First Siege of Jinju (1592)

The first siege of Jinju was a battle for a walled town in south-eastern Korea in late 1592, the first year of the Imjin War, the war that began when Toyotomi Hideyoshi of Japan sent a vast army into Korea. By then the Japanese had taken the capital and Pyongyang, but their hold on the countryside was weak. In the tenth lunar month of 1592, a Japanese army gathered from garrisons around Busan, Gimhae and Changwon marched on Jinju, the leading town of the western half of Gyeongsang Province. Korean records call this half Gyeongsang Right Province.

Inside the walls, Kim Si-min, the newly promoted magistrate of Jinju, commanded a few thousand troops and the townspeople. For about six days, from the fifth to the tenth day of the tenth lunar month (8–13 November 1592), the attackers fired muskets from the hills, built bamboo screens, scaling ladders and a three-storey siege tower, and finally launched a great night assault. The defenders answered with cannon, arrows, gunpowder fire-bundles, stones and boiling water. On the last day the Japanese burned their dead, lifted the siege and withdrew.

The victory had a price. During the final assault Kim Si-min was struck in the forehead by a bullet, and he died of the wound some time later. The siege is useful for understanding the whole war because it shows two things at once: how far the Japanese armies had advanced in a few months, and how hard it was for them to turn that advance into control of the land, its roads and its harvests.

Jinju was more than one town among many. It was a great town (daeeup) of Gyeongsang Right Province, a walled administrative and military centre, and it lay on the overland road leading west into Jeolla Province. The fortress stood beside the Nam River. A moat on the north linked several ponds, and in 1591, on the eve of the war, the governor of Gyeongsang, Kim Su, had enlarged the walls eastwards with an outer wall.

Jeolla mattered because it was rich in rice and, during the first invasion, most of the province, including its capital Jeonju and its main rice-growing districts, stayed out of Japanese control, although there was serious fighting on its north-eastern edge at the Ungchi and Ichi passes. It was also the base of the Korean naval forces still fighting at sea, including Yi Sun-sin’s Jeolla Left fleet and Yi Eok-gi’s Jeolla Right fleet. The Encyclopedia of Korean Culture describes Jinju as the gateway that protected Jeolla, the region supplying military grain, which is why both sides fought so hard for it.

Jinju also mattered as a base. According to the Encyclopedia of Korean Culture, the Japanese knew that the main force of Gyeongsang Right Province was concentrated there. A month after the siege, Joseon’s Border Defence Council noted that “among the large towns of Yeongnam [Gyeongsang], Jinju alone is intact”, and argued that its commander should not be moved elsewhere in case the town became vulnerable again.

Before the siege, the Japanese defeated the Gyeongsang Right army commander, Yu Sung-in, near Changwon. When the Japanese closed in on Jinju, Yu rode to the walls and asked to come in. Kim Si-min, worried that two commanders inside would split the chain of command, kept the gate shut and asked him to support the town from outside. Yu turned back, met the enemy and was killed together with the magistrate of Sacheon, Jeong Deuk-yeol, and the officer Ju Dae-cheong. According to the Revised Annals of King Seonjo, Gwak Jae-u later praised Kim’s decision as the reason the fortress survived.

On the fifth day of the tenth lunar month (8 November) Japanese troops appeared on the Mahyeon ridge east of Jinju, and the attack began the next day. Kim Seong-il, the senior Joseon official in the province, reported that the enemy came in three columns. Kim Si-min ordered old people and women to dress as men so that the garrison would look larger, and told his soldiers not to waste arrows. On the seventh the attackers advanced behind rows of bamboo shields with wooden boards between them, shouting and firing, and attacked again after dark.

The Revised Annals describe the Japanese building thousands of wide bamboo ladders and a three-storey wooden tower from which to look down into the fortress, and piling up screens of bamboo and pine filled with earth. Kim had prepared for this. His men fired a heavy gun, the hyeonja chongtong, at the tower, burned the screens with bundles of gunpowder wrapped in straw, smashed ladders, and poured boiling water and dropped large stones on the climbers.

The decisive night came on the tenth (13 November). The Japanese attacked the east and north sides, and Kim Seong-il’s report says the west gate too. They tried to dig under the wall and set fires whose smoke covered the sky. While directing the defence, Kim Si-min was hit in the forehead. Yi Gwang-ak, the magistrate of nearby Gonyang, took over and shot down a richly dressed mounted commander; at another threatened point the officer Choe Deok-ryang rallied soldiers who had begun to flee. By the afternoon the Japanese had burned their dead and about a thousand houses outside the walls, lifted the siege and marched away.

Jinju was not defended by its garrison alone. Kim Seong-il, who had been sent to the province to rally resistance, sent forces to approach the town from all four directions: from the east the righteous-army leader Yun Tak of Samga, Gwak Jae-u of Uiryeong and Jeong Eon-chung of Chogye; from the north Kim Jun-min of Hapcheon; from the west the Jeolla righteous army of Choe Gyeong-hoe; and from the south Jo Eung-do of Goseong and the ambush commander Jeong Yu-gyeong. The Revised Annals add that Kim Seong-il had asked Jeolla for help and that Choe Gyeong-hoe and Im Gye-yeong came.

Their results were mixed, and the records say so. Yun Tak and Jeong Eon-chung, with about three hundred men, ran into the enemy at Mahyeon and were routed. The report does not place Gwak Jae-u himself at Jinju: it says he sent his vanguard commander, Sim Dae-seung, to the hill north of the town, where his men lit torches, blew horns, fired guns and shouted to the defenders that over ten thousand men from Jeolla and the “red-robed general” (Gwak’s nickname) would arrive the next morning; the people inside shouted back. Kim Jun-min, with some eighty men, drove off a Japanese band that was burning the offices of Dangye county, and Jo Eung-do’s troops took up a position about ten li (a few kilometres) from the Nam River, making their presence felt from a distance and picking off scattered raiders.

The Revised Annals sum up the role of the relief forces carefully. They held strong positions, and at night climbed nearby peaks to light fires and beat drums together with the people in the fortress, but “did not dare to attack outright”. When the Japanese split up to plunder neighbouring districts, the relief forces ambushed them. In other words, their value lay less in a dramatic rescue than in keeping up the defenders’ spirits, threatening the besiegers’ rear and making foraging dangerous.

Gwak Jae-u had been fighting this kind of war for months. He raised one of the first righteous armies at Uiryeong on the twenty-second day of the fourth lunar month (1 June 1592). In the second half of the fifth month (early July) he ambushed a Japanese force crossing the Nam River at Jeongamjin, helping to block a route towards Jeolla, and he raided boats carrying Japanese supplies on the rivers. Cooperation between such volunteers and the government was not automatic: early in the war the governor of Gyeongsang, Kim Su, branded Gwak’s band as rebels. At Jinju, by contrast, government troops inside the walls and volunteers outside fought as parts of one plan.

Kim Si-min was born in 1554 and passed the military examination in 1578. In 1583 he served under Jeong Eon-sin against the uprising of the Jurchen leader Nitanggae on the northern frontier. As a junior officer at the military training office he made proposals about weapons and training that were not adopted, and he resigned. In 1591 he became panggwan, the deputy magistrate, of Jinju.

When the war broke out in 1592, Kim withdrew with the magistrate of Jinju, Yi Gyeong, into the Jiri mountains; when Yi died of illness, Kim took over his duties on the orders of Kim Seong-il. He calmed the people, brought refugees back, repaired the walls and organised the soldiers. Over the summer he fought alongside the Gonyang magistrate Yi Gwang-ak and the righteous-army leaders Yi Dal and Gwak Jae-u, helping to recover Goseong and Changwon, and defeated a Japanese force at Sarangam near Geochang. In the eighth lunar month he was promoted to magistrate. The Encyclopedia of Korean Culture says that he then made about 500 geun of saltpetre and some 70 guns (chongtong) by imitating Japanese methods, and trained picked men to use them.

The court appointed him army commander (byeongsa) of Gyeongsang Right Province, but, in the words of the Revised Annals, he “soon died”. The sources differ on when the appointment was made and received: the Encyclopedia of Korean Culture mentions it after his victory at Jinhae in the ninth month, and also says that the order naming him to the post reached his funeral procession. His date of death is also disputed: the Encyclopedia of Korean Culture says a few days after the battle, while other accounts give later dates. In 1604 he was posthumously enrolled as a second-class Seonmu (military merit) subject, and he later received the posthumous title Chungmu, the same title given to Yi Sun-sin, which sometimes causes confusion between the two.

Jinju makes most sense against the larger problem the Japanese faced in 1592. After taking the capital they divided the provinces among their divisions: the Sixth Division was to conquer Jeolla and the Seventh to secure Gyeongsang, among others. Advancing along main roads was one thing; holding the land around them was another. Every garrison had to be fed, every message and supply column had to pass through hostile country, and from the summer the Korean navy made sending supplies by sea along the west coast impractical.

An entry in the Revised Annals describes how the Japanese tried to hold Gyeongsang. They set up chains of garrisons: ten towns along the left road from Dongnae to Hyeonpung and twelve along the right road from Ungcheon to Mungyeong. No single garrison exceeded about a thousand men, apart from a few coastal bases with several thousand each, and the whole force left in Gyeongsang was put at no more than forty to fifty thousand. To look more numerous, they shifted units back and forth by day and night. This was a network of roads and strongholds, not control of every village.

Such a network was vulnerable to exactly the kind of war that the righteous armies were fighting: ambushes at river crossings, raids on foragers and supply boats, and the recapture of small towns. To attack Jinju, the Japanese had to combine troops from garrisons around Busan and other bases, and when the siege failed they withdrew again. According to the Encyclopedia of Korean Culture, Gwak Jae-u’s defence of Gyeongsang Right allowed farmers there to keep cultivating their fields as in peacetime.

It would be wrong to say that Jinju alone decided the war. Jeolla was also shielded by the navy and by battles such as Jeongamjin and the defence of the Ichi pass in the summer of 1592, and in early 1593 the main Ming army arrived. When the Japanese left the capital in the fourth lunar month of 1593, the Encyclopedia of Korean Culture lists several reasons together: righteous-army uprisings everywhere, the arrival of Ming troops, supply difficulties and epidemic disease. Jinju was one link in that chain: the place where the Japanese attempt to break through to the south-west by land failed in 1592.

An event of 1592 in north-western China had a direct effect on the war in Korea. In March 1592 the garrison of Ningxia, a Ming frontier region facing the Mongols, mutinied under the officer Liu Dongyang and the Mongol-born commander Pubei. The Ming besieged the city from May, built a dam around it and flooded it, and took it on 20 October 1592. The general Li Rusong, who commanded in the final phase, was then given command of the expedition to Korea, and many troops from Ningxia went with him. Here the link is documented, not a coincidence: while Ningxia held out, the Ming could spare only a small force for Korea, and it was defeated near Pyongyang in late August.

This means that when Jinju was besieged in November 1592, no Ming army was there to help; Li Rusong’s army crossed into Korea only around the turn of 1592–93. The defence of Jinju was fought by Korean government troops, volunteers and townspeople alone.

Elsewhere, events of the same year had no connection with Korea and are mentioned only to set the scene. On 9 November 1592, the day after the Japanese appeared before Jinju, Pope Clement VIII issued the bull Cum sacrorum with a revised Latin Bible, the Clementine Vulgate, which remained the Catholic Church’s official Latin text until 1979. In the same year the young mathematician Galileo Galilei moved to the University of Padua, where he would teach until 1610.

Jinju’s victory did not last. When Kim Si-min’s condition worsened, Kim Seong-il appointed in his place Seo Ye-won, the magistrate of Gimhae, who had abandoned his own town and followed Kim Seong-il. The Revised Annals, compiled decades later, describe Seo in harsh terms and say that from then on Jinju’s defences were never as good as before; this judgement should be read with the same caution as other praise and blame in that source.

In 1593, while Ming and Japanese envoys were negotiating, Hideyoshi ordered Jinju to be taken, and the Encyclopedia of Korean Culture calls the attack retaliatory. From the twenty-first to the twenty-ninth day of the sixth lunar month (19–27 July 1593) a far larger Japanese army besieged the town. Accounts commonly give 90,000 to 100,000 men for the operation (90,000 in English Wikipedia; nearly 100,000 in the Encyclopedia of Korean Culture’s entry on Kim Cheon-il), but estimates differ in what they count. The Encyclopedia’s entry on Jinju fortress says about 40,000 attacked, and Japanese Wikipedia, citing a Shimazu-family list, counts about 42,500 men in the units surrounding the fortress, with other units assigned to building forts, transport and garrison duty. The fortress fell. Hwang Jin was killed by a bullet on the twenty-eighth, the day before the fall, and when the walls were breached Kim Cheon-il and Choe Gyeong-hoe threw themselves into the Nam River, according to the Encyclopedia of Korean Culture. Many other commanders also died, and Korean records report the deaths of tens of thousands of soldiers and civilians, although figures vary widely.

Even this victory did not open Jeolla. According to the Japanese Wikipedia account, Japanese troops pushed on to Gurye and Gokseong but turned back when they found Namwon strongly defended. The second siege, with its own heroes and the later story of Nongae, deserves separate treatment; for the first siege it is a reminder that holding Jinju in 1592 bought time rather than permanent safety.

In 1619, at the wish of Jinju’s people, a stele recording Kim Si-min’s achievements was set up inside the fortress, with a text by Seong Yeo-sin. It is now a tangible heritage of South Gyeongsang Province. Its inscription says that fewer than a thousand men drove off an army of a hundred thousand, while other records speak of about 3,800 defenders and 20,000 attackers. The gap shows how memory can enlarge a victory within a single generation.

The royal certificate issued in 1604 to enrol Kim Si-min as a merit subject left Korea during the Japanese colonial period. In 2006 it was bought back with money raised by the public, designated a Treasure, and placed in the Jinju National Museum, a museum devoted to the history of the Imjin War. The Encyclopedia of Korean Culture describes it as the first cultural property to return from Japan through a public fundraising campaign.

The siege left a trace in Japan too. Kim Si-min’s title, moksa (magistrate), was heard by the Japanese as “Mokuso”, and during the Edo period a character called Mokuso hōgan, loosely inspired by him, appeared in kabuki plays. In Korea the battle is counted, together with Hansan Island and Haengju, among the “three great victories” of the Imjin War, and Jinju fortress, restored in the 1970s, is today a historic site with shrines to the dead of both sieges.

The closest thing to an eyewitness report is Kim Seong-il’s dispatch to the court, entered in the Annals of King Seonjo under the fifth day of the twelfth lunar month of 1592. It gives day-by-day details, names the commanders inside and outside the walls, and says the enemy came with “over ten thousand troops, divided into three” (兵萬餘, 分爲三起). A second report, entered in the first lunar month of 1593, recommends the officers for reward. These are official reports by the man who organised the relief: precise and early, but written to establish merit.

The Revised Annals of King Seonjo (Seonjo sujeong sillok) tell the story more fully and more literarily. Begun in the 1640s and completed in 1657 because the original annals were thought to be biased, they say “several tens of thousands” of Japanese attacked a garrison of “over three thousand”, that the town was besieged for ten-odd days with four or five major battles, and they even date the final night assault to the “tenth day of the ninth month”, which does not match the other records. Modern reference works give about 3,800 defenders (English Wikipedia, citing Samuel Hawley, adds more than 2,200 irregulars) against 20,000 (the Encyclopedia of Korean Culture’s article on Kim Si-min) or about 30,000 (its article on Jinju city and the English and Japanese Wikipedia articles), the stele of 1619 claims under a thousand against a hundred thousand, and English Wikipedia’s count of more than 10,300 Japanese dead has no clear basis in the Korean reports, which say only that the dead were beyond counting and that the Japanese carried away or burned their bodies.

Reports about weapons and gunners also differ, though they measure different things and cannot be compared directly. English Wikipedia’s articles on the siege and on the Imjin War say Kim had recently acquired about 170 arquebuses; the Encyclopedia of Korean Culture speaks of some 70 locally made guns, and a Border Defence Council memorial a month after the siege reports, as something it had heard, that he then had about eighty gunners. The image of Gwak Jae-u arriving in person with thousands of guerrillas, found in some English accounts, does not match Kim Seong-il’s report, which mentions only a detachment that Gwak sent under Sim Dae-seung.

Japanese sources on this defeat are thinner than Korean ones. English Wikipedia quotes the Taikōki, a biography of Hideyoshi published in 1626, on ladders nearly breaking under crowds of men racing to be first on the wall, but it is a later, literary work. The dates themselves need conversion: Joseon and Japan both used lunisolar calendars, and the Gregorian dates used here (8–13 November 1592) were converted with the Korea Astronomy and Space Science Institute table, while Japanese accounts count the siege from the fourth day of the tenth month.

How decisive was the siege? Korean accounts commonly say that the victory stopped the Japanese from occupying the Jeolla region and upset their overall plan, and English Wikipedia’s main article states that it largely saved Jeolla from devastation during the first invasion. Other evidence suggests caution. Jeolla had already been shielded in the summer of 1592 by the navy, by Gwak Jae-u at Jeongamjin and by the defence of the Ichi pass, and the Japanese motives in the autumn seem to have been mixed: opening a route west, but also destroying the main base of Gyeongsang Right’s resistance. The second siege shows that taking Jinju did not by itself open Jeolla. A fair conclusion is that Jinju was one important barrier among several.

Whose victory was it? Royal honours and the 1619 stele centre on Kim Si-min. Contemporary reports also highlight Yi Gwang-ak, who took command after Kim was wounded, Seong Su-gyeong, who held the east gate for five days, and Choe Deok-ryang and Yi Nul, who stopped a breach. Kim Seong-il’s second report even records that on the eighth Kim Si-min proposed secretly opening the water gate to let the old and weak escape, and that Yi Gwang-ak objected that this would break the soldiers’ resolve: a reminder that the defenders were not certain of success. Kim Seong-il himself organised the relief, and the righteous armies and townspeople, including the women who dressed as soldiers, were essential participants.

The siege also sits inside wider debates. Korean narratives of the war have long given the righteous armies and popular resistance a central place, and the role of volunteers at Jinju fits that story well. Kenneth Swope has argued that the Ming intervention was decisive in the land war as a whole; Jinju in 1592 was fought before Li Rusong’s main Ming expedition arrived (an earlier, smaller Ming force had been defeated at Pyongyang), and no Ming troops took part, so it is evidence of what Joseon forces could achieve on their own in a well-prepared fortress, but not a refutation of the importance of the Ming. JaHyun Kim Haboush argued that the righteous armies and their writings helped create a proto-national Korean identity, and the cooperation of volunteers from Gyeongsang and Jeolla at Jinju can be read alongside her argument, although this connection is this article’s interpretation rather than hers.

Finally, Jinju illustrates a question central to the whole war: why did an army that could take the capital in three weeks fail to conquer the country? No single battle answers it. But the sources on Jinju show the mechanisms clearly: thinly spread Japanese garrisons, the need to concentrate forces for each new offensive, local forces able to raid and harass, a fortified town that would not fall to assault, and a grain-rich province that remained out of reach.