Gwanggaeto the Great

Gwanggaeto was the nineteenth ruler of Goguryeo, which Korean history counts among the Three Kingdoms together with Baekje and Silla; in his time Goguryeo ruled lands in northern Korea and Manchuria. His personal name was Damdeok. He is usually said to have been born in 374; he was named crown prince in 386 and, according to the stone monument set up after his death, came to the throne in 391 at the age of eighteen by traditional East Asian age reckoning.

The name Gwanggaeto is not really a name but a description. It is a shortened form of a long title written on his monument, which can be translated roughly as “the king buried at Gukgangsang, who widely opened up the territory and brought peace and security, the good great king”. In his lifetime, because he used the reign era Yeongnak, he was also called the Yeongnak Great King. Koreans today usually call him Gwanggaeto Daewang, “Gwanggaeto the Great”.

His reign lasted only about twenty-one years, but it changed the map of north-east Asia. Goguryeo pushed into the Liaodong region west of the Yalu River, defeated Baekje, placed Silla under its protection and conquered peoples to the north and north-east. His son and successor Jangsu built on these gains and ruled for almost eighty years, until 491.

Almost everything we know about Gwanggaeto’s deeds comes from two kinds of writing: a stone inscription made by his own family two years after his death, and history books written in China and Korea centuries later. Reading these sources side by side, and noticing where they disagree, is part of what makes him interesting to historians.

Gwanggaeto is remembered as a warrior king, and he fought repeated wars against Baekje. When he became king, Goguryeo was still recovering from a disaster: in 371 the Baekje army had killed his grandfather, King Gogugwon, in battle at Pyongyang.

His monument lists campaigns in almost every direction. In the west he fought the Khitan, a people living along the upper Liao River (the Xar Moron), and, as Chinese and Korean histories record, the Later Yan state for control of Liaodong. In the south he attacked Baekje and, in 396, forced its king to surrender and swear loyalty. In 400 he sent a large army of infantry and cavalry, fifty thousand soldiers according to the stele, to help Silla, whose lands had been overrun by warriors from Wa (the islands of Japan) and their allies. In the north-east he sent a small force against the Sushen in 398 and personally campaigned against Eastern Buyeo in 410.

Not every change was about war. Under his uncle, King Sosurim, Goguryeo had officially received Buddhism in 372 and founded the Taehak academy; Chinese writing was already in use well before that. Korean records say Gwanggaeto founded nine Buddhist temples at Pyongyang in 393 and created new central offices, and his family recorded rules for the households that looked after royal tombs, which shows a growing, organised government.

By the end of his reign the stele claims he had captured 64 walled towns and about 1,400 villages. We should read such numbers as the proud boasts of a royal monument, but they show what Goguryeo’s rulers wanted people to believe about their king.

The stele records that Gwanggaeto’s body was moved to the royal tomb on the twenty-ninth day of the ninth lunar month of 414, the gabin year, and that King Jangsu then erected a huge stone pillar nearby in the Tonggou basin by the Yalu River. It is a natural, roughly shaped block of volcanic stone (Korean and Japanese reference works describe it as tuff or tuff breccia), about 6.39 metres high and between about 1.35 and 2 metres wide on each side. Characters were carved into all four faces in 44 vertical lines, originally an estimated 1,775 in total, in a style close to Chinese clerical script.

The text has three parts. The first tells the origin legend of Goguryeo: its founder Chumo was the son of the Heavenly Emperor and the daughter of a river god, and when he reached a river, reeds were bound together and turtles rose to the surface so that he could cross and found the kingdom. (The later Samguk sagi tells the same story with fish and turtles.) The second part lists Gwanggaeto’s campaigns year by year. The third part lists 330 households, drawn from both old Goguryeo people and newly conquered peoples, whose duty was to guard and maintain the royal tombs, and forbids anyone to buy or sell them.

After Goguryeo fell in 668, the region became a borderland. The stele is mentioned in Korean sources of the fifteenth and sixteenth centuries, in a note to Yongbieocheonga (1447) and on a map of 1596, but it was then taken for a monument of the Jurchen Jin dynasty. It drew renewed attention around 1880, when the Qing government lifted its ban on settlement in the region and scholars began studying its inscription. Since 1928 it has been protected by a pavilion; today it stands behind glass in Ji’an, Jilin Province, China, and is part of a UNESCO World Heritage Site.

Goguryeo had grown since the first century BCE from a confederation of mountain peoples along the middle Yalu into a powerful kingdom, but the fourth century brought crises. In 342 the Former Yan, a state founded by the Murong branch of the Xianbei, sacked the Goguryeo capital. In 371 King Geunchogo of Baekje, at the height of Baekje’s power, invaded the north and King Gogugwon was killed defending Pyongyang.

The response under Gogugwon’s son, King Sosurim (r. 371–384), was a programme of reform. In 372 Buddhism was officially received from the Former Qin, a rising power in northern China, and a national academy, the Taehak, was founded to teach the Confucian classics to aristocratic youths. In 373 Goguryeo issued written laws. These measures strengthened the throne and gave the aristocracy a shared education and religion. Gwanggaeto’s father, King Gogugyang, reigned briefly after Sosurim.

Gwanggaeto therefore inherited a state that had rebuilt its institutions but not yet its prestige. Since the early fourth century northern China had been divided among competing, often short-lived states, many founded by non-Han peoples, while the Eastern Jin ruled the south from 317. When the Former Qin collapsed after its defeat at the Fei River in 383, Liaodong and the northern frontier became a contested zone between Goguryeo and the revived Murong state of Later Yan. This fluid situation gave an ambitious king room to expand.

The stele dates events by the reign era Yeongnak, counting from 391. In Yeongnak 5 (395) the king personally led an attack on the Baili, a Khitan group on the Liao River, and returned with livestock and captives. In Yeongnak 6 (396) he led a naval and land campaign against Baekje, which the stele calls by the insulting name Baekjan, “Baekje remnants”. The inscription claims the capture of 58 walled towns and 700 villages; the Baekje king (Asin) surrendered, handed over a thousand captives and a thousand bolts of fine cloth, and swore to be Gwanggaeto’s “slave-retainer” (nogaek) for ever; Gwanggaeto then returned home with Asin’s younger brother and ten high ministers. Many historians think the 58 towns sum up gains made over several years, from 392 to 396.

In Yeongnak 8 (398) a force was sent against the Sushen in the north-east. In Yeongnak 9 (399), the stele says, Baekje broke its oath and made an alliance with Wa, and Silla sent word that Wa troops filled its lands. In Yeongnak 10 (400) Gwanggaeto sent 50,000 infantry and cavalry south; they drove the Wa forces out of Silla and pursued them to a fortress in “Imna-Gara” in the Gaya region at the southern tip of the peninsula, and afterwards, the stele says, the Silla king came to pay tribute in person. Many characters in this passage are damaged. For Yeongnak 14 (404) the stele records that Wa forces entered the old Daebang region (generally placed in southern Hwanghae Province) and were routed; the damaged text mentions ships and royal troops, but the course of the fighting is unclear.

In Yeongnak 17 (407) the stele records a great victory in which some ten thousand suits of armour were captured, though the enemy’s name is lost in a damaged passage; proposals include Baekje (the view of many researchers), Later Yan, the Wa and the Gaya states. In Yeongnak 20 (410) the king marched against Eastern Buyeo, which had stopped paying tribute, and received its submission. Summing up, the stele claims 64 walled towns and 1,400 villages taken during the reign.

The stele says little about the west, but Chinese and Korean histories record a changing relationship with Later Yan. In 396 its ruler Murong Bao granted Gwanggaeto Chinese-style titles, and the two states exchanged envoys; in 400 Later Yan attacked Goguryeo’s western fortresses, Goguryeo struck back across the Liao River in 402 and 404, and it repelled Yan attacks in 405 and 406. By the end of this period Goguryeo controlled Liaodong, a rich agricultural plain that it would hold until the seventh century and defend against the Sui and Tang empires.

The expedition of 400 had consequences far beyond one campaign. Silla, still a small kingdom centred on the Gyeongju basin, came under strong Goguryeo influence in the early fifth century, though it later sought greater autonomy and cooperation with Baekje. Korean records say Silla sent princes as hostages to Goguryeo (in 392 and 412), and Goguryeo objects appear in Silla tombs of this period.

One striking object is a bronze bowl found in 1946 in the Houchong tomb in Gyeongju. Its base carries sixteen characters, 乙卯年國罡上廣開土地好太王壺杅十, naming the eulmyo year and King Gukgangsang Gwanggaeto Hotaewang and calling the vessel a hou; the bowl was made to honour or commemorate the king rather than by him. The year is usually taken as 415, one year after the stele was erected. The final character, 十 (“ten”), has several proposed explanations, from “ten vessels” to an auspicious or filler sign, and none is certain. The bowl ended up in a Silla tomb built in the early sixth century, about a century later: a physical trace of how closely Silla’s elite was tied to Goguryeo.

For the Gaya polities on the southern coast, the campaign of 400 is often seen as a turning point. Many historians connect the decline of Geumgwan Gaya (at modern Gimhae), the leading Gaya state until then, with the Goguryeo pursuit to “Imna-Gara” mentioned on the stele, after which leadership among the Gaya states shifted inland to Dae Gaya at Goryeong. The exact scale of the damage is debated.

The stele’s picture of the Wa as frequent invaders also raises questions about Japan in this period. The islands were in the Kofun period of great keyhole-shaped burial mounds, and their elites depended on iron and technology from the peninsula. Whether the “Wa” of the stele were armies sent by a central ruler in the Kinai region, warriors from western Japan, or mixed forces including peninsular allies is one of the problems discussed in the advanced sections below.

The stele shows how Goguryeo’s rulers saw themselves. The king is called taewang, “great king”, and the founder is presented as the son of Heaven’s ruler. The use of a reign era, Yeongnak, rather than the era names of a Chinese emperor, signals that Goguryeo claimed its own place at the centre of a world order: neighbouring peoples are described as its subjects who came to pay tribute. Historians describe this as Goguryeo’s own “world view” or tianxia, modelled on, but independent of, Chinese imperial ideas.

At the same time, Goguryeo kings accepted titles from Chinese courts when it suited them diplomatically; in 396, for example, the Later Yan ruler named Gwanggaeto “Governor of Pingzhou and King of the Two States of Liaodong and Daifang”. This double stance, using Chinese forms of kingship at home while bargaining with Chinese states abroad, was common among East Asian kingdoms and is important for the later historiographical debates.

The final third of the inscription, about the tomb-keeping households, reads almost like a legal document. It explains that earlier kings had drawn the guardians only from old Goguryeo people; that Gwanggaeto ordered that his own tomb be kept only by the Han (韓) and Ye (穢) people he had brought back from the communities he conquered in the peninsula; and that Jangsu, worried that the newcomers would not know the rules, added 110 households of old Goguryeo people to the 220 new ones. The 330 households were divided into 30 “state households” (gugyeon) and 300 “watch households” (ganyeon). The text also bans the buying and selling of these people. For social historians this is one of the most valuable parts of the stele, showing how Goguryeo organised labour and absorbed conquered communities; some scholars think the 330 households served all the royal tombs of the capital, not his alone.

Where exactly Gwanggaeto was buried is less certain than the stele’s location suggests. Two great stepped stone tombs nearby have been proposed: the Taewangneung (“Tomb of the Great King”), about 300–400 m from the stele, where a brick reading “May the tomb of the Great King be as secure as a mountain and as firm as a peak” was found in 1913 and where a bronze bell inscribed with the year sinmyo and the title Hotaewang was later excavated; and the better-preserved Janggunchong (“General’s Tomb”), about 1.7 km away. Opinion is divided: the Encyclopedia of Korean Culture’s entry on the stele notes that many Korean researchers favour the Janggunchong, while its entry on the Taewangneung notes that many favour that tomb, dated by its finds to the late fourth or early fifth century.

Gwanggaeto’s reign, 391–412, unfolded amid struggles among the states around Goguryeo. To the west, the Murong state of Later Yan contested control of Liaodong and fought Goguryeo in the early fifth century. To the south, Baekje tried to resist Goguryeo’s advance, while Silla sought Goguryeo’s help against attackers from Wa and their allies. These linked conflicts explain why the campaigns recorded on the stele reached in several directions and why control of routes across the peninsula and Liaodong mattered to Goguryeo.

The spread of Buddhism is a clearer connection between Goguryeo and the wider region. Goguryeo had officially received Buddhism from Former Qin in 372, before Gwanggaeto took the throne. Buddhist teachings and Chinese writing gave its court ways to record royal authority and maintain ties with neighbouring states. Korean records also credit Gwanggaeto with founding temples at Pyongyang in 393. These developments belong to the same regional network of monks, envoys and texts that later carried Buddhism to other kingdoms on the Korean peninsula and to Japan.

In China the same years saw the rise of the Northern Wei, founded in 386 by the Tuoba Xianbei, which would reunify northern China and later become Goguryeo’s main neighbour. The great translator Kumārajīva arrived in Chang’an in 401 and produced the Chinese versions of Buddhist scriptures that Korean and Japanese monks would read for centuries. Buddhism, which had reached Goguryeo in 372, was thus spreading along the same routes that armies and envoys used.

Closer to Goguryeo, Gwanggaeto’s southern campaign changed relations among Silla, Gaya and Wa as well as Baekje. The stele says that his army entered Silla in 400 and pursued Wa forces towards Gaya, but it tells the story from the Goguryeo court’s point of view. Historians compare that account with Korean and Chinese records and with archaeological finds on both sides of the Korea Strait. The evidence shows a connected region, while leaving the size and origin of the Wa forces open to debate.

His son Jangsu (r. 412/413–491) inherited a large and confident kingdom. In 427 he moved the capital south to Pyongyang, closer to the peninsula’s farmland and to the rival kingdoms. In 475 Goguryeo captured Hanseong, the Baekje capital near modern Seoul, and killed the Baekje king, forcing Baekje to move south to Ungjin (Gongju). Silla and Baekje had already drawn together against Goguryeo before the fall of Hanseong; their alliance is often dated to 433, though scholars differ. The balance of power that Gwanggaeto created thus shaped the next two centuries of peninsular politics.

In Korean memory Gwanggaeto became a symbol of a time when a Korean kingdom ruled wide lands in Manchuria. Since the early twentieth century Korean study of his stele has itself often been driven by the wish to prove Korean pre-eminence and victory over the Wa, just as Japanese study stressed a Wa advance into the peninsula. His name is also used in modern commemoration: South Korea’s first domestically built destroyer, Gwanggaeto the Great, was delivered to the navy in 1998. These modern uses tell us more about twentieth- and twenty-first-century concerns than about the fifth century.

For scholars, his lasting legacy is above all the stele: one of the longest surviving texts from the Three Kingdoms period, and almost the only contemporary written sources for the history of Goguryeo, Baekje, Silla, Gaya and Wa around the year 400.

The stele re-entered history around 1880, after the Qing lifted its ban on settlement in the region and set up Huairen county. A farmer reported the monument, which was covered in moss and vines; workers cleared it by burning, which cracked the stone and made parts of the surface flake away. Partial rubbings and “ink-filled” copies, made by tracing the outlines of the characters and inking the spaces around them, began to reach collectors in Beijing around 1882. Around 1883 (some sources say 1884) Sakō Kagenobu (酒匂景信, also read Kageaki), a lieutenant of the Japanese army general staff on an intelligence mission, obtained such an ink-filled copy (not a true rubbing) and sent it to Japan, where the general staff studied it and published a full reading in 1889.

Local rubbing-makers, notably the Chu family (father and son Chu Tianfu and Chu Junde), began around 1902 to apply lime plaster to the uneven surface so that clearer, more saleable rubbings could be made, and they even retouched some characters. This coating altered the visible shapes of some characters, a problem already noted by Japanese researchers such as Sekino Tadashi and Imanishi Ryū who surveyed Ji’an in 1913. In 1972 the Korean-Japanese historian Lee Jin-hui (Ri Shinki) argued that Sakō had altered characters and that the Japanese army general staff had later coated and altered the stone to strengthen passages about the Wa. His thesis caused a sensation and made the stele a symbol of colonial manipulation of history.

Later research did not support the thesis of deliberate falsification by the Japanese army. The Chinese scholar Wang Jianqun, who studied the stone on site from 1981 and published a monograph in 1984, concluded from fieldwork and interviews that the lime had been applied by the Chu family to obtain clearer rubbings, not to distort history, and that the coating was not the work of the Japanese general staff. Since the 1980s a number of rubbings made before the lime was applied have come to light, in collections in Beijing, Taipei, Japan and Korea, and studies of the rubbings, such as Xu Jianxin’s, have refined their dating. Today few researchers in Korea, Japan or China accept the army-falsification thesis. Two things remain true, however: the Japanese army was deeply involved in the early study of the stele, and local rubbing-makers did coat the stone and retouch some characters, so lime-period rubbings must be used with great caution and compared with the early, pre-lime ones.

Philologically, the key problem is that at least about 140 characters are damaged or illegible. Reconstructions depend on comparing early rubbings of different dates, photographs and the stone itself, and on arguments about grammar and the stele’s internal structure. Several passages, not only the famous “sinmyo year” sentence, therefore allow more than one reading.

The sinmyo passage. A sentence of 32 characters placed before the campaign of 396 is often transcribed 百殘新羅舊是屬民由來朝貢而倭以辛卯年來渡海破百殘□□新羅以爲臣民, although 海 and parts of the following words are uncertain. The traditional reading, set out in the 1880s–1890s by Japanese scholars working from the army general staff’s copy, such as Yokoi Tadanao and Naka Michiyo, takes the Wa as the subject throughout: “Baekje and Silla were formerly [Goguryeo’s] subject peoples and paid tribute; but in the sinmyo year (391) the Wa came, crossed the sea and defeated Baekje, [..] and Silla, and made them their subjects.” In pre-war Japan this reading was linked with the legend of Empress Jingū’s conquest of Korea and used as evidence for an ancient Japanese dominion over the southern peninsula, the idea later associated with “Mimana Nihonfu”, and it served colonial ideology.

Korean scholars proposed alternative readings from the 1950s, in response to the Mimana theory. Jeong In-bo (Chŏng In-bo), whose study was published in 1955, argued that the subject changes in mid-sentence: the Wa came in 391, and Goguryeo then crossed the sea and defeated them. North Korean scholars in the 1960s offered further variants, for example that it was Baekje, allied with the Wa, that made Silla its subject, or that the one who “crossed the sea” was Goguryeo. Others have questioned the reading of individual damaged characters, such as 海 (“sea”). These interpretations differ in how they supply omitted subjects, where they divide the sentences and how they read damaged characters, and Korean scholars do not agree on a single interpretation.

Since the 1980s the most common view in Korean, Chinese and Japanese scholarship has been a third one: it accepts the Wa as the grammatical subject of “came and crossed the sea and defeated”, but treats the claim that Baekje and Silla became Wa subjects as exaggeration, written to justify Goguryeo’s wars. Read in context, the sentence is a preamble to the campaigns of 396 and later, and its other claims are also doubtful: only about twenty years earlier, in 371, Baekje had killed a Goguryeo king, so Baekje had hardly “always” been a tributary subject. On this reading the stele is royal propaganda, not a neutral report, and most scholars no longer treat it as proof that the Wa conquered and ruled southern Korea. Some critics object that it is inconsistent to mine the inscription for facts while declaring parts of it fiction. None of this settles how far Wa forces really operated in the peninsula, a question that archaeologists now approach through the distribution of weapons, armour and keyhole-shaped tombs on both sides of the strait.

Whose history is Goguryeo? A second debate concerns the kingdom as a whole. In Korean historiography Goguryeo is one of the Three Kingdoms of Korean history; the later kingdom of Goryeo took its name from it, and Korea’s English name comes from Goryeo. From 2002 to 2007 the Chinese Academy of Social Sciences, with three north-eastern provinces, ran the state-backed “Northeast Project”. Working from the ideas of China as a unified multi-ethnic state and of history belonging to present-day territory, it presented Goguryeo as an ancient regional ethnic regime within Chinese history, pointing also to the fact that most of its territory and early capitals lie in present-day China and that its kings received Chinese titles. Korean scholars and governments objected strongly, arguing that tributary titles did not mean political subordination and that Goguryeo’s own world view, visible on the stele itself, place it within Korean history. In 2004 the two governments reached an oral understanding to keep the issue from becoming political, and South Korea set up the Koguryo Research Foundation (2004), merged in 2006 into the Northeast Asian History Foundation.

In 2004 UNESCO inscribed both China’s “Capital Cities and Tombs of the Ancient Koguryo Kingdom”, which includes the stele, and North Korea’s “Complex of Koguryo Tombs”; the Chinese property’s official description calls Koguryo “a regional power and ethnic group”, while the North Korean one calls it “one of the strongest kingdoms in nowadays northeast China and half of the Korean peninsula”. Goguryeo incorporated people of different origins over its history, and at times Khitan and Mohe groups lived under its rule; but this diversity, and the modern locations of its monuments, should be kept distinct from arguments about historical succession. Readers should notice that textbooks in Korea, China and Japan still frame Gwanggaeto and his stele quite differently.

The stele is a contemporary source, carved within two years of the king’s death, but it is a commemorative text commissioned by his heir. It selects victories, names enemies with contempt and says nothing about defeats; some of its claims, such as Eastern Buyeo having been subject to Goguryeo since the founder’s time, are difficult to accept as fact. It also uses its own dating: counting from the stele, Gwanggaeto acceded in 391 at 18 and died in 412 at 39, ages given by traditional East Asian reckoning.

The Samguk sagi, compiled in 1145 under the Goryeo dynasty, preserves annals of Goguryeo that date his accession to 392 and his death to 413, one year later than the stele, and records events not on the stele, such as a campaign against the Khitan and the capture of Baekje’s Gwanmi fortress in 392, and fighting with Later Yan. Its compilers drew on Chinese histories and lost Korean records; where it conflicts with the stele, most scholars prefer the stele for dates but use the Samguk sagi for events the stele omits.

Chinese dynastic histories, such as the histories of the Jin and Later Yan period compiled later, give the view from the other side of the Liao frontier and mention Goguryeo mainly in the context of war and diplomacy. Japanese chronicles like the Nihon shoki (720) contain traditions about the peninsula in this period, but their dating of events before the late fifth century is widely regarded as pushed back; for some Korea-related entries, especially in the Empress Jingū section, the usual view is that they should be moved 120 years later, though each passage has to be assessed on its own.

Archaeology adds an independent check. Objects such as the 415 bronze bowl from Gyeongju, Goguryeo-style goods in Silla tombs, the stepped stone tombs of Ji’an and the fortresses of Liaodong allow historians to test and sometimes correct the written sources. Much of the relevant ground is now divided among China, North Korea and South Korea, which makes shared fieldwork and open access to finds a continuing challenge.