same time, there were also prestige goods from the Japanese archipelago, such as swords and beads, in the same tombs. There was considerable variation in the piece-meal eclectic nature of Yŏngsan River basin tombs in the late fifth–early sixth century. The keyhole-shaped tombs there simply could have represented one of these variations as local elites in the Yŏngsan River basin competed for recognition from the Paekche or Yamato court or simply tried to represent their strong ties with other local elites in Kyushu or other parts of the Japanese archipelago.

We see some possible evidence of this in the Nihon shoki but for the neighboring region of Kaya. Although the Yŏngsan River basin is never mentioned, local authorities carrying title and ranks from both the Paekche and the Yamato courts from the Kaya region are recorded. It is entirely possible that local authorities from the Yŏngsan River basin could have similarly held title and ranks from Paekche and Yamato courts as well.

The earliest example of this recorded in the Nihon shoki is Hozumi no Omi Oshiyama 穂積臣押山,²⁵ who was the regional authority of Tari 哆唎國 in the Kaya region. In 512, the Yamato court had Oshiyama send forty-four horses from Tsukushi as tribute to Paekche.²⁶ It seems odd that the Yamato court would ask a Kaya regional authority to procure horses from Kyushu to send to the Paekche court, unless he had strong connections with both courts and Kyushu. It may also indicate Oshiyama's influential position on the Paekche-Kaya-Kyushu-Yamato trading network. In the twelfth lunar month of the same year, Oshiyama accompanied a Paekche envoy to Yamato requesting ownership over the following "four districts of Mimana" 任那四県: Upper Tari 上哆唎, Lower Tari 下哆唎, Sat'a 娑陀 (J. Sada), and Moru 牟婁. Oshiyama, as the authority of Tari, made the following impassioned plea:

[512] Hozumi no Omi Oshiyama, the governor of the land of Tari, petitioned the [Yamato] emperor saying, "These four districts are closely connected to Paekche but distantly separated from Japan. Morning and night [Paekche and Tari] exchange communications, and our chickens and dogs [are so close to each other] it is difficult to distinguish [whose are whose].²⁷ If [these counties] are now bestowed to Paekche and joined together to form the same country, then there is no better policy. True there may be a threat to future generations if [these counties] are bestowed to make a unified country. But how many years can [they] be defended if they are separate?

哆唎國守穗積臣押山奏曰「此四縣、近連百濟、遠隔日本、旦暮易通、 鶏犬難別。今賜百濟合爲同國、固存之策、無以過此。然縱賜合國、 後世猶危、況爲異場、幾年能守。」²⁸

The request was subsequently granted.²⁹

Oshiyama's impassioned plea tells us several things. One, Paekche at this time had expanded far enough to neighbor Tari on the Sŏmjin River. Second, Oshiyama had calculated that merging his territory with Paekche would be more beneficial than being independent or continuing relations with Yamato. This provides an example of a regional elite willing to merge his territory into Paekche and join the ranks of Paekche's elites for political and/or economic reasons and not within the context of a conquest. A similar situation may have been developing in the Yŏngsan River basin as well.

On the other hand, some Japanese scholars believe that the "four districts of Mimana" are in the Yŏngsan River basin.³⁰ All of these theories, however, are based on very arbitrary and questionable linguistic work done by Suematsu Yasukazu, who looked at Paekche place names and tried to find ones that sounded similar to the ones of the four districts in present-day South Korea. One example is his arbitrary decision that Moru was the same as present-day Muan in South Chŏlla Province. According to the "Monograph of Geography" in the Samguk sagi, the old Paekche name for Muan was Murahye 勿阿兮. Suematsu argues that this name would have been pronounced "mur-a-xoi" at the time, which happens to sound similar to Moru.³¹ No other evidence is presented, and no further analysis or even sources for his reconstruction are given. Yet many Japanese nationalist scholars have taken his geographic designations as canon and do not question them. In addition, if Oshiyama is considered a regional authority of the Yŏngsan River basin, his participation in the following Paekche tribute mission to Yamato is problematic:³²

[513] Seventh Year, Summer, Sixth Month. Paekche sent General Chŏmi Mun'gwi [J. Sami Monki] and General Churi Chŭgi [J. Tsuri Soni], along with Hozumi no Omi Oshiyama (the Paekche Annals say, "Commissioned Lord Oshiyama.") to present a scholar of the Five Confucian Classics Tan Yangi³³ and a separate memorial that said, "the land of Panp'a [Tae Kaya]³⁴ has seized our land of Kimun [J. Komon].³⁵ We humbly wish that your Heavenly favor can restore it to its original jurisdiction."

七年夏六月、百濟遣姐彌文貴將軍・洲利卽爾將軍、副穗積臣押山[百濟本記云、 委意斯移麻岐彌]貢五經博士段楊爾、別奏云「伴跛國、略奪臣國己汶之地。伏願 、天恩判還本屬。」³⁶

If this passage is accurate, Tae Kaya (i.e. Panp'a) and Paekche were both expanding into the upper reaches of the Sŏmjin River basin and apparently clashing with each other. Not only did Paekche absorb Kimun, but several months later, thanks to Yamato diplomacy, it had also absorbed Taesa 帯沙 (present-day Hadong, South Kyŏngsang Province) at the mouth of the Sŏmjin River.³⁷ Oshiyama's participation in this would only make sense if he were already a regional authority of the Sŏmjin River and could negotiate on behalf of Paekche. As a local authority of the Yŏngsan River basin, he would have very little influence on the outcome of this negotiation.

The presence of Oshiyama as a regional authority of the Sŏmjin River basin and mediator between Paekche, Yamato, and the Sŏmjin River authorities is also a possible model for something similar happening in the Yŏngsan River basin. It is also interesting to note that Oshiyama is still recorded with his Yamato title Hozumi no Omi while also holding a Paekche title as well. Kimun or Taesa was not one of the "four districts" that Paekche had acquired the previous year, but it is clear that Paekche's expansion into the Sŏmjin River basin was nearing completion by this point, according to the Nihon shoki. A similar expansion may have also been occurring in the Yŏngsan River basin as well, since the Yŏngsan River basin is geographically adjacent to the Sŏmjin River basin. Based on Paekche's expansion into the Sŏmjin River basin, the local authorities in the Yŏngsan River basin would have been keenly aware of Paekche's southern intentions. Whether some embraced it, as Oshiyama had, or resisted it is a question that cannot be answered with the extant texts.

PAEKCHE OFFICIALS OF WA-DESCENT

Other theories on the identity of the occupants of the keyhole-shaped tumuli in the Yŏngsan River basin argue that they are Paekche officials of Wa-descent. Wa being loosely defined as natives of the Japanese archipelago who subscribed to the keyhole-shaped tumuli culture. There are several variations of this, but they all try to explain the eclectic nature of the YSR keyhole-shaped tumuli (i.e. combination of Paekche, Japanese archipelago, local elements) by creating a scenario where someone from the Japanese archipelago is recruited by Paekche and is essentially bicultural, which is then reflected in their burial style.

Interestingly, the Nihon shoki does mention a number of Paekche official of Wa-descent, and they suddenly appear in the early part of Yamato ruler Kinmei's reign (r. 539–571), which overlaps with Paekche's King Sŏng's reign (r. 523–554). Paekche's foreign relations situation at the time consisted of fighting with Koguryŏ in the north and confronting Silla in the Kaya region in the south. In the early sixth century, Silla began expanding into the Kaya region, which, according to the Nihon shoki, alarmed Paekche and Yamato. As a result, many envoys were sent from Paekche and Yamato to the Kaya region to try to find a solution to the Silla crisis. Unable to find a diplomatic solution, the situation quickly deteriorated into a conflict, which the Kaya polities lost. The following is a chronological list of Paekche officials of Wa-descent found in the Nihon shoki during the reigns of Keitai and Kinmei:

[512, 513, 529] Hozumi no Omi Oshiyama 穗積臣押山[516] Shinano Ahita 斯那奴阿比多³⁸[541,542, 544] Nasol Ki no Omi 紀臣奈率³⁹[543] Sidŏk Mononobe no Makamu 物部施徳麻哿牟⁴⁰[544; 545] Nasol Mononobe no Yōkata no Muraji 物部連奈率用歌多⁴¹[544] Nasol Kose Kama 許勢奈率歌麻⁴²[544] Nasol Mononbe no Kahi 物部奈率歌非⁴³[553] Nasol Shinano Shishu of the Upper Division 上部徳率科野⁴⁴[553] Kawachibe no Ashihita 河内部阿斯比多⁴⁵[553] Nasol Shinano Shiragi of the Upper Division 上部奈率科野新羅⁴⁶[554] Nasol Mononobe no Kaku of the Upper division 上部奈率物部烏⁴⁷[554] Governor of the Eastern Province Mononobe no Makamu no Muraji 東方領物 部莫哥武連⁴⁸

The lineages of some of these Paekche officials of Wa-descent are the same as those that form the core of the Yamato court, such as Ki 紀氏, Mononobe 物部氏, Shinano 科野(斯那奴, 斯柰奴)氏, Kose 許勢氏, which suggests that they had connections at the Yamato court as well.⁴⁹ With the exception of Mononobe no Makamu no Muraji, these Paekche officials of Wa-descent functioned primarily to request military assistance from the Yamato court or were sent to Kaya to devise strategies to restore the Kaya polities that were being annexed by Silla.

Among the above-listed historical figures who had dual Paekche-Yamato ranks and title, Mononobe no Makamu no Muraji is notable because he served as a high-ranking Paekche military officer and was even governor of Paekche's Eastern Province 東方. Since provincial governors had to have held at the rank of talsol [second highest rank out of sixteen], Mononobe no Makamu no Muraji would have been the highest-ranking dual Paekche-Yamato official ever recorded. In 554, as the governor of the Eastern Province, he was assigned to attack Silla's Hamsan Fortress (present-day Okch'ŏn), where Paekche's King Sŏng was killed. Mononobe no Makamu no Muraji was of the Mononobe lineage, who handled military affairs in Yamato. The headquarters of the Eastern Province was in present-day South Ch'ungch'ŏng Province. Therefore, if he had been buried in a keyhole-shaped tumulus, it would have been there or in the Paekche capital of Sabi but not in the Yŏngsan River basin.

Starting in the 540s, there was a sudden jump in the number of recorded Paekche officials of Wa-descent. This also happens to coincide with Paekche's annexation of the Yŏngsan and Sŏmjin River basins. This suggests that there may have been local authorities of Wa-descent, who had connections with both Paekche and Yamato and were absorbed into the Paekche central elite. In fact, Paekche's countryside already seemed to have been quite diverse as indicated by the Chinese dynastic histories Suishu 隨書 (compiled in 636) and the Nanshi 南史 (compiled in 659):

A province [in Paekche] has [up to] ten prefectures. Each prefecture has a general. The people are a mix of Silla, Koguryŏ, Wa, etc., and there are even Chinese.

方有十郡，郡有將。其人雜有新羅、高麗、倭等， 亦有中國人。⁵⁰

Paekche could have found these local authorities of Wa-descent, quite useful for their connections with other regional authorities and their positions on the Paekche-Kaya-Kyushu-Yamato trade network. Yŏn Minsu speculates that these Paekche officials of Wa-descent would have also been useful for their language skills as well, although the Paekche and Wa seemed to have had no trouble communicating previously.⁵¹

The main problem connecting these figures to those buried in the YSR keyhole-shaped tumuli is the chronology. These Paekche officials of Wa-descent appeared after the last of the YSR keyhole-shaped tumuli were constructed, so they could not be the ones buried in them. The only possibilities are that the above Paekche officials of Wa-descent are the offspring of those buried in the YSR keyhole-shaped tumuli or that the tomb chronology is incorrect. The chronology of the texts is difficult to question since the circumstances of their activities (i.e. Silla's annexation of the Kaya region) is cross-verified in other texts, such as the Samguk sagi.

HISTORICAL FIGURES OF MIXED HERITAGE

Current theories regarding the identity of those entombed in the YSR keyhole-shaped tumuli have been restricted to limited identity labels found in historical texts, such as Wa, Paekche, or Mahan. I argue that some of the entombed may not have fallen so easily into either textually-derived category of identity and could be a mixed category or something different altogether, which may explain the eclectic nature of the burials. Evidence for people of mixed heritage between "Japanese" people and Han 韓, a general term for people from the southern Korean peninsula,⁵² can be found in the Nihon shoki. One example is a memorial sent by the ruler of Imna (J. Mimana) to the Yamato court complaining about Kena no Omi, the Yamato envoy to Imna:

[530] [Kena no Omi] is negligent in administering governmental affairs. There are many complex disputes regarding children between Japanese and Imna people. None have been resolved. Kena no Omi enjoys trials by boiling water saying, "Those who are true will not be scalded. Those who are false will surely be scalded. Due to this, many have been scalded to death by being plunged into the boiling water. Furthermore, he killed Nadari [J. Natari] and Sap'ori [J. Shifuri], the Han children of Kibi. (Those born of Japanese marriage with barbarian women were called "Han children."

懶聽政焉。爰以日本人與任那人頻以兒息、諍訟難決、元無能判。毛野臣、樂置 誓湯曰、實者不爛、虛者必爛。是以、投湯爛死者衆。又、殺吉備韓子那多利・ 斯布利[大日本人娶蕃女所生、爲韓子也、恆惱人民、終無和解]。

Another example found in the Nihon shoki regarding Paekche is Nasol Ki no Omi:

[541] Fall, Seventh Month. Paekche heard that the Japanese authorities of Alla [J. Ara] and Silla were scheming together, so it sent Nasol Piri Makko of the Forward Division, Nasol Sŏnmun, Nasol Mokhyŏp Misun of the Middle Division, and Nasol Ki no Omi Mimasa.

(The Nasol Ki no Omi was probably the son of a Ki no Omi by marriage with a Han 韓 woman, who then remained in Paekche and became a Nasol.)

秋七月、百濟聞安羅日本府與新羅通計、遣前部奈率鼻利莫古・奈率宣文・中部 奈率木刕眯淳・紀臣奈率彌麻沙等、[紀臣奈率者、蓋是紀臣娶韓婦所生、因留百 濟、爲奈率者也。未詳其父。他皆效此也。]⁵³

Nasol Ki no Omi Mimasa is half "Wa" and half "Han" 韓. Mimasa represents a third identity that may have been quite common for this time, having a Yamato name and title, residing in Paekche and also carrying a Paekche title. This could also equally apply to Korean peninsula immigrant groups residing on the Japanese archipelago.

Another example of the fluid exchange between the Japanese archipelago and the Korean peninsula is Nichira:

[583] "The late emperor planned to restore Imna but died before [his plan] came to fruition and could not complete his will. Therefore, I must undertake his divine plan. Talsol Nichira, the son of Arishito, ruler of the province of Ashigita in Hi⁵⁴, is now in Paekche. He is wise and brave."

先考天皇謀復任那、不果而崩、不成其志。是以、朕當奉助神謀復興任那。今在 百濟、火葦北國造阿利斯登子達率日羅、賢而有勇。⁵⁵

Nichira's father is a regional authority in Kyushu, and yet his son decided to pursue a political career in Paekche. Even as late as 583, this shows that there was considerable mobility between Kyushu and Paekche.

It makes it even more difficult to classify someone from the Yŏngsan River basin as "Wa," "Paekche," "Mahan," etc., and trying to identify them as such is most likely a foolhardy endeavor. It is entirely possible that the inhabitants of the Yŏngsan River basin were themselves an eclectic mix of many different groups stemming from the jar burial culture that had existed there several centuries prior. It is a difficult question to answer archaeologically and an even harder one to answer through the existing historical texts. But it is unlikely that those buried in the YSR keyhole-shaped tumuli came directly from Paekche or Yamato because the construction methodology is identical to non-keyhole-shaped tumuli in the same region.⁵⁶

PAEKCHE'S MOVE TO SABI (538) AND THE YŎNGSAN RIVER BASIN

In 538, Paekche King Sŏng (r. 523–554) moved the Paekche capital approximately 30 km southwest from Ungjin to Sabi (present-day Puyŏ) and changed the name of the state to Nam Puyŏ 南夫餘.⁵⁷ Several factors precipitated this move, but the primary concern seemed to be breaking free from the militarily defensible but restrictive geography of Ungjin, which was surrounded by mountains and difficult to access, to Sabi, which was on an open plain and closer to the oceanic transportation routes to China and the Japanese archipelago.⁵⁸ The move to Ungjin from the previous capital Hansŏng was not by choice and hastily prepared, with the primary goal at the time to reconstitute the government as quickly as possible in a militarily defensible position. However, as the population grew, the militarily advantageous rugged terrain became a limiting factor in Ungjin's growth and its ability to foster trade and diplomacy outside the Ungjin region. The move to Sabi, on the other hand, was carefully planned and reflected Paekche's new confidence in its political and military strength.⁵⁹

King Sŏng's move of the capital also bookmarks the end of the construction period of the YSR keyhole-shaped tumuli. Jar burials are also all replaced with Paekche-style tombs, indicating that the local elites have been fully integrated into the Paekche political structure. It may seem coincidental that the beginning and end of the YSR keyhole-shaped tumuli are timed to changes in Paekche's capital, but I think it is no accident. The geopolitical and economic changes that accompanied each move most likely had an impact on the internal changes already taking place within the Yŏngsan River basin.

CONCLUSION

After examining the available textual data concerning Paekche and Yamato in the late fifth to early sixth century, there are several events and figures that may provide clues regarding the relationship between Paekche and the Yŏngsan River basin. First, Paekche's move to Ungjin and loss of all its territory in the Han River basin in 475 as well as the persistent threat from Koguryŏ in the north would have encouraged Paekche to focus on the south for allies and resources. According to the Samguk sagi, Paekche at the time was too weak in the late fifth century to have absorbed the polities in the Yŏngsan River and most likely formed relationships with the local authorities in the Yŏngsan River basin instead, which is supported by the archaeology. Second, local authorities in the Yŏngsan River basin (as well as Kyushu) started building ties with both Paekche and Yamato and neighboring local elites. This can explain the sudden explosion of eclectic burial goods and the incorporation of tomb structural innovations that started in the late fifth to early sixth century, which includes the YSR keyhole-shaped tumuli.

The Nihon shoki also shows that there was considerable mobility of people between Paekche and the Japanese archipelago. Groups from the Japanese archipelago could easily migrate into the southern Korean peninsula and integrate with the local authorities. Local authorities in the Kaya region, Kyushu, and presumably the Yŏngsan River basin could simultaneously hold rank and title in the Paekche and Yamato courts. In other words, there was considerable flexibility of identity across the Paekche-Yŏngsan River Basin-Kyushu-Yamato trade corridor. This also meant that local authorities in the Yŏngsan River basin had access to groups that specialized in the construction of keyhole-shaped tumuli via their connections in Kyushu. The Chinese dynastic histories also confirm that there was considerable ethnic diversity in the Paekche countryside, which further supports the idea of a fluid multi-ethnic/multicultural world at this time.

The historical sources, however, provoke more questions than answers. The identity of those buried in the YSR keyhole-shaped tumuli is still difficult to answer, and the process that led up to the adoption of the keyhole-shaped tumuli in certain areas of the Yŏngsan River basin and the relationship between the local authorities, Paekche, and polities on the Japanese archipelago that might have influenced it is not entirely clear. The absence of historical sources regarding the Yŏngsan River basin limits their usefulness regarding these questions, which forces us to look at archaeological approaches. However, even the archaeology has its limitations. But based on the limited data that we have, it is seem most likely that the YSR keyhole-shaped tumuli were constructed for local elites who had strong ties with both Paekche, Yamato, and Kyushu. And the decision to use a keyhole-shaped tumulus with Paekche prestige goods, Kyushu horizontal corridor stone chambers, and local ritual pottery and construction methods is reflective of a very multicultural period in the Yŏngsan River basin in the late fifth and early sixth centuries.