The Third Man

A familiar name appears in an unfamiliar place

The Third Man, which was released in theaters in early 1950 and starred Joseph Cotten, Alida Valli, and Orson Wells, is on the short-list of my all-time favorite films.

One of the most famous exchanges in the film comes when the Joseph Cotten character, an American pulp fiction writer named Holley Martins, tells the detective investigating a recent murder about a major discovery he made relating to the case.

“There was a third man there. And I suppose that doesn’t sound peculiar to you?”

“I’m not interested in whether a racketeer like Lime was killed by his friends or by an accident. The only important thing is that he’s dead.”

That scene was on my mind as I drilled down on a crucial detail about the Moon Duck Harr OSS personnel file.

Remember, my hypothesis is that Moon Duck Harr is a pseudonym for Karl Harr, Jay Reid’s cousin, who would go on to serve in a number of high-level government positions under President Eisenhower. An open code translation would be:

• Moon - Korean agent (cover)

• Duck - indication that the name is a pseudonym

• Harr - real surname

Getting to the heart of this OSS matter is critical to unspooling the entire story of Jay’s still-classified relationship with the CIA.

The OSS training center

During the holiday break, I puzzled over the training center listed on the top of a student evaluation document found in the Moon Duck Harr personnel file at the National Archives.

My first thought was that Area “WA”, when combined with West Coast Training Center, pointed to a site somewhere in Washington state.

We know that is wrong because the OSS did not have any training locations in Washington.

Perhaps the “West Coast” part was a ruse, and WA really meant Washington, D.C.?

Wrong again.

Eventually I found the answer in an extensive report written for the National Parks Service by Dr. John Whiteclay Chambers II.

Area “WA” is to be read “Area W-A”, as in West Coast, Area A, which fits the naming convention for sites along the East Coast (A, B, C, etc.).

This particular training site was located on Santa Catalina Island, about 22 miles off the coast of Southern California, near Los Angeles.

Chambers summarized the training situation on Catalina Island in an article he published with the blessing of the CIA that was based on the Parks Service report:

The number of OSS training camps in the United States increased to 16 in the last 12 months of the war as the original training areas and assessment stations in Maryland and Virginia were augmented by a communications school, designated Area M, at Camp McDowell, near Naperville, Illinois, and eight relatively new training facilities in southern California. The most prominent of these “W” areas was on Santa Catalina Island, as the focus of war effort shifted to the defeat of Japan.

When Phillip Allen, head of West Coast schools, arrived from S&T [Schools & Training] headquarters, he was able to institute a well-coordinated program there. His success was due in part because, except for the Maritime Unit, which already had its own school there, the other operational branches did not have training facilities there, and this enabled Allen largely to start afresh. His training program began with the new basic, unified, two-week E Course. This was followed by an advanced course in SI [Secret Intelligence, aka intelligence gathering], SO [Special Operations, later called covert operations], or MO [Morale Operations, aka psychological warfare or propaganda], or a combination of them.

In the summer of 1944, Allen was able to obtain as instructors seasoned veterans who had real experience and information on current conditions in the war zones and who could provide practical advice to their students. Training concluded with extremely demanding field problems, as some of the students—Korean Americans, Japanese Americans, and some Korean prisoners of war—were preparing for infiltration into Japanese occupied Korea or Japan itself.

When I say that the article quoted above was blessed by the CIA, I mean that there is literally a disclosure form on page 2 indicating that Chambers wrote the article for the Agency.

This is not to say that because the CIA signed off on the text that the article is propaganda. It is not.

But keep in mind that unclassified articles published for the CIA’s Center for the Study of Intelligence can present an incomplete picture of the true situation.

Chambers goes into more detail in the Parks Service report (emphasis added):

In 1944-1945, the OSS, using a Korean American, recruited a number of volunteers from among the Korean POWs interned at Camp McCoy for infiltration and subversion of the Japanese in Korea or Japan itself. Men with sufficient patriotism, intelligence, and hatred of the Japanese to become good agents were, if they were accepted, sent to the OSS secret, secluded training facilities on Catalina Island, which was declared off limits for tourists for the duration. Captain Robert E. Carter, 26, from Alexandria, Virginia, who had been with the OSS German OG at Areas F and B in late 1943 but then was reassigned for advanced training in Britain, was put in charge of two of the Catalina training camps, Howland’s Landing and 4th of July Cove, in 1945 (the main training camp was at Toyon Cove, 3 miles from the town of Avalon, but it and its Korean trainees were kept strictly separated from the other two). Under the supervision of Major Vincent Curl, Colonel Carl Eifler’s representative, Carter and his staff spent seven months providing intensive training and field exercises for nearly two dozen Koreans, who Carter described as “dedicated, serious students.”

Note that the first group of Koreans was kept “strictly separate” from the Koreans at other training camps on Catalina Island. This was not done by accident. Note also that Chambers does not provide any details about the secluded group.

Another important clue is that Toyon Cove is listed as “the main training camp”. Logic would suggest that this camp is a good candidate for Area W-A, where Moon Duck Harr was trained.

Additionally, earlier in the report Chambers writes that the POWs were mostly young men who had been conscripted into the Japanese army.

Chambers cites as one of his key sources for this section of his article a book called The Deadliest Colonel, written by Tom Moon and Carl Eifler, the latter of whom oversaw the training.

When the book went to press in 1974, the CIA prepared an internal information sheet, declassified in 2011, on Eifler’s background and the book’s claims.

Someone at the agency underlined three facts that stood out above the rest:

1. Eifler accepted an assignment to assassinate Nationalist Chinese leader Chiang Kai-shek during WWII, which would have allowed the U.S. to back Mao Tse-tung; the operation was called off by the plan’s originator

2. Eifler was assigned to kidnap Germany’s top nuclear scientist; this too was called off at the last minute

3. Eifler was assigned to “plan a major revolution in Korea”

Notice that while Eifler’s assignment was to plan a revolution in Korea, most of his “Korean agents” were aimed at Japan and some actually ended up infiltrating the United States mainland, where “none were ever detected or caught”. Keep this in mind as we return to the Chambers report.

A different group, non-English speaking Koreans, but probably Korean exiles approved by Syngman Rhee, rather than the POWs, underwent paramilitary training at Area C-1 in Prince William Forest Park [Virginia] in the summer of 1945. Many of these men later went on to become high officials in Rhee’s postwar government of South Korea.

Chambers points out that this third group of Koreans did not speak English, which suggests that some portion of the Catalina Island “Koreans” did speak English.

I doubt that few, if any, POW conscripts spoke English.

Taken at face value, the Korean section of Chambers’ report does not make logical sense.

Syngman Rhee held three degrees from American universities. He was no fool. Why would he want the high ranks of his government-in-waiting to be filled with paramilitary types? And why wouldn’t the POWs on Catalina Island, men with military backgrounds who were selected in part due to their hatred of the Japanese, receive paramilitary training?

It makes no sense to have the “dedicated, serious students” be the ones primed for battle with the Japanese while the paramilitaries prepared to run a country.

Colonel Carl Eifler claimed that his ten units of OSS trained Korean POWs turned SOs, were ready to leave their Catalina Island training camps, led by himself and thirty American OSS officers in August, and head for Japan itself. However, the mission was cancelled when the Japanese surrendered.

According to Chambers, around two dozen Korean POWs were trained over seven months at Catalina Island’s main camp. That would make them a rounding error of the total group supposedly destined for Japan.

These men were led by 30 American OSS officers who now appear on the scene for the first time. Chambers does not elaborate on whether any of these Americans were recently trained, perhaps on Catalina Island.

However, earlier in the report, he states that Col. Eifler trained “teams of Korean and American saboteurs to be sent into Korea and Japan itself” (emphasis added).

A third man emerges

The Parks Service report was published in 2008. That same year a huge tranche of OSS personnel files was declassified and released to the public.

Chambers’ summary article was published in 2009, and it expanded on the “missions” embarked on by the trainees who landed on the coast in Orange County, California (emphasis added).

Advanced SI [Secret Intelligence] students, accompanied by radio operators, had to infiltrate northern Mexico and obtain and relay important information. Advanced SO [Special Operations] men were sent on survival problems, dispatched into desolate areas with only a minimum of food, forced to live on fish they could catch or game they could shoot. Subsequently they were tested on preparing effective plans to sabotage military facilities in San Pedro harbor and the Orange County coast. Lt. Hugh Tovar, SI, a Harvard ROTC graduate, was one of those OSS trainees in the interior of rugged, windswept Santa Catalina Island in 1945. “They gave me a carbine with one bullet and told me to survive on my own out there for several days,” he recalled. He did and went on afterward to China and Indochina.

When the files opened up, Chambers appears to have gone in search of a living person who had been a member of one of the Korean cohorts trained on Catalina Island in 1945.

The person he found was not Korean and did not have his mission cancelled due to the Japanese surrender. Indeed, William Sanford White¹ confirmed that:

From Toyon Bay [the secluded group] the trainees went to Burma and China where they did behind-the-lines intelligence work for British and Indian armies fighting to recapture Burma.

They recruited, trained and led a guerrilla force of native Kachin fighters whose exploits have never been equaled in the history of guerrilla warfare, and who were largely responsible for the opening of the Burma Road to China, and retaking of Northern Burma.

Bernardo Hugh Tovar was born on December 27, 1922, in Bogota, Columbia and moved to the United States as a child. He would go on to a nearly three-decade long career in the CIA, during which he served as station chief in Malaysia, Indonesia, Laos, and Thailand.

Tovar’s war-time record bears a striking resemblance to that of Karl Harr (emphasis added).

His ROTC class was called to active duty in June 1943, though he was awarded a degree in 1944 because he passed his comprehensive exams. He graduated as a second lieutenant. After various assignments and training in Infantry, Artillery and Communications, he was selected to join the Office of Strategic Services (OSS) in 1945. He was flown to Kunming, China, and was active in the search for prisoners of war held by the Japanese forces in various parts of Asia. In late 1945 he parachuted into Vientiane in French Indochina to try and keep the French from re-entering the region (unsuccessfully). In September the OSS was disbanded, and he returned to Illinois, taking up graduate studies in political science and economics at Northwestern.

Tovar clarified the particulars in a 2007 oral history interview.

I went to Harvard and I went into the Army before I graduated, but I graduated after I was in the Army. A little hard to explain. I was in the ROTC. And I was eventually commissioned at Fort Sill as a second lieutenant for field artillery. And let's see, then I was in the Army, you know, in the artillery for a couple years and then OSS reached out and grabbed me and brought me into the field there…They gave me a little more training and sent me off to China. And from China, because I spoke French, I spoke schoolboy French, they sent me down to Indochina in a parachute and I ended up in Vientiane in late '45 and got my first direct involvement in what became later on the Indochina [Vietnam] War.

Unlike Harr, Tovar did not finish his Master’s degree.

Claiming that he was fed up with his department head, Tovar who had been receiving feelers from the newly minted CIA, left school, moved to Washington, D.C., and returned to the intelligence world.

After retiring from the Agency, Tovar published numerous books and articles on a wide range of intelligence and covert operations topics.

What I think happened on Catalina Island

The pieces of the OSS West Coast training of Korean agents, as assembled by Dr. John Chambers, do not fit. I think he has collected many of the main pieces, though, and if we rearrange them, the picture begins to make sense.

During the summer of 1945, up to 50 Ivy League-educated Americans were given Korean pseudonyms and secluded at the main training camp at Toyon Cove. These “dedicated, serious students” were secretly being prepared for missions in the China-Burma-India theater. Simultaneously, they were evaluated for further grooming in preparation for holding high government office, not in Korea, but rather in the U.S.

Among these students, the Special Operations group was placed on a track leading to positions inside the United States, while the Secret Intelligence group was prepped for roles on the periphery of the empire.

Meanwhile, a separate cover operation was held over a period of months in other training camps on the island. This operation provided non-English speaking Korean POWs with paramilitary training in preparation for an assault on Japan that would never take place.

The training of future Rhee-government officials in Virginia is likely unrelated. However, one can see why the OSS would want to train future officials in a friendly foreign government close to the halls of power in Washington, D.C., while keeping American agents prepping to run its own government at a geographical arm’s length.

I want to make clear that the above is not an accusation or a supposition. It is a hypothesis; one I am preparing to test.

Meanwhile, I will leave you with one final tidbit.

Although Hugh Tovar would spend significant chunks of his career working overseas on behalf of the Agency, he did maintain a residence in the U.S.

That residence was a short walk from Jay Reid’s house on Blacklock, Rd. in Bethesda, Maryland.

1. White, William Sanford, Santa Catalina Island Goes To WAR: World War II 1941-1945. White Limited Editions, 2002.