Battle of Jiksan

The Battle of Jiksan was fought during the second Japanese invasion of Korea, which began in 1597 after peace talks between Ming China and Japan had failed. It took place on the 7th day of the 9th lunar month of 1597, according to a Korean liaison officer’s report and Japanese accounts. The date is usually given as 16 October 1597; converted with the table of the Korea Astronomy and Space Science Institute (KASI), the same lunar day is 17 October. Korean historians also call it the Battle of Sosa, after the Sosa plain north of Jiksan.

After taking Namwon and Jeonju in the south-west, part of the Japanese Right Army marched north through Chungcheong Province toward Hanseong, the Joseon capital. The Ming superintendent Yang Hao, who had just arrived in Hanseong, sent picked cavalry south to meet them. Near Jiksan, a small town north of Cheonan, the Ming horsemen ran into the vanguard of Kuroda Nagamasa. Reinforcements joined on both sides, but the accounts differ about how the fighting ended.

No Korean unit is recorded as fighting in the battle in the sources used here; it was a clash between Ming and Japanese troops. Ming and Korean records called it a victory that saved the capital. Japanese accounts say the Ming were driven off. A few days later Japanese raiders reached Anseong and the edge of Juksan, closer to Hanseong than Jiksan, but they did not attack the capital, and soon the Japanese armies turned back toward the southern coast.

Why the Japanese turned back is still debated. The sources offer several explanations, including Ming resistance at Jiksan, Hideyoshi’s plan for the campaign, the coming winter, supply difficulties and diplomacy, but they do not show how much each mattered. The later sections of this page look at what each source says and where they disagree.

The second invasion had a limited aim. According to Japanese Wikipedia, which quotes Hideyoshi’s order of the 21st day of the 2nd month of 1597, the Japanese armies were to subjugate the “Red Country”, Jeolla Province, completely, and to move into the “Blue Country”, Chungcheong Province, and elsewhere as far as they could; after that they were to build castles, assign garrisons and send the remaining troops home.

In the late summer of 1597 two Japanese armies advanced inland. The Left Army under Ukita Hideie took Namwon on the 16th day of the 8th lunar month (26 September); the Right Army under Mōri Hidemoto, with Katō Kiyomasa, Kuroda Nagamasa and others, took the mountain fortress of Hwangseoksan. Jeonju, abandoned by the Ming commander Chen Yuzhong, fell without a fight on the 19th (29 September). From Jeonju the Right Army marched north to Gongju. The Ming History says that Ma Gui had sent the youji (mobile corps commander) Niu Boying to join Chen at Gongju; Japanese accounts add that Niu fell back to Hanseong when the Japanese advanced.

At Gongju the Right Army divided. The Encyclopedia of Korean Culture says that Mōri’s and Katō’s troops went on to Jeonui and Jincheon, and that part of them, Kuroda’s force, reached Jiksan. Japanese accounts summarised by Japanese Wikipedia say that Katō went to Cheongju, while Mōri, with Kuroda as his vanguard, reached Cheonan on the 6th day of the 9th month.

The Ming side was weak and divided. Ma Gui, newly made commander (tidu) of the Ming army in Korea, told King Seonjo on the 27th day of the 8th month that the Ming troops at Cheonan were outnumbered and exhausted and that he would send Xie Sheng and Yang Dengshan to relieve them. The Ming History says that Ma Gui even proposed abandoning Hanseong and falling back to the Yalu River, until the official Xiao Yinggong hurried from Pyongyang to stop him. The civil official Yang Hao, in overall charge of the Ming army as superintendent (jingli), arrived in Hanseong on the 3rd day of the 9th month. According to the Revised Annals of King Seonjo, he rebuked Ma Gui for not fighting and, with him, secretly chose picked horsemen and sent them south under Xie Sheng, Niu Boying, Yang Dengshan and Po Gui; neither the other Ming units nor the Koreans were told. Bai Sai, another youji, followed with 2,000 cavalry.

The first news reached Yang Hao’s headquarters on the 8th day of the 9th month. According to the Korean officials who passed it on, the Ming troops had lain in ambush in narrow, rugged ground about 10 li south of Jiksan and had met a Japanese vanguard of unknown command; they dismounted and fought, then both sides drew back for a time. A Ming soldier reported that 30 heads had been taken, that countless Japanese had been shot, and that some of the dead wore gold helmets and armour, a sign that they were commanders.

On the 9th two fuller reports arrived. Jang Un-ik, the Korean official attached to Ma Gui, repeated what a Ming soldier back from the battlefield had told him: between Cheonan and Jiksan, the Japanese vanguard, all dressed in white, came over the fields; the Ming at first took them for Koreans and did not advance, but when the Japanese fired first, the Ming cavalry charged. After a long fight, nearly 500 or 600 Japanese had been killed by arrows or clubs and more than 30 heads taken. Then the Japanese climbed a hill and raised a white flag, the main army from Cheonan quickly gathered “like clouds”, and, outnumbered, each side withdrew to hold its ground. Many Ming soldiers had also been killed, and Xie Sheng and the other three generals had left Jiksan in the night.

The liaison officer Sin Chung-il reported that on the 7th the Japanese vanguard had come up from Cheonan, and that the four generals, with 2,000 picked men and 15 officers, had met them about 15 li from Jiksan and taken 31 heads; Xie Sheng took two himself, Yang Dengshan one and Po Gui three. On their way back the Ming troops passed Jinwi, reached Suwon in the early evening, ate, and rode on at the second watch of the night.

Japanese accounts, as summarised by Japanese Wikipedia, tell the story from the other side. Before dawn on the 7th Kuroda Nagamasa sent his vanguard under Kuroda Naoyuki and Kuriyama Toshiyasu toward Jiksan; at sunrise they saw Ming troops filling the hills and decided that attacking was safer than retreating. Nagamasa hurried up with 3,000 men, and the Ming were reinforced by Bai Sai. The fighting was undecided until Mōri Hidemoto, at Cheonan, sent his vanguard under Shishido Mototsugu and Yoshimi Hiroyuki against the Ming flank and rear; the Ming then fell back toward Suwon, and after about 3 p.m. the Japanese returned to Cheonan without pursuing. English Wikipedia, citing Kenneth Swope, adds that Kuroda and Mōri reported that their muskets could not pierce the iron shields of the Ming soldiers.

In Hanseong the news of the Japanese advance caused panic. Yu Seong-ryong wrote in Jingbirok that when the enemy reached Jiksan, the people of the capital all fled, and that on the 9th day of the 9th month the queen left the city for the west. The Revised Annals add that the Ming commanders asked King Seonjo to come out to the Han River, that he went unwillingly while people stood ready with their bundles, and that the city calmed down only when news of the victory arrived. The crown prince had already left with the ancestral tablets of the royal shrines, and the king told a Ming officer that he had sent his family out of the city because the danger was so great.

The Ming officers were divided too. On the 12th, according to the Veritable Records, Yang Hao rode out to the Dongjak ferry on the Han River. Ma Gui followed him out of the South Gate, warning that if the Japanese came round by another road the Ming would be cut off, but Yang answered, “I have my own plan”, and the king followed them. Across the river Yang and Ma Gui had about a thousand Ming horsemen manoeuvre on the hills. After crossing back, Yang held an archery contest between Ming and Korean archers and had cannon fired beside the river. The record explains that many Ming officers, hearing how close the enemy was, had wanted to withdraw on the grounds that their forces were few and weak, and that Yang wanted to show that there would be no retreat.

Did Korean troops fight at Jiksan? The sources used here do not describe any Korean unit in the battle. The Revised Annals stress that the Koreans were not even told that the cavalry had been sent, and the reports in the Veritable Records reached the court through Korean liaison officers who repeated what Ming soldiers told them. Korean forces were in poor shape: on the 13th the commander-in-chief Gwon Yul told the king that the Chungcheong army commander Yi Si-eon had taken over 2,000 men into Gyeongsang Province, but that by the time he came back toward Jeolla most had deserted and only about 50 remained. The Ming History says that the Korean court sent the high commissioner Yi Won-ik through the Joryeong pass into Chungcheong to block the enemy, but that is not a report of fighting at Jiksan. For this reason the battle is best described as a Ming–Japanese clash rather than a battle of Ming–Korean allied forces.

Jiksan was not the furthest point the Japanese reached. The Veritable Records note, under the 14th day of the 9th month, that the Japanese had plundered Anseong on the 10th and pushed on into the district of Juksan, both closer to Hanseong than Jiksan. At court, officials feared that the Japanese, blocked by the Ming at Jiksan, would go round by Anseong and Juksan and then take the Yongin road straight to the capital; the Office of the Censor-General urged that Gwon Yul be sent back to guard the roads through Yangji and Juksan.

Then, unexpectedly, the Japanese began to pull back. On the night of the 15th King Seonjo called his officials to discuss reports that the enemy was suddenly withdrawing and that the Ming commanders were taking their best troops south in pursuit. The king suspected a trick, a feigned retreat to draw the Ming forward; one official, Kim Su, thought the Japanese would go back to Gyeongsang Province by way of Chungju. On the 20th Korean officials reported Ming claims that pursuing units had taken 155 heads, and that the large Japanese camps at Cheongju and Gongju had all left, some heading into Jeolla Province and some by way of the Joryeong pass, killing everyone they met along the way. On the 22nd the king wrote that the enemy had begun to withdraw from Gyeonggi and Chungcheong, leaving the dead piled up in the villages.

The Encyclopedia of Korean Culture says that Kuroda stayed at Jiksan for several days, plundering the area, then joined Mōri and withdrew by way of Gongju and Sangju; Mōri and Kuroda fell back to Seongju and in the middle of the 10th month to Yangsan. Jingbirok sums up the result: the Japanese went back from the borders of Gyeonggi Province, Katō Kiyomasa again held Ulsan, Konishi Yukinaga held Suncheon and Shimazu Yoshihiro (written 沈安頓吾 in Jingbirok) held Sacheon, a line of fortresses stretching some 700 or 800 li along the southern coast.

Korean accounts have long linked the Japanese withdrawal to Jiksan. The Encyclopedia of Korean Culture’s article on the war says that the clash at Sosapyeong completely stopped the Japanese advance north, and that Yi Sun-sin’s victory at Myeongnyang blocked their advance west by sea. It dates the two battles to the 5th and the 16th of the month, eleven days apart, although its text calls the interval ten days. The sources offer several explanations for the Japanese withdrawal, but the evidence used here does not establish which factors actually contributed or how much each mattered.

First, the campaign plan. Hideyoshi’s order of early 1597, quoted in Japanese Wikipedia, specified subjugating Jeolla and campaigning in Chungcheong and elsewhere as far as possible, followed by castle-building and garrisons; it did not name Hanseong as a target, though it did not expressly exclude it either. An instruction not to attack the capital appears only in a prisoner’s testimony, below. Japanese Wikipedia argues that by reaching Anseong and Juksan the Right Army had achieved its objective and turned back as planned. Fukuda Kansuke (福田勘介), a Japanese prisoner captured by the Chungcheong army commander Yi Si-eon, is recorded in the Veritable Records entry for the 3rd day of the 10th month as saying that Hideyoshi had ordered his commanders not to attack the capital, to kill everyone in their way until the 9th month and to return to Seosaengpo or Busan within the 10th month, and that this was why the army turned back when it was only three days’ march from the capital. Such testimony, given by a captive, cannot be taken at face value, but it is broadly in line with the order quoted in Japanese sources.

Second, Ming resistance. Samuel Hawley writes that once it was clear that the Chinese had returned to Korea, the Japanese made no effort to hold the ground they had covered, which would have left them open to counter-attack with winter coming on. Another prisoner, a servant of the Mōri, said that more than 500 Japanese had died at Jiksan when the Ming blocked their advance on the capital, and that Katō Kiyomasa sent orders to withdraw. Third, season and supply: English Wikipedia cites a Japanese study that explains the halt by the approaching winter and food shortages, and the Ming History notes that Korea had been laid waste, so the Japanese found little to seize as they marched.

The Ming History gives yet another explanation: that the Japanese withdrew after a Ming envoy brought them a letter from the captured negotiator Shen Weijing, and that Konishi Yukinaga blamed Katō for the rash advance. Myeongnyang, by contrast, cannot explain the first turn south. The naval battle was fought on the 16th day of the 9th month (26 October), but the Korean court was already discussing the Japanese withdrawal on the 14th and 15th. Myeongnyang kept the Korean navy in being and dealt a blow to the Japanese fleet’s advance west, but the withdrawal on land had begun before news of it could have arrived.

Even the Korean records do not agree. The early reports in the Veritable Records describe a sharp fight with heavy losses on both sides, after which the Ming were outnumbered and pulled back to Suwon. The Revised Annals, compiled decades later, present a planned ambush on the Sosa plain in which the Japanese were struck before they could form up, fled with very many dead, and were beaten again in a pursuit with Bai Sai’s 2,000 horsemen. The Encyclopedia of Korean Culture follows this version and adds that the fleeing Japanese threw away their muskets.

The Ming record is divided as well. The Ming History says that Ma Gui reported “great victories at Cheongsan and Jiksan”, and that the official Xiao Yinggong objected that the Japanese had withdrawn because of Shen Weijing’s letter and that no engagement had taken place at either Cheongsan or Jiksan, so no merit could be claimed; the supreme commander (zongdu) Xing Jie and the superintendent Yang Hao, angered, had Xiao impeached and arrested. In its biography of Yang Hao, the same history says that the deputy commander Xie Sheng and others “repeatedly checked the enemy” and that the Korean forces also had several successes, after which the Japanese withdrew to Ulsan; it does not say that those Korean actions took place at Jiksan. Yang himself was dismissed in 1598, after another official accused him of deceiving the throne, chiefly over the failed siege of Ulsan.

Japanese accounts, as summarised in the narrative of Japanese Wikipedia, describe a Japanese success: the Ming were driven back toward Suwon with about 200 dead, while Kuroda lost 29 men. Japanese Wikipedia also cites the Veritable Records’ figure of 500–600 Japanese dead, and in its summary of the result calls the battle undecided. The prisoner Fukuda Kansuke said that Kuroda’s army had lost many men at Jiksan and was hiding it out of shame, though he admitted he did not know the details. English Wikipedia notes that most Western, Korean and Chinese works treat Jiksan as a Ming victory, while Japanese works are mixed, and it describes the result as a withdrawal by both sides. Hawley calls it a clash with “an advance unit of Chinese troops”.

The figures cannot be reconciled. Korean reports count heads taken (30 or 31) and estimate those killed by arrows and clubs (500–600); the Japanese figure of 29 dead counts only Kuroda’s men; and none of the sources used here gives a reliable Ming total. The detailed narratives agree that Ming troops and Kuroda’s force fought near Jiksan, but they disagree about the course and result of the fighting: the early reports in the Veritable Records describe a Ming withdrawal to Suwon, the Japanese accounts describe both sides leaving the field, and the Revised Annals describe a Ming pursuit and a further victory, while Xiao Yinggong denied that any engagement took place at all. What is clear is that Japanese troops then reached Anseong and Juksan without attacking Hanseong, and that by the middle of the month the court was discussing their withdrawal.

The date of the battle is not certain to the day. Sin Chung-il’s report gives the 7th day of the 9th month, as do the Japanese accounts summarised by Japanese Wikipedia; the Encyclopedia of Korean Culture gives the 5th. Converting the 7th into the Western calendar also gives two answers. In the original Veritable Records the cyclical day names of this month correspond to 16 October, the date English Wikipedia gives, citing Swope; in the Revised Annals and the KASI table, which put the first day of the month on 11 October, it is 17 October. The same one-day difference affects other dates of this month, such as the raid on Anseong on the 10th (19 or 20 October). Myeongnyang, on the 16th, is dated 26 October here.

The place is also described in different ways. The first dispatch put the ambush about 10 li south of Jiksan; Jang Un-ik’s informant said “between Cheonan and Jiksan”; Sin Chung-il said about 15 li from Jiksan. The Revised Annals name the Sosa plain of Jiksan, and the Encyclopedia of Korean Culture places the battle at Honggyeongwon (弘慶院) on the Sosa plain, north of Jiksan. The reports may describe different stages of a moving fight, but the sources used here do not settle where the main fighting took place.

Each kind of source has its problems. The Korean court heard about the battle second- or third-hand, from liaison officers repeating Ming soldiers; the court badly needed good news and depended on Ming goodwill, which may have coloured how such reports were received. The Revised Annals were compiled decades later and tell a tidier story. The Ming History was compiled much later still, and its account of the 1597 campaign records the quarrels among Ming officials over who deserved credit. The Japanese battle narrative used here comes through Japanese Wikipedia. It cites Nihon senshi: Chōsen-eki for some figures, but the underlying Japanese works were not examined for this article. The prisoners’ statements may reflect rumour or what their captors wanted to hear: when one of them said that Hideyoshi had ordered the army to embark on the 2nd day of the 10th month and reach Japan within 20 days, Ma Gui called his story absurd.

At the time, the Korean court treated Jiksan as a Ming victory and thanked the Ming officers in person. On the 25th day of the 9th month King Seonjo visited Xie Sheng’s quarters and drank with him. On the 3rd day of the 10th month he called on the youji (mobile corps commander) Po Gui, who, the record notes, had taken many heads at Jiksan, and told him that the enemy had come close to the capital, the army and people had collapsed and the dynasty had almost fallen, but that thanks to his might the enemy’s spearhead had been blunted and they had all withdrawn. Po Gui, 55 years old, answered that in all his campaigns he had never met an enemy as fierce as the Japanese, but that if armoured cavalry charged and trampled them, they could not use their courage.

Not every Ming officer thought the battle worth much. A Ming deputy commander surnamed Li told the king, who mentioned the victory at Jiksan, that it was hardly worth speaking of: only when an army of a hundred thousand had been wiped out, as at Pyongyang, would there be cause for congratulation.

The Encyclopedia of Korean Culture’s article on the war speaks of a great victory (daecheop) at Sosapyeong. The encyclopedia’s article on the Battle of Sosa counts it, with Pyongyang and Haengju, among the three great land victories of the Imjin War, and notes that it was a pitched battle in the open, unlike the sieges and defences of the other two. This list differs from the “three great victories” of Hansan Island, Jinju and Haengju that are traditionally counted in Korea. The Encyclopedia’s article on the war pairs Jiksan with Myeongnyang on the 16th of the same month: one blocked the Japanese advance on land, the other at sea. Japanese Wikipedia, for its part, calls Jiksan the only battle of the war in which the two armies met head-on in the open field; its narrative describes a Japanese success, but its summary of the result calls the battle undecided, with both armies withdrawing at dusk.