Battle of Byeokjegwan

The Battle of Byeokjegwan was fought during the Imjin War, the Japanese invasion of Korea that began in 1592. It took place on the 27th day of the 1st lunar month of 1593. Converted with the table of the Korea Astronomy and Space Science Institute (KASI), that day is 27 February 1593. Japanese Wikipedia gives the Japanese date as the 26th day of the 1st month and equates it with the same Western day; the sources used here do not explain the one-day difference.

Byeokjegwan was the guesthouse of the Byeokje relay station on the main road from Kaesong to Hanseong, north-west of the capital; envoys from China spent their last night there before entering the city. The fighting took place mainly at Yeoseongnyeong, a low pass about 3 km south of the station, so Korean historians also call it the Battle of Yeoseongnyeong.

A few weeks earlier, a Ming army under Li Rusong, together with Korean troops, had retaken Pyongyang. Li then marched south to recapture the capital. On the morning of the 27th he rode ahead of his main army with about a thousand horsemen of his own guard to join his advance troops, who had clashed with the Japanese. The Japanese commanders in Hanseong had decided to come out and fight, and their vanguard, under Kobayakawa Takakage, attacked the Ming cavalry from several directions on narrow, muddy ground. Li barely escaped, and the Ming fell back.

The battle itself was short, and its toll is disputed. Its consequences were larger: Li withdrew first to Kaesong and then to Pyongyang, the Ming did not try to take Hanseong by force again, and the war moved toward negotiations. The later sections of this page look at what each source says and why the accounts differ so much.

In January 1593 (the 12th lunar month of 1592) a Ming army of more than 40,000 under Li Rusong crossed the Yalu River. With Korean government troops and warrior monks it stormed Pyongyang on 8 February, and Konishi Yukinaga escaped south across the frozen Daedong River. Kuroda Nagamasa pulled back from Hwanghae Province, and on 19 February the allies retook Kaesong with little fighting. The Japanese commanders in the north fell back on Hanseong. Only Katō Kiyomasa’s army, of about 20,000, was still far away in Hamgyeong Province.

In Hanseong the mood was grim. According to Yu Seong-ryong’s Jingbirok and the Revised Annals, on the 24th day of the 1st month (24 February) the Japanese, angry at their defeat and suspecting the city’s people of secretly helping the Ming, killed large numbers of the Korean men still in the city and burned most of its buildings. They also burned the hills and fields near the capital, which later left the Ming cavalry with little fodder.

The Japanese commanders then had to decide whether to wait behind the walls or march out to meet the Ming. According to the Encyclopedia of Korean Culture, Kobayakawa Takakage, who had commanded at Kaesong, had wanted to hold that city and withdrew only after repeated urging from the overall commander Ukita Hideie; he then refused to enter Hanseong and resolved, with Tachibana Muneshige, to fight the Ming outside it. The encyclopedia puts the Japanese gathered in Hanseong at about 50,000, of whom some 40,000 set out with Kobayakawa as vanguard. Japanese Wikipedia, summarising Japanese accounts, adds that Ishida Mitsunari and Ōtani Yoshitsugu argued for defending the city, and that the commanders finally divided their forces, leaving part in Hanseong.

On the Ming side, the Encyclopedia of Korean Culture says that after a council of war at Kaesong the army set out for Hanseong on the 25th day of the 1st month (25 February), aiming to destroy the Japanese main force and retake the capital. The compiled account in the Veritable Records says that Li crossed the Imjin River by a ford downstream on the 26th and camped at Paju. Yu Seong-ryong, who was in charge of supplies, had grain and fodder brought up, and because the ice on the Imjin was melting, a bridge of ropes made from arrowroot vines was laid across it for the troops. Korean officials, eager for the Ming to press on, played down the Japanese numbers: English Wikipedia says that Yu put the Japanese in Hanseong at only about 10,000, and a few weeks later a Korean interpreter told the Ming commanders that officials had asked him to say the Japanese were few.

Early on the 27th, Ming advance troops under Zha Dashou, with the Korean defence commander Go Eon-baek, went ahead toward Hanseong to scout. Near the Byeokjegwan road they met a Japanese detachment and cut it up, taking heads. Reports disagree about the number: Jingbirok and the Revised Annals give more than 100, Yu Hong more than 400, the compiled account of the Veritable Records more than 600, and Yi Deok-hyeong about 1,000. Encouraged, they pressed on, but larger Japanese forces came up behind, and Zha fell back toward Byeokjegwan.

Li Rusong was riding south to see the road to the capital for himself. When he heard of the clash he hurried forward with his household cavalry, leaving the main army behind. Crossing the Hyeeumnyeong pass, his horse stumbled and threw him; Korean reports say he hurt his face. He then led the horsemen who had arrived into battle near Yeoseongnyeong. According to Jingbirok and the Revised Annals, the Japanese had hidden most of their troops behind the ridge and showed only a few hundred men on it. When the Ming advanced, about ten thousand Japanese suddenly appeared on the hills.

The Encyclopedia of Korean Culture describes Kobayakawa’s army attacking in three groups and surrounding the Ming. The Ming horsemen had no firearms with them except fire arrows (singijeon); their artillery had not arrived, and the road was narrow and deep in mud. Yu Seong-ryong wrote that Li’s northern horsemen carried only short, blunt swords, while the Japanese foot soldiers fought with sharp swords three or four cheok (a traditional unit of length) long, and Korean accounts mention the Japanese guns firing from all sides. The Ming were driven back in hand-to-hand fighting.

Li himself was nearly caught. The Revised Annals say that his officer Li Yousheng (the Joseon records write the name in several ways) and more than 80 of his bravest men were killed, and the Encyclopedia of Korean Culture that Li escaped only because Li Yousheng gave his life. That evening at Paju, Li called for Li Yousheng’s son-in-law, stroked his back and wept, saying, “He was a good man, and he died for me.” Reinforcements under Yang Yuan arrived as the Ming fell back, and according to the Veritable Records the Japanese, who had pursued them as far as the Hyeeumnyeong pass, turned back when they saw the main Ming army and returned to Hanseong.

The sources disagree widely about the losses, and they do not always count the same things: some give the dead, some the dead and wounded together, some the losses on both sides. The figures fall into four groups.

Li Rusong’s own report put Ming losses at 264 killed, 49 wounded and 276 horses lost; English Wikipedia cites this figure from Li Guangtao’s 1972 study published by the Institute of History and Philology of Academia Sinica. The Ming commanders’ statements to the Koreans were similar in tone: on the day after the battle Li told Yu Seong-ryong and other Korean officials that “our army defeated the enemy yesterday and suffered no real defeat.” Yu Hong, the Right State Councillor, who was serving as chief inspector, reported that Li lost “only two company commanders” (pachong), and Yu Seong-ryong at first spoke of Ming casualties of a few dozen.

Korean officials who questioned others gave larger numbers. The envoy Yun Geun-su reported what a Ming staff officer, Xu Yiguan, had told him: about 300 Ming soldiers were lost and about as many Japanese killed. Yi Deok-hyeong, who watched the fighting from the Hyeeumnyeong pass, told King Seonjo that losses on the two sides were about equal and came to “nearly 500 or 600”. The compiled account in the Veritable Records speaks of several hundred Ming dead. A much higher figure came from Wu Weishan, an officer of the Ming southern troops holding the rank of qianhu (千戶), who had been sent as a scout: he said that only some 120 Japanese heads were taken while Ming casualties numbered more than 1,500, including 14 officers. The Korean officials who passed this on warned that the southern troops were at odds with Li and that the report might not be reliable.

Japanese accounts give the largest figures. War chronicles written in the mid-seventeenth century or later, such as the Chōsen seibatsuki, say that 6,000 or 10,000 Ming soldiers were killed while the Japanese lost only a few hundred, and the history of the war compiled by the Imperial Japanese Army general staff estimated over 6,000 Ming dead. Japanese losses are not known with any certainty. Muster lists of March 1593 show Tachibana Muneshige’s contingent much reduced, but Japanese Wikipedia points out that units that fought no battles also shrank greatly, so the drop cannot simply be counted as battle deaths. Several contemporary reports put Ming losses in the hundreds, while another reports more than 1,500 killed and wounded. These figures count different things, and the evidence used here does not establish the total number killed.

The next morning, 28 February, Li prepared to withdraw across the Imjin River to Dongpa. Yu Seong-ryong, Yu Hong, the commander-in-chief Kim Myeong-won, Yi Deok-hyeong and others went to his headquarters and begged him not to retreat, warning that the people who had begun to return home would be abandoned to the Japanese. Li answered that he had only come to reconnoitre, that Paju was short of fodder, muddy after rain and had a river at its back, and that he would rest his men for a few days and then advance again. He showed them his draft report to the Ministry of War, which put the Japanese in Hanseong at almost 200,000; when Yu protested that there were no more than about 10,000, Li replied that the figure came from Korean documents. The Ming withdrew to Dongpa that day and within the next few days to Kaesong, leaving Zha Dashou with a few hundred men to guard the Imjin crossing.

Supply was now the Ming army’s worst problem. Yu wrote in Jingbirok that heavy rain had fallen for days, that the Japanese had burned the hillsides so that there was no grass, and that disease among the horses killed nearly 10,000 within a few days; the compiled account in the Veritable Records gives more than 12,000 dead horses. At Kaesong the grain ran out. When Li’s officers asked to go home, Li had Yu and other Korean officials made to kneel in the courtyard and threatened them with military law; then he turned on his own officers, reminding them that in the Ningxia campaign they had gone days without food and never spoken of going home. Soon afterwards grain boats arrived from Ganghwa. Yi Deok-hyeong told the king that the army had withdrawn because of the shortage of grain and fodder, and the king agreed that this was not Li’s fault alone.

The Koreans kept asking Li to attack again. Li answered that he would advance when the weather cleared and the roads dried. Yu also reported Li telling an interpreter in secret that the first month clashed with his horoscope, which was why he had fallen from his horse and why he could not advance until the next month. According to the Encyclopedia of Korean Culture and Jingbirok, a rumour then spread that Katō Kiyomasa’s army would cross from Hamgyeong by way of Yangdeok and Maengsan and strike Pyongyang. Li declared that Pyongyang was the army’s base and must be saved, left the southern general Wang Bidi at Kaesong, told the Korean commanders to pull back north of the river, and returned to Pyongyang. Yu sent an officer after him with five reasons why he must not withdraw; Li listened in silence and left.

The rumour was false. The Encyclopedia of Korean Culture, in its article on the battles in the northern frontier region, says that Katō spread it himself so that he could withdraw safely from Hamgyeong, and he reached Hanseong around the end of the 2nd lunar month. Jingbirok goes further, saying that Li already wanted to go back north and used the rumour as an excuse. When a Korean plan for a combined attack on the Japanese camps around Hanseong was put to Wang Bidi, he thought it a clever scheme, but, according to the Revised Annals, Li would not agree to it.

The defeat set off arguments within the Ming army. Li’s army was made up of two very different parts: northern cavalry from Liaodong, including Li’s own household troops, and southern infantry from Zhejiang, skilled with firearms, who had led the storming of Pyongyang. Yu Seong-ryong complained to the court that Li used the southern troops to take cities but ranked the northerners first when rewards were given, and Yi Deok-hyeong told the king that the southern general Wu Weizhong deserved most credit for Pyongyang and that Zha Dashou’s rash advance had caused the defeat at Byeokjegwan. The Korean court heard that Li had been impeached, and English Wikipedia notes that one of his officers, Wang Bidi, accused him of starving his men, directing battles from the rear and withholding the rewards he had promised for Pyongyang.

Li in turn blamed the civil official Song Yingchang, who as military commissioner was responsible for the whole campaign and for its supplies. Yi Deok-hyeong told the king that Li complained to the interpreters that Song had not crossed the Yalu to support him, had held back a thousand war horses sent by the court and had obstructed his reports of merit. English Wikipedia’s article on the battle also says that Li shifted the blame to Song.

Korean opinion was divided. King Seonjo criticised Li for rushing to Kaesong after Pyongyang without leaving troops behind and for unfair rewards, but also admitted that the grain shortage was Joseon’s responsibility. The Encyclopedia of Korean Culture notes that the commander-in-chief Kim Myeong-won and others, seeing that Li was underestimating the enemy and advancing recklessly, kept their troops at a distance, so that Korean losses were avoided; the Korean main forces did not take a large part in the battle. Some Koreans, according to English Wikipedia, wanted Li replaced by the southern general Luo Shangzhi.

With the Ming army gone, the Japanese in Hanseong turned on the Korean forces around the capital. On 14 March 1593 they attacked the hill fortress of Haengju, where Gwon Yul, governor of Jeolla Province, beat off repeated assaults. The Revised Annals report that when Li heard of this victory, and of Gwon’s later stand at Paju, he regretted having pulled back so hastily and sent Zhang Shijue back to Kaesong with Yi Deok-hyeong to stockpile grain.

For nearly two months the two sides faced each other between Kaesong and Hanseong, and both suffered from hunger. The Japanese camps ran short of food and were struck by disease, and there are accounts of a Ming raiding party burning Japanese granaries at Yongsan. The Ming army’s grain came from China only as far as Uiju, so Joseon had to gather and carry the rest, a heavy burden on ordinary people.

Both sides now looked to negotiation. The Revised Annals say that while Li stayed at Kaesong he repeatedly sent men to the military commissioner Song Yingchang, apparently to revive the earlier peace talks, and Japanese Wikipedia sees the defeat as the point at which Li gave up on driving the Japanese out by force. The Ming envoy Shen Weijing resumed negotiations with Konishi Yukinaga, and the Japanese agreed to leave Hanseong, release two captive Korean princes and receive Ming envoys. On 18 May 1593 (the 18th day of the 4th lunar month) they left the capital and withdrew to fortresses on the south coast. Joseon, which wanted to fight on, was largely shut out of these talks.

The fullest early evidence comes from the reports that reached King Seonjo’s court in the weeks after the battle and were copied into the Veritable Records of King Seonjo. They include dispatches from Yu Seong-ryong and Yu Hong, who were with Li at Paju; Yi Deok-hyeong’s eyewitness account given to the king in person; reports passed on by the officials who received Ming troops, quoting Ming officers such as Chen Fangzhe and Wu Weishan; and a compiled account of the whole episode filed under the 5th day of the 2nd month. These are close to the events, but most were written by people who wanted the Ming to advance again, and the Ming officers they quoted had their own quarrels.

Yu Seong-ryong’s Jingbirok, written after the war, gives the best-known account of the ambush, the short swords and the horse plague, and the Revised Annals of King Seonjo, compiled decades later, describe the ambush and the Ming cavalry’s difficulties in closely similar terms. The Encyclopedia of Korean Culture lists the Veritable Records, the Revised Annals, Yeollyeosil gisul (a much later compilation) and Yi Hyeong-seok’s Imjin jeollansa (1976) among its references. Details differ between them: the number of Japanese heads taken in the opening clash, the place of that clash (Yeongseo station, the Changneung tombs or Yeoseongnyeong), and the name of the officer who died for Li.

Ming evidence includes Li Rusong’s own report of his losses, which modern Chinese-language scholarship cites (English Wikipedia gives it from Li Guangtao’s study of 1972), and the collected letters of Song Yingchang, the Jinglüe fuguo yaobian, one of the main first-hand Ming sources for the campaign. Japanese evidence is mostly later. The detailed stories of the battle, with Tachibana Muneshige and his men in the leading part and the very high Ming death tolls, come from war chronicles and family histories written from the mid-seventeenth century onward, and from the army general staff’s history of the invasion. They are valuable for the Japanese side of the battle, but their numbers and heroic episodes are hard to check.

In Korea the battle is usually seen as the moment the allied counter-offensive lost its momentum. The Encyclopedia of Korean Culture concludes that because of this defeat the Ming army became less active and missed the chance to destroy the Japanese. Contemporaries said much the same: a few weeks after the battle, Yu Seong-ryong reported that since Byeokjegwan the Ming general’s spirit had shrunk day by day and that he kept promising to advance when the weather cleared.

Japanese accounts present the battle as a great victory over a large Chinese army, with Kobayakawa Takakage and Tachibana Muneshige as its heroes. High estimates of Ming deaths also appear in Western writing: Samuel Hawley, in a 2003 article, described an “epic clash” and gave 10,000 Ming dead “by one estimate”. Modern summaries, such as English and Japanese Wikipedia, set these figures beside the much smaller Ming and Korean ones and treat them with caution.

Historians who work mainly from Ming sources put more weight on the Ming side of the story. Kenneth Swope’s study of the war argues that the Ming intervention on land, above all the recapture of Pyongyang, turned the war, and the low losses in Li’s own report support the view that Byeokjegwan was a sharp but limited defeat for one part of the army. Seen this way, what stopped the Ming included the defeat and the problems it exposed: a cavalry force that had outrun its guns and supplies, dying horses, grain that Joseon could not deliver in time, quarrels between northern and southern troops and between Li and Song Yingchang, and a court in Beijing where the Minister of War, Shi Xing, favoured a negotiated settlement.

What the sources agree on is narrower. On 27 February 1593, near Byeokjegwan, a Japanese army that chose to march out of Hanseong caught Li Rusong’s advance force in difficult ground and drove it back with losses whose size cannot be fixed. Li did not attack Hanseong again, and the allied campaign of early 1593 ended not with an assault on the capital but with the Japanese withdrawal under the peace talks.