National Archives
National Archives

Americans learned a hard lesson when North Korean prisoners took over their compound-and kidnapped a general.

The Korean prisoners of war stood in sullen ranks, disciplined, belligerent, ready for battle even though their only weapons were homemade spears, clubs, and incendiary grenades. Their enemy-also disciplined and far better armed, with bayoneted rifles, tear gas, and tanks-stood ready to assault the POWs and recapture Compound 76 of Camp One, Koje-do, a hilly 150-square-mile island 20 miles off the southeastern coast of Korea. In May 1952, the Korean War continued hundreds of miles to the north, but on Koje-do prisoners were waging war as tenaciously as on Sniper Ridge or Porkchop Hill-and here the Communists were winning. N Modern Western ideas about POWs had developed during the American Civil War. The Hague Conventions of 1899 and 1907 transformed these into international law, further refined after World War I in the Geneva “POW Convention” of 1929. That prisoners of war could be a strategic asset was a legacy of Hitler’s Germany and Stalin’s Soviet Union, and the Geneva Convention of 1949 defined the ultimate responsibility of a detaining power to return POWs to the nation that put them in uniform. Conferees adopted these revisions because the Soviet Union was holding German and Japanese POWs as slave laborers, reparations for the damage inflicted on Russia in World War II. Even though tens of thousands of non-Germans-largely Soviet citizens-had served in the Wehrmacht and resisted repatriation in 1945 and 1946, the 1949 Convention revisions were largely silent on the right of POWs to refuse repatriation and on the detaining power’s right to forcibly repatriate unwilling prisoners.

The Geneva Convention of 1949 assumed that prisoners would want to be liberated or exchanged and did not anticipate that the POWs might actually see themselves as unarmed combatants. Although the convention addressed attempts to escape or to attack other prisoners, it never foresaw prison camp violence on a mass scale directed against camp authorities. It was even more unthinkable that POWs would delay their own repatriation with such attacks, or that POWs refusing repatriation would resort to violent resistance. But even as an armistice loomed in Korea in 1952, prisoners in a U.S. Army-run POW camp were scheming to seize the American who ran the camp, Brig. Gen. Francis T. Dodd, and then extort from him a confession that prisoners were abused on his watch. Indeed, the senior officers of the United Nations Command in Korea were about to get a startling education in a POW war behind the wire.

Brutality characterized the Korean conflict for years before the North Korean invasion of June 25, 1950. During the postwar American occupation of Korea (1945-1948), American troops, the Korean National Police, and the Korean Constabulary (the forerunner of the South Korean Army) crushed a major Communist-directed rebellion in October and November of 1946. In March and April of 1948, the South Korean Labor (Communist) Party began a continuous insurgency to prevent United Nations-sponsored elections that would establish the Republic of Korea.

Although the Communists could not stop South Korea from gaining independence on August 15, 1948, the withdrawal of all but 5,000 U.S. Army troops accelerated the partisan war. Trained and organized as guerrillas, Communist Koreans could field up to around 10,000 fighters in 1948 and 1949, supported by probably five times as many South Korean Communist Party sympathizers. Korean security forces, assisted by American weapons and more than 500 advisers of the U.S. Army Military Advisory Group Korea, finally suppressed the insurgency in April and May of 1950.

All the belligerents committed atrocities. The ardently anti-Communist Korean National Police and guerrilla bands led by dedicated South Korean Communist Party members were the worst offenders.

The South Korean government acknowledges the deaths of 7,235 security forces members, with all other deaths in this period estimated at 15,000 to 30,000. South Korean President Syngman Rhee’s critics put the “innocent” deaths at no less than 30,000 and perhaps as high as 100,000. When the Rhee government declared the insurgency crushed in May 1950, there were five to six thousand insurgents and suspected sympathizers in South Korean jails, but more than a thousand remained in hiding, ready to help the impending North Korean invasion. These South Korean Communist Party partisans would play a central role in the fate of the Communist POWs.

In June 1950, as nine divisions of the Korean People’s Army rolled south across the Han River valley toward Pusan, the North Koreans dragooned many South Korean enlisted men into their army. The Communists shot or imprisoned South Korean leaders and “class enemies.” During the capture of Seoul, North Korean soldiers shot the wounded in two hospitals.

Meanwhile, as they withdrew south toward the Taegu-Pusan enclave, South Korea’s national police jailors and South Korean MPs executed their Communist prisoners in Seoul, Wonju, and Kwangju rather than take them south or risk their escape. Only the intervention of an American colonel prevented a mass execution in Pusan. Alan Winnington and Wilfred Burchett, Western journalists sympathetic to the Communists, saw a mass grave near Taejo?n with 1,000 to 1,500 victims. Helpless U.S. Army advisers verified that South Koreans had executed those buried there.

When American infantrymen entered the war near Osan on July 5, they became both POWs and victims. The first North Korean killing-of four captive GIs-took place at Chonui on July 9. Gen. Douglas MacArthur, heading the U.S. Army’s Far East Command and the UN Command, called on all the combatants to observe the Geneva Convention in an announcement broadcast in English and Korean on July 19, 1950. He ordered American commanders to investigate atrocities and to ensure that their troops treated POWs well.

The UN Command, whose forces were retreating in July and August of 1950, did not have many Korean POWs. However, the number of prisoners mounted to 1,899 by the end of August and soared with the Communist defeats of September and October of 1950. By October 31, the UN Command had custody of 176,822 POWs (essentially, any detained Korean), concentrated in three areas: the captured North Korean capital of Pyo?ngyang (80,647) and the southern ports of Inchon (33,478) and Pusan (62,697).

The POW administrators could not provide for their wards-in addition to more than 150,000 refugees-so any South Korean civilian who could convince an interrogator that he had been forced into service as a North Korean soldier or supply bearer was released. South Korean soldiers forced into the North Korean army were turned over to the South Korean army’s military police and intelligence officers for further screening; most remained in custody, along with northern anti-Communist guerrillas who had fled south. All were terrified that their South Korean jailors would execute or torture them, so they were initially docile and cooperative.

[continued on next page]

The Communist defeat in the autumn of 1950 also created complex problems in categorizing the detainees. North Korean soldiers appeared easy to identify; they surrendered in uniform on the battlefield. Yet interrogators soon realized that many of them were impressed South Koreans who were not Communists (or who had made rapid reconversions to anti-Communism). Others were North Koreans, also impressed into the North Korean army, who would have fled North Korea if they could have.

South Koreans posed other problems. When the U.S. Eighth Army and five South Korean army divisions advanced across the 38th Parallel in October, they left behind the U.S. 2nd and 25th Infantry Divisions, three reconstituted South Korean divisions, and the Korean National Police in South Korea to deal with North Korean army stragglers (estimated at 10,000) and thousands of South Korean Communist Party partisans. All Communist Party members and sympathizers in South Korea became fair game. American and British soldiers-and Western journalists-witnessed mass summary executions around Seoul. For any Korean who thought he could be suspected of collaborating, surrendering to the Americans looked attractive.

The Chinese military intervention of October and November of 1950 threw the United Nations’ prisoner management into even greater confusion. As the Chinese People’s Volunteers Force liberated Pyo?ngyang from the Eighth Army and forced the U.S. X Corps and South Korean I Corps to evacuate northeastern Korea by sea, the UN Command evacuated its Pyo?ngyang POW population, along with more than 900,000 additional refugees from North Korea. No longer shooting suspected Communists-at American insistence-the Rhee government wanted refugees and POWs treated alike until their loyalty could be established. Not unreasonably, South Korean counterintelligence officers suspected that some of the refugees were Communist infiltrators and long-term moles sent to reestablish the South Korean Labor Party.

Most of the POWs and shipborne refugees went directly to Koje-do Island, with some shipped to Cheju-do Island, the traditional last bastion of Koreans against foreign invaders and the potential evacuation site for Rhee’s government. Temporarily, the rest of the POWs went to the Pusan area. At the end of December 1950, UN Command had 137,175 Koreans and 616 Chinese in custody.

The Koje-do camps, managed by the U.S. Army 3rd Logistical Command, had been evacuated directly from the areas of Inchon and Seoul. The first 53,588 Korean POWs built their own camps, primarily tents and wooden barracks. Even though UN Command’s military fortunes began to improve by February 1951, the UN Command jailors continued transferring POWs to Koje-do because of threats of guerrilla raids and jailbreaks. By March, 28 different Koje-do compounds had become the home of 139,796 captives, mostly North Koreans-and far more than their intended maximum capacity. The Pusan camps held 8,000 hospital patients, 420 political figures, 2,670 high-ranking officers, 3,500 intelligence targets, and 2,500 administrative and medical personnel.

Because the POWs still seemed docile and cooperative, the camps remained undermanned with ill-trained, poorly armed guards: one South Korean guard for every 26 prisoners and one American guard for every 200-plus prisoners. The POWs and civilian refugees mingled daily while constructing the camp, cooking, and disposing of waste. More refugees served as clerical and medical personnel and interrogator-translators. The camp administrators, focused on short-term management challenges, judged the prisoners cooperative. The Red Cross reported the camps minimally acceptable. A joint U.S.-Korean counterintelligence team warned, however, that the North Korean POW population included a large group of Communist militants capable of violent resistance to the UN Command prison policies.

Three Korean POWs later emerged as leaders of the resistance movement. Sr. Col. Yi (or Lee) Hak-ku had led a mutiny against the commander of the North Korean army’s 13th Division, who would not surrender his battered, helpless division, trapped north of the Pusan Perimeter. Division Chief of Staff Lee had shot (but not killed) that commander and fled to Eighth Army lines with his division’s classified documents. He requested a transfer to the South Korean army, which was denied. Off he went to a prison pen, knowing his former employer, North Korea, would put him to death if he returned home.

Sr. Col. Hong Chol, an intelligence officer, had also been cooperative on surrender, and may have surrendered on purpose, to organize the camps. Pak Sang-hyon, the ultimate commander of the Koje-do resistance, a civilian and Soviet citizen, had been born and raised in the Asian Soviet Union, where his family of radical nationalists had found refuge from the Japanese. He became a full-fledged member of the Communist Party USSR in 1940.

In 1945 Pak joined the Red Army (unwillingly) as a Korean language interpreter, with the nominal rank of captain, for the Soviet invasion of Manchuria and northern Korea. He became vice chairman of the North Korean Labor Party’s branch in rich, populous, and unhappy Hwanghae Province, south of Pyo?ngyang. In October 1950, even before the advancing Eighth Army arrived, Pak fled his station. In the crisis, North Korean leader Kim Il-sung decreed summary execution to any Communist Party member who tore up his party card and deserted his post. To avoid being shot, Pak surrendered to an American patrol on October 21, 1950, in the uniform of a North Korean army private.

Three UN Command initiatives in 1950 and 1951 contributed to the ballooning of the resistance movement among the Communist prisoners of war: an effort to reduce epidemic gastrointestinal diseases among the POWs; the political-religious reeducation programs conducted for the POWs; and the investigation into war crimes committed by the North Koreans and Chinese. None of these UN Command programs alone was responsible for the resistance movement, but all provided extra opportunities for the Communists to reduce UN Command access to the POWs, as Communist planners turned the POWs into weapons of warfare.

Jailors’ access to their prisoners was the issue at the heart of the violence that was about to envelop Koje-do. American and South Korean prison authorities needed access to the prisoners to determine if they wanted to be repatriated to North Korea or China after the war ended.
But the Communist leaders, both in China and in North Korea, sought to prevent embarrassing mass defections to South Korea or Formosa. So they decided to try to force the UN Command to abandon any sort of screening process, thereby prolonging the war and undermining U.S.-South Korean relations.

Given the primitive conditions and overcrowding of the Koje-do prison complex in early 1951, the UN Command’s public health doctors were not surprised when dysentery, diarrhea, and a variety of enteric fevers broke out in epidemic proportions. The key figure in treating the fevers was a U.S. Navy physician, Lt. Gerald A. Martin. Jerry Martin’s medical training in the United States was first-rate, and his family had been medical missionaries in Korea for two generations. His father, Dr. Stanley H. Martin, had been a prominent staff physician at Seoul’s Severance Hospital, Korea’s preeminent medical center.

Jerry Martin, who spoke Korean, rallied Korean Christian doctors, nurses, and technicians from Severance Hospital to establish a major clinic at Koje-do in May 1951. After identifying different strains of parasitic diarrhea, Martin and his staff developed different treatment regimes that cured many of the stricken POWs.

Martin’s fame spread from Koje-do to Korea and to Japan through his military sponsors, the United Nations Civil Assistance Command and Far East Command’s Civil Intelligence and Education Service (CI&E), a program ostensibly for educating POWs about life outside China and North Korea, but in essence an indoctrination program that promoted democracy and Christianity. One of the program’s most effective representatives was the Rev. Edwin W. Kilbourne, Martin’s brother-in-law.

The Communists, however, charged that Martin was conducting germ warfare experiments in the same spirit of Japan’s heinous Unit 731 during World War II. Communist propagandists in Pyo?ngyang condemned Martin as a war criminal even before they announced that the UN Command had waged a bacteriological warfare campaign in Korea and China. They especially focused on an inoculation program Martin’s staff of medical “researchers” had begun for the POWs.

By the autumn of 1951, the public health teams found it harder and harder to reach the POWs; some even attacked them in the compounds. Providing effective medical treatment had become a battleground for the hearts and minds of the POWs.

[continued on next page]

The need to improve living conditions and to keep the POWs busy gave the Civil Intelligence and Education staff plenty to do on Koje-do. Gen. Douglas MacArthur believed that only an ideological campaign that fused democracy, capitalism, and evangelical Christianity could stop Communism in Asia. The indoctrination program became the UN Command’s principal opportunity for prisoner reeducation. Begun with 500 Korean POWs at Yongdungpo, an industrial suburb of Seoul, in October 1950, the indoctrination program combined literacy education (in Korean and Chinese), vocational training, recreational sports and physical education, musical events, and political indoctrination.

The reeducation program began in March 1951 with North Korean civilian internees, refugees from Communism who were expected to become interpreters and camp service workers. The screening immediately took on a religious dimension since the refugees included an estimated 1,000 ministers and 20,000 lay leaders, predominantly Presbyterians and Methodists. The U.S. Army and CI&E civilians found it difficult to sort out the self-proclaimed Christians among the POWs, but Korean Christians in CI&E’s employ could do a more dependable job of testing POWs’ religious convictions.

Indoctrination program participants-estimated at 150,000 POWs and internees in 1951-were trained as metalworkers, brick makers, furniture makers, instructors, textile workers (mats and rice bags), tailors, electricians, musicians, and artists. The indoctrination and industrial program participants became so industrious that Far East Command’s civil affairs administrators feared that Koje-do’s captives were taking work away from the refugees and mainland Koreans. In secret, the POWs also made hand weapons.

Given the separation of church and state in the United States, Far East Command could hardly publicize the indoctrination program’s evangelical character. The Far East Command’s reports provided statistics aplenty on all of the indoctrination program’s literacy, civic education, and vocational training efforts, but they never mentioned that an estimated 60,000 POWs and civilian internees “accepted Christ.”

Communist Chinese and North Korean political leaders fully appreciated that Christianity represented a more serious ideological challenge than the vague promises of Western democracy and capitalism. In broadcasts and publications of the Chinese People’s Committee on World Peace, in speeches at the Panmunjo?m negotiations, and through sympathetic journalists Winnington and Burchett, they repeatedly attacked the indoctrination program staff as religious fanatics, dupes, fellow travelers with the Chinese Guomintang fascists on Formosa, and capitalist agents.

The Far East Command’s pursuit of North Korean and Chinese war criminals also spurred the Communist POW resistance movement. By August 1950, the bodies of executed GIs and a handful of massacre survivors proved that the North Koreans were flouting the Geneva Convention on POW treatment. Evidence of repeated, large-scale massacres mounted in October 1950 as the United Nations forces advanced to Seoul and then into North Korea. That month the North Koreans killed 138 Americans in one incident near Pyo?ngyang, and they killed South Korean civilians by the thousands at Taejo?n, Chongju, Hamyang, Mokpo, Kwangju, Pyo?ngyang, and Wonsan, where they had been jailed as “enemies of the people.” The Communists also eradicated suspect North Korean civilians as they retreated north.

The number of incidents and the horrific evidence of mass burials required a quick response, so in October 1950, General MacArthur directed the Eighth Army commander, Lt. Gen. Walton H. Walker to organize a War Crimes Division within his Judge Advocate General’s office and collect evidence of POW murders. He intended to prosecute Communist war criminals. Investigators concluded they had evidence of some 400 atrocities among the nearly 2,000 incidents investigated. The lawyers concluded they could try 326 cases and compiled a list of suspects and witnesses.

The investigators, desperate for results, developed a network of informants at Koje-do. No security precautions could disguise that perhaps a thousand or more POWs might be tried as war criminals, based on their own confessions and the testimony of their fellow prisoners. The South Korean investigation teams, aided by POWs, conducted long, painful sessions with suspects and witnesses. The South Korean army lieutenant who supervised the largest number of agents admitted using torture to extract confessions and expose Communists, as well as suspected murderers.

The hard-core Communist POWs-though a fraction of the total Koje-do prison community-had plenty of reasons to disrupt the 1951 camp routine. The first collective violence against camp personnel occurred on June 18 and 19, 1951, when some North Korean officers protested having to dig latrines and garbage pits. When a South Korean guard detail entered Compound 76, the prisoners stoned the guards. Soldiers opened fire, killing three POWs.

More incidents followed: demonstrations within the compounds, work refusals, threats against camp personnel, and some 15 murders of Korean prisoners. In July and August of 1951, the guards killed eight more POWs. On one occasion, the guard force had to rescue 200 POWs from Compound 78, where hard-core Communists had executed three supposed collaborators in a plan to control the compound. Camp administrators noted that the demonstrations and protests followed the first visits of Red Cross inspectors and journalists to Koje-do. The public relations crisis stirred Gen. Matthew B. Ridgway, who replaced MacArthur in April 1951, to order Lt. Gen. James A. Van Fleet, Eighth Army commander, to fix Koje-do-which meant quiet things down there.

In late September, Van Fleet and his staff visited Koje-do and concluded that the physical conditions were adequate but that there were too few guards and they were poorly disciplined. POWs had too much free time and independence, and the indoctrination program instruction was too classroom oriented. Van Fleet sent a new U.S. Army military police battalion to the island, which brought the 8137th Military Police Group up to three battalions and four escort companies.

In December 1951, a battalion of the crack U.S. 23rd Infantry Regiment augmented the guard force. More South Korean army MPs arrived, too. The guard force now numbered 9,000 officers and men, but it was still 40 percent below the force requested by Brig. Gen. Paul F. Yount, commanding general, 2nd Logistics Command. Yount relieved the U.S. Army camp commander and persuaded the South Korean army to find another colonel to lead its guards. Reinforcements, reforms, and reliefs-the traditional military response to a crisis-should have made a difference. They did not.

In July 1951, Sino-Korean allies and the Soviet Union and the United States agreed to discuss an armistice. Neither the Chinese nor the North Koreans liked negotiating but Stalin, providing the munitions and limited air defense for the Chinese and Korean armies, had enough leverage to draw Mao Zedong and Kim Il-sung into negotiations. The Sino-Korean allies also needed some relief from combat.

The Chinese demanded to manage the talks, and the negotiators established a rough agenda. Agenda Item Four would deal with the issue of a prisoner exchange. The South Koreans had screened all but the hard-core Communists and concluded that they could release almost 38,000 civilian detainees, all from South Korea. Indoctrination program leaders confirmed the estimate, reporting (too optimistically) that 98 percent of the Korean civilian internees had become militantly anti-Communist and wanted to remain in South Korea. General Van Fleet endorsed their release, which would have simplified his security problems, but Ridgway, with guidance from Washington, rejected a total release. About 37,000 civilian internees were finally released in November 1951, but thousands remained.

Their anti-Communism now exposed, several thousand internees, most from North Korea, remained behind the wire-essentially as hostages to the truce talks. When General Ridgway issued “Articles Governing United Nations Prisoners of War,” effective November 1, 1951, he stressed that the articles applied to everyone detained by United Nations Command. Any act of violence or noncooperation might be punished-with the death penalty available for mutiny, murder, rape, assaulting a member of the UN Command, and organizing a mutiny.

The Communists had been equally busy turning the POWs to their advantage. They could see that the United Nations Command had difficulty censoring or shaping the reports of Western journalists, not only Communist champions like Winnington and Burchett but also non-American reporters and representatives of American newspapers critical of President Harry Truman’s administration. The North Koreans, very junior partners on the battlefield, wanted to sponsor the POW resistance and increase their influence at the truce talks. They set their ultimate goal as capturing Koje-do and staging a Great Escape, aided by South Korean Communist guerrillas.

The organizational instrument was the Guerrilla Guidance Bureau, an agency of the Korean People’s Army’s political department, directed from Panmunjo?m by its leader, Gen. Nam Il, who was actually a Soviet citizen. His deputy, Gen. Kim Pa, established contact with the Communist POW leaders.

In the autumn and winter of 1951-1952, the Guerrilla Guidance Bureau sent an estimated 280 male and female agents to Koje-do, some to be POWs, others to establish local support units on Koje-do and the mainland. Col. Bae Chul, a Soviet citizen and Red Army officer, managed the infiltration through Pusan.

[continued on next page]

Agents informed Pak Sang-hyon in Compound 62 that General Nam had named him chief resistance leader. With the help of colonels Lee and Hong, Pak was to seize control of as many of Koje-do’s compounds as possible and use violence to antagonize the Koje-do guards and administration. POW losses would be inevitable, and retribution was encouraged to ensure press coverage. The ultimate goal was a combined POW-guerrilla attack to seize the island. The resisters-especially the leaders-would be heroes welcomed back to the ranks of the Communist faithful. If they failed, their families would die-or suffer worse.

The Communist resisters and their anti-Communist counterparts tested their strength over the issue of screening the remaining Korean civilian internees, most of them former South Korean soldiers and guerrilla suspects. Compound 63 produced the vanguard anti-Communist camp government, supported by an ultra-patriotic paramilitary society called the Hwarang Association. When South Korean screening teams approached Communist-dominated Compound 62, 300 members of the Hwarangs arranged a transfer to Compound 62 to assist the screening by doing away with the Communist organizers.
The compound had already been the scene of a reign of terror against the non-Communists, a minority of the 5,000 captives. Deaths and torture had been inflicted by beating, burning, electrocution, castration, mutilation, and drowning.

The coup attempt went awry and produced a riot. After three hours of violence on December 18, 1951, that left 14 dead and 24 wounded, the South Korean guards rescued the Hwarangs and 100 other anti-Communists. Three days later, all civilian internee screening ended in Compound 62, now under complete Communist control. Pak Sang-hyon became the first commissar of the resistance leadership.

The camp commander, Col. Maurice J. Fitzgerald, continued to feed the captives but attempted no other contact with Compound 62. Instead, Fitzgerald surrounded the compound with guard strongpoints designed to prevent a breakout. Pak and his cadre displayed protest signs in Korean and English, drilled with spears, sang and chanted with revolutionary fervor, and threw rocks at passersby. The civilian internee screening and the indoctrination program began to falter in other Korean compounds.

The resumption of armistice negotiations on November 27, 1951, brought a dramatic escalation of POW resistance. The issue was the final disposition of the POWs. The Communists, citing the Geneva Convention of 1949, argued that each side had to return all captives and internees without any consideration of the POWs’ personal preferences. The Communists feared a potential propaganda disaster-multitudes of Korean and Chinese voluntarily and loudly rejecting Communism.

The UN Command saw the issue differently. General Ridgway suspected that he held far more POWs than the Communists, even if as many as 41,000 Koreans were reclassified from POW to civilian internee status and held out of the POW exchange, not being enemy combatants. The UN Command held at least 100,000 POWs, while the Communists had perhaps 6,000 American and allied troops and maybe 28,000 South Korean soldiers.

The U.S. Joint Chiefs of Staff favored a one-for-one exchange, but would accept “all for all,” and as quickly as possible to save lives. Ridgway had ample evidence that Communists were committing atrocities against prisoners in North Korea, and allied POWs certainly lived in miserable conditions. Yet “all for all” also meant returning war criminals, collaborators, intelligence operatives, and many innocents to Communist control.

The POW exchange issue escalated when South Korea’s Syngman Rhee demanded no forced repatriation, the immediate release of cleared civilian internees, and the prosecution of war criminals. The Truman administration divided; only the president himself could resolve the issue.

In December 1951, Truman chose voluntary repatriation. With the president’s approval, the actual plan envisioned a two-phase exchange: “one for one” until all UN Command POWs had been released, and then voluntary repatriation. The Chinese negotiators originally considered the proposal a stalling tactic related to other issues like the location of the future military demarcation line. The North Koreans accused the UN Command of covering up its own war crimes and of POW intimidation. Nevertheless, all sides agreed to prepare a list of POWs in custody.

The exchange of these hurriedly compiled lists surprised, shocked, and outraged each side. The South Korean army estimated it had lost 70,000 soldiers to captivity and could not account for 88,000 others. Yet, the Communists claimed to hold only 7,412 South Korean soldiers. (In fact, the Communists had impressed thousands into their own armed forces, where many died in action. Thousands of others, mainly “class enemies,” died in slave labor camps from which POWs were still escaping four decades later.)

The lists had surprises. The UN Command had more than 10,000 MIAs and thought at least 6,000 were-or had been-in enemy hands. Nevertheless, only 3,198 American names were listed, 1,219 from other allied units, and the list excluded some people known to be POWs.

The Communists were equally shocked (or put on a convincing act) to learn that the UN Command held only 95,531 Communist Koreans, 16,243 uncleared South Koreans, and 20,700 Chinese soldiers. They also heard estimates from Western sources that more than half their POWs would refuse repatriation.

As the Communist POW resistance organized in early 1952, the conditions for exchanging prisoners came to dominate the Panmunjo?m negotiations. The UN Command position, formed in Washington, was that no POW would be repatriated against his will. The Communists insisted that the Geneva Convention of 1949 required that all POWs be returned. The Communist POW resisters, now with plenty of incentive, tested the camp administrators in a violent protest on February 18, 1952. Armed with rocks, clubs, and spears, between 1,000 and 1,500 inmates of Compound 62 attacked a South Korean team trying to rescue non-Communist civilian internees in the compound. A battalion of the U.S. 27th Infantry fought back, killing 75 captives and wounding 139 others. One GI died in the melee and 22 were wounded. Five days later General Nam charged the UN Command with “barbarously massacring” large numbers of harmless civilian internees.

With neither side wavering on POW exchange at Panmunjo?m, Ridgway planned to screen all the Koje-do captives for their preference on repatriation, at Washington’s direction. He ordered “a detailed plan to provide for the selection, segregation and protection of those North Korean and Chinese Communist prisoners of war who would violently oppose repatriation to Communist control.” The South Korean POWs or civilian internees would be allowed to elect repatriation or to stay in South Korea. Ridgway warned Van Fleet to ensure that disorder, riot, and bloodshed were kept to an absolute minimum. He advised taking as little time as possible for the screening, “preferably during daylight hours of a single day.”

[continued on next page]

The task went to Brig. Gen. Francis T. Dodd, Eighth Army deputy chief of staff, reassigned to command Koje-do Camp One with Col. Maurice J. Fitzgerald as his deputy. Dodd’s staff drafted a plan, Operation Spreadout, which would send an estimated 82,000 POWs and civilian internees to new camps on the mainland and to Cheju-do Island. Dodd thought that screening would be inevitably violent and didn’t think it could be done quickly, but that the screening could be supported by promises to protect the anti-Communist POWs.

On March 13, resisters in Compound 76 stoned a passing work detail. South Korean guards fired, killing 12 and wounding 26. In trying to stop the shooting, a Korean indoctrination program staffer and a U.S. Army officer were wounded. The entire U.S. 38th Infantry regiment then joined the guard force because of intelligence assessments that the Communists wanted to destroy the screening process with a mass jailbreak. Yet Ridgway rejected Van Fleet’s request to postpone screening. Instead, he ordered strict new controls on the indoctrination program and on the activities of Chinese nationalist agents. He also wanted recommendations on reducing the South Korean guard force, Korean service personnel, and refugees at Koje-do.

The screening began April 8 in the 11 compounds deemed most friendly. On April 10, however, Koreans in Compound 95 captured a medical party, and South Korean soldiers armed with clubs had to rescue them. The melee spread when other South Korean soldiers opened fire on the mob. A U.S. Army officer with a jeep-mounted machine gun stopped a rush to the gate. Three POWs died, and 60 fell wounded while one South Korean soldier disappeared and four were wounded. Operation Spreadout started separating repatriates from those preferring not to return. In command of the entire POW system, General Yount began moving the Koreans who refused repatriation from Koje-do to mainland camps at Pusan, Masan, Yongchon, Kwangju, and Nonsan. The captive Chinese, divided into repatriates and nonrepatriates, would be sent to new camps on opposite sides of Cheju-do Island.

Heavily protected, the UN Command screening teams worked their way through 22 of 28 compounds not firmly under Korean Communist control. They deemed six compounds-four North Korean army compounds and two for Communist Party members and guerrillas from all over Korea-to be too well-armed, well-organized, and belligerent to enter until the prospective battlefield had been cleared of all nonrepatriates and refugees. The nonrepatriate Chinese left the island first for Cheju-do, many unhappy they were not headed to Formosa. Some 7,000 Chinese POWs who wanted to go home to their families were left behind, only a few of them Communists.

By April 19, 1952, General Dodd’s teams had screened 106,376 POWs and civilian internees. Of these, 31,244 chose to return to Communist custody, while 75,132 preferred being sent to South Korea, Formosa, or elsewhere. Screening at the 64th Field Hospital in Pusan shows the strength of the resistance movement there: 4,774 POWs wanted repatriation; 1,738 did not.

Forwarding Dodd’s report to Washington, Ridgway warned that the Koje-do guard force still faced 43,000 violent, vicious North Korean resisters in six compounds: 37,628 POWs commanded by colonels Lee and Hong and 5,700 civilian internees directed by the mysterious, unidentified Mr. Pak. Yet many POWs had already been shipped from Koje-do to Pusan, Masan, Yongchon, Kwangju, Nonsan, and Cheju-do. It appeared Operation Spreadout was nearing completion in relative peace.

But this progress on nonrepatriates galvanized the resistance leaders to take desperate action in May 1952. Whether the resistance leaders received specific directions from General Nam is unclear, although reporters Winnington and Burchett later claimed that Pak, Lee, and Hong acted without orders, which brought them disgrace. The Koje-do Three-as Pak, Lee, and Hong came to be known-decided to kidnap General Dodd and force him to confess to mistreating the Communist POWs. At a minimum, they expected to create a media sensation and win a propaganda victory. Perhaps there would be a breakout. The plotters could also assume that they still had informers in their midst, requiring them to move quickly.

On April 29, the North Korean officers of Compound 76 asked to meet with Lt. Col. Wilbur Raven, a military police officer and enclosure commander. The meeting was supposedly to resolve Raven’s suspension of a cigarette ration after North Korean army officers refused to serve on work detail. Raven and a South Korean interpreter entered the “headquarters” hut just inside the wire and began listening to a barrage of demands. Suddenly more than a hundred cadremen flooded the building. They screamed at Raven and one tried to force-feed him bean soup. Then the POWs produced an EE8 field phone and told him to call General Dodd. After Raven passed on the prisoners’ demands-which Dodd rejected-the POWs released Raven. The whole bizarre episode was a rehearsal.

On May 7, the diehards of Compound 76 seized General Dodd. Responding to another POW request to discuss prison conditions and screening, Dodd and Raven met a delegation at the compound’s front gate. Discussions through the outer wire lasted more than an hour. Dodd, following Raven’s advice, was unarmed but armed GIs protected him. Then a “honey bucket” crew came by and a guard opened the outer gate. With a yell, POWs grabbed Dodd and almost captured Raven, who grabbed a gatepost and kicked his assailants away until the guards rescued him. As he was being carried away, Dodd ordered his soldiers not to shoot. None did.

The Dodd affair-called a mutiny by Western writers-threw unwelcome light on the UN Command’s POW management and the voluntary repatriation policy. It occurred as General Ridgway was turning over the Far East and UN Commands to Gen. Mark W. Clark. Ridgway ordered the use of “whatever degree of force” necessary to gain Dodd’s release. His demand to turn over Dodd was met with catcalls when read to the POWs. The next day a new Koje-do commander, Brig. Gen. Charles F. Colson, arrived with an American infantry battalion. An able, decorated infantry commander of World War II, Colson warned the POWs not to harm Dodd, his friend. He deployed all his infantry and mounted machine guns to deter a mass jailbreak, predicted by Dodd’s G-2.

Dodd reported his status to Colson, first by notes and later by telephone. His captors were treating Dodd with respect, but he found himself discussing prison reform, repatriation policy, and capitalist injustice-with his release at stake. The POWs also threatened him with a “trial” as a war criminal. The Communists placed festive protest signs around Compound 76 warning that Dodd would die if Colson tried to rescue him.

In May 9, Van Fleet, the Eighth Army commander, came to Koje-do to review the plans to break Compound 76’s resistance and save Dodd, though the latter was a secondary concern. The general had company: Ridgway and Clark. Both urged Van Fleet to give the POWs plenty of opportunity to surrender. They agreed there would be no media coverage of the Koje-do crisis. The critical matter, they thought, was massing firepower to neutralize the 19,000 POWs outside the hard-core Compounds 76, 77, and 78. Only then would American soldiers enter Compound 76, using tear gas and other riot-suppression equipment.

Van Fleet gave Colson another day for negotiations. Company B of the 64th Tank Battalion, 3rd Infantry Division, had not arrived from the mainland with its 22 tanks, including five flamethrower tanks. General Van Fleet also knew, however, that generals Ridgway and Clark wanted the Koje-do problem to disappear.

For reasons that are unclear, Ridgway did not tell Clark of Dodd’s capture until May 8. Ridgway knew that whether Dodd lived or died, the Communists would probably fight relocation, and he knew they were prepared to die in large numbers, which could not be kept secret. Ridgway may have preferred to let the inevitable massacre occur on Clark’s watch, because Ridgway wanted to be U.S. Army chief of staff. Clark wanted to retire.

Not briefed about the POW rebellion on his way to Tokyo, Clark was dismayed by the Koje-do surprise, which he called “the biggest flap of the entire war.” After negotiations the next day with the captors and several telephone conversations with Dodd, the two American brigadier generals made a deal. Eventually the Koje-do Three accepted a signed statement by General Colson that United Nations guards had killed and wounded “many prisoners of war.” Colson promised to “do all within my power” to treat the POWs according to international law; he said he had no authority to modify the UN Command position on voluntary repatriation.

He promised he would conduct no more “forcible” screening sessions, nor would he force nonrepatriates to again bear arms, and he would recognize POW representatives chosen by the POWs themselves. At nine thirty in the evening on May 10, Dodd walked out of Compound 76, unharmed. He and Colson were not, however, out of harm’s way.

Always public relations-conscious, Clark wanted Ridgway to explain the crisis to the press, but Ridgway issued no statement until May 12, when Clark coaxed him to approve a memo Clark’s staff had prepared, announcing that Dodd had bought his freedom two days earlier with what amounted to a confession of war crimes that implicated Ridgway and Van Fleet. Colson had signed the statement only to recover Dodd.

General Clark immediately disavowed Colson’s “confession” and produced an account he and Ridgway had agreed on even before Van Fleet and Dodd talked to the press in Seoul. Clark’s statement, however, revealed such a degree of ignorance of conditions on Koje-do that he could have been judged either a fool or a liar by anyone who knew about the events, including newsmen. For example, Clark claimed that the POW violence had been exaggerated and had no bearing on the armistice negotiations, when clearly they were central to those meetings.

[continued on next page]

Dodd’s account of his captivity revealed a more complicated picture of the Communist POW resistance movement. Dodd also said that Colson’s concessions were “of minor importance” and the POWs’ demands inconsequential, a view shared by Van Fleet. An inevitable investigation by a board of Eighth Army officers found Colson’s conduct meritorious but Van Fleet rejected this finding, under pressure from Clark. Gen. Omar N. Bradley, chairman of the Joint Chiefs of Staff, strongly suggested that Clark punish Dodd and Colson. Clark convened a board of Far East Command generals that recommended sanctions. Without further investigation or formal charges, Dodd and Colson returned to their permanent rank of colonel and left the army for ignominious retirement. General Yount, who was not represented, received a letter of reprimand for being a propaganda embarrassment, in Ridgway’s words as “humiliating and damaging a defeat as any that might have been imposed in bloody battle.”

The Koje-do Three and the hard-core North Korean Communist leadership had succeeded in pulling off the resistance movement’s most theatrical and deadly propaganda victory. But now the UN Command had to tackle the problem of its central prison being run by armed prisoners.

To relieve the Eighth Army of its long-term POW management responsibilities, Clark appointed a new commander for Koje-do Camp One, Brig. Gen. Haydon L. “Bull” Boatner. Boatner specialized in Chinese language and culture and had spent 10 years in Asian tours, giving him special insight in the psychology of Asian soldiers. With his eyeglasses, thinning hair, and flabby physique, he had an avuncular look that belied his profane, bullying, perfectionist, professional approach to command. He chose tough Col. Harold Taylor as his deputy. Thinking the enlisted men were “the poorest quality of American soldiers with whom I’ve ever served,” Boatner got quality replacements. He also had Clark lend him a judge advocate schooled in the Geneva Convention to review Koje-do operations, and Boatner intended to follow his lawyer’s advice.

Clark contributed the paratroopers of the 187th Regimental Combat Team. Boatner wanted more crack troops, and Van Fleet agreed, ordering the British Commonwealth Division to produce some troops to pacify Koje-do. As he learned about the POW resistance, Boatner came to view the insurgency as a “self-inflicted mess” created by American commanders who knew nothing of Asians or POWs. Boatner was appalled to see undisciplined GIs rush to the wire to scream catcalls and throw rocks at demonstrating POWs. He relieved three of the four senior MP officers and culled NCOs from the MP brigade. To counter Communist atrocity charges, he opened the new camps to inspection by a team from the International Red Cross, and Boatner insisted that 40 to 50 war correspondents come to Koje-do and report his actions.

Before he could attack Compound 76, Boatner had to establish new quarters for five compounds of North Korean soldiers and two for civilian internees, in all almost 70,000 determined repatriates. He planned new, smaller, closely guarded camps that were more isolated, more secure, and that would depend less on Korean service personnel and refugee labor. The small islands of Yoncho-do and Pongam-do would house 12,000 segregated leaders and troublemakers. He ordered another small camp (Chogu-ri) to be built for the same purpose on Koje-do. The equivalent of solitary confinement, it nonetheless complied with the Geneva Convention.

On June 10, 1952, General Boatner assaulted Compound 76 with a tank platoon and two battalions of paratroopers. Although Communist “shock troops” charged the GIs with handmade flails and spears while others threw firebombs from trenches and dugouts, Boatner’s troops defeated 6,500 North Korean officers and NCOs with relatively few casualties.

First they ordered the POWs into their barracks on the threat of machine gun fire. Some POW resisters fought on for three hours, although their comrades-including Sr. Col. Lee Hak-ku-streamed from the compound to surrender. Thirty-one POWs died and 131 were wounded. Army investigators later decided that fellow prisoners had murdered about half the dead for sympathizing with the Republic of Korea. One paratrooper bled to death from a stab wound and 14 others collected Purple Hearts. In the next two weeks, residents of six other POW and civilian internee compounds moved to the three new camps in the Koje-do system, without resisting (but not before murdering 15 more prisoners). Around 48,000 POWs remained in old Camp One.

The completion of what had been dubbed Operations Spreadout and Breakup restored United Nations Command control over the prison population. Van Fleet felt secure enough to allow the South Koreans to free 27,000 South Korean civilians who had proven their identities and loyalties in June and July of 1952. Another 11,000 South Koreans who had been impressed into North Korean army service went home. The camps came much closer to Geneva Convention and Red Cross standards.

Still, POW resistance did not disappear, because its basic external causes (voluntary repatriation and UN Command coalition politics) had not disappeared. The North Korean political officers at Panmunjo?m had ample reason and opportunity to demand more resistance. President Syngman Rhee of South Korea demanded that Van Fleet release all nonrepatriates, provided South Korean investigators cleared them. Indoctrination program teams and South Korean agents fed the growing anxiety of the “detainees” that they were pawns to the Panmunjo?m talks. They spread the rumor that the UN Command negotiators had promised to return no less than 76,000 POWs. Rhee used the detained anti-Communist Koreans to stop or slow the armistice negotiations, which he opposed.

The British, who regarded the American handling of POWs as dangerously inept, began to criticize UN Command policy, starting with the report of Maj. D.R. Bancroft, the commander of a two-company task force deployed on Koje-do from May 23 to July 10, 1952. These British and Canadian troops were assigned to control Compound 66 (militant North Korean army officers led by Col. Hong Chul) and thought American and South Korean treatment of the Chinese and Koreans beyond contempt. They found the POWs in charge of everything behind the wire. When the British searched a barrack, they found money, escape maps, medical supplies, weapons, and civilian clothes, all provided by Korean guards.

The British stopped the flow of gasoline and nail-studded firewood to the compound and shut down the ordnance factory that had been disguised as a blacksmith shop. Responding to the new media access, the island sprouted protest signs. As the British grasp tightened on 4,000 POWs in four new compounds, they found that strict standards and human decency paid dividends when the Red Cross cited the group for its adherence to the Geneva Convention. Major Bancroft left Koje-do more impressed by the dedication and discipline of the North Koreans than by the Americans and the South Koreans.

His report caused a furor in the British, Canadian, and Australian defense and foreign ministries and made POW management an item of interest for several Commonwealth generals.

Segregating the Chinese POWs on Cheju-do brought the Chinese resisters to life. Relieved from the domination of the nonrepatriate majority and supported by Chinese nationalist agents and indoctrination program personnel, the leaders of the Chinese resisters organized a series of protests that began in August 1952. On October 1, the anniversary of the creation of the People’s Republic of China, the Cheju city compounds turned red with makeshift flags, banners, and decorations.

[continued on next page]

The American camp commander ordered his guards, MPs, and a battalion of the 35th Infantry to tear down these symbols. The Chinese fought back with spears and gas bombs. In a melee that swept over three compounds, 56 Chinese were killed and 91 more were hospitalized. Nine GIs were wounded.

Internal political struggles within the POW resistance movement probably prolonged the revolt. By autumn 1952, it looked as if there would be an armistice and a POW exchange. The Koreans and the Chinese knew repatriates would have to prove that their capture had been unavoidable and their resistance heroic, up to the highest Communist standards.

In December 1952, the resisters sparked 48 incidents that left 99 POWs dead and 302 wounded. The worst of these occurred on Pongam-do, where Pak still commanded the civilian party faithful. On December 14, 85 resisters died rushing the wire in a breakout attempt, the last such escape effort. Suicidal resistance continued at the POW hospitals at Pusan where, by refusing treatment, resisters were making political statements. Undercover Communist doctors and nurses killed patients they regarded as traitors. Over the winter, as armistice negotiations stalled again, incidents dwindled but for the occasional murder of Koreans and Chinese of suspect loyalty. In the spring of 1953, everyone in custody became ever more anxious about their fates. The Koreans who wanted repatriation numbered 66,754 POWs and civilian internees, and there were 8,840 Chinese POWs. Virtually all the nonrepatriates were former soldiers: 35,597 Koreans and 14,280 Chinese.

In March, Josef Stalin died-the single most critical event leading to the armistice. The Soviet politburo warned Mao Zedong that he could no longer count on military assistance. Syngman Rhee made demands for buying his acceptance of an intolerable peace: a mutual defense agreement with the United States and the promise of over $1 billion in economic and military assistance. As part of his pressure on the United States, Rhee began to threaten to release all the Korean nonrepatriates. This encouraged noncooperation in all the camps, not just those of the repatriates.

On June 13, Korean nonrepatriates beat eight Communist agents, killing one. On June 17-19, the South Korean Counterintelligence Corps and the Military Police Command organized the breakout of 27,000 of the 35,000 Koreans who had refused repatriation. The only Americans who tried to stop the escapees were marines of the 1st Shore Party Battalion, a temporary guard force at a minor compound at Ascom City near Seoul. The marines thought the Koreans had weapons and wanted to fight. They opened fire, killed 44 escapees, and wounded more than 100. At all the other camps, only 17 other escapees died. In the aftermath of the great escape, an additional American infantry regiment joined the guard force, for fear of a Chinese jailbreak at Cheju-do. In all the camps, anti-Americanism became an epidemic.

Operation Little Switch, an agreed-upon exchange of sick and wounded POWs in April, had established the norms of all the subsequent exchanges. The sick and wounded Chinese and Koreans postured and posed, spit and cursed, sang and chanted and, even on stretchers, tried to tear off their prisoner garb. The exchange, conducted before the armistice, did not require further screening, a critical difference from the post-armistice Big Switch. There were 6,670 Chinese and Koreans going north and 684 UN Command personnel (471 South Koreans) returned to allied care.

The POW resistance movement staged its last protests as part of Big Switch, the POW exchange that followed the ceasefire on July 27, 1953. After much acrimonious debate, the Communists accepted the principle of voluntary repatriation, but only if it included rescreening those North Koreans and Chinese POWs who had rejected repatriation: some 14,704 Chinese and 7,900 Koreans. The Communists held 359 prisoners who opposed repatriation, including 335 Koreans.

Article III of the Korean Armistice Agreement established the exchange. The first phase would transfer prisoners who chose repatriation to the neutral territory around Panmunjo?m. The Committee for Repatriation of Prisoners of War, three representatives each from the two belligerent coalitions, supervised the movement. The second phase focused on those captives who rejected repatriation. Within a 60-day period, Red Cross teams would go to all the POW camps, escort the prisoners to Panmunjo?m, and supervise interrogating the nonrepatriates to ensure their decisions were truly voluntary (if that could be determined). The Communist political officers would try to persuade the POWs to change their minds. In addition to the belligerents’ representatives, officer teams from Switzerland, Sweden, Poland, and Czechoslovakia (the Neutral Nations Repatriation Commission) would direct activities. Three thousand Indian troops were to arrange required interviews under neutral protection.

Much to the displeasure of the United Nations and the South Korean government, the Communists immediately turned the Indians into impotent onlookers. The Communist delegation demanded that the Indian Custodial Force break up the militant anti-Communist prisoner leadership organization. They published a list of 400 UN Command collaborators and agents among the 22,604 prisoners. The Indians replied that they would not use force to keep order, an invitation for a renewed war behind the wire. With only 1,300 troops to guard 55 compounds over three and a half square miles, the Indians could not stop attacks on waverers. The Indians did shoot and kill two escapees in October, and wounded three demonstrators. The Indians tried to court-martial seven Chinese for murder but witnesses were scarce. On December 12, four Korean bodies were found in a ditch inside the compound.

The Communists harassed interrogators and took names and addresses, giving a strong impression that they would attack the families of any staying behind. The prisoners reacted with suicides and suicide attempts, and sometimes attacked Chinese and Korean officers. Under impossible conditions, the screening teams talked to only about 5,000 POWs before giving up on December 31, without requesting an extension.

The commander of the Custodial Force, Lt. Gen. K.S. Thimayya, reported to the Neutral Nations Repatriation Commission that the Communists had turned the process into a show, and 38 POWs had died. Four hundred forty Chinese and 188 Koreans changed their minds and went north; 86 others went to India and then scattered around the world. Repatriates and nonrepatriates alike used Panmunjo?m as a final stage for mass demonstrations and gang battles for the members of the media to observe.

Banners flying, marching to songs and chants of protest and revolution, the Communist POWs stripped off their hated UN Command uniforms and marched into an unknown future. The Chinese resisters found themselves pariahs, condemned by their army and the Communist Party for being captured. One-third of the Chinese People’s Volunteers Force captives, dedicated Communists or not, chose repatriation in the face of death, beatings, mutilation by tattooing, and self-hate. They returned to a Communist China that regarded them as traitors for 40 years.

The North Korean soldiers, guerrillas, and party cadremen who returned home had no better luck. Despite leading the resistance, Pak Sang-hyon and Col. Lee Hak-ku were shot for treason (which was also not an isolated event in the post-armistice Democratic People’s Republic of Korea), although they may have been allowed to see their families first. Col. Hong Chul, who may have been a planted POW commissar, simply disappeared.

For UN Command/Far East Command, the challenge of the Communist POW resistance movement produced ample lessons in handling captives who still considered themselves combatants. The lessons faded in the 1950s, marginalized by the rant in the United States over the alleged misbehavior of captive GIs by “brainwashing,” the “war crimes” confessions of American airmen, and the imagined moral corruption of American youths in uniform. The Code of Conduct (1953) applied to American captives, not their Communist counterparts.

The principle of voluntary repatriation influenced the POW exchanges in 1954 in French Indochina, freeing thousands of Viet Minh cadremen, who later subverted Vietnam. Indeed, since the Korean War, the war behind the wire has continued in conflicts throughout the world.



Click For More From MHQ!
Click For More From MHQ!