Tuesday, June 11, 2019

It's Oh-Dark-Thirty, All Right

     Today, I learned that if the Japanese military had been in the habit of reading the Chicago Tribune, WW II might've run a lot longer.

     You see, in June 1942, a Trib reporter spilled the beans about the U. S. reading encrypted IJN radio traffic, thanks to a buddy who was XO on the USS Lexington.  It's still unclear exactly how the Tribune's Stanley Johnston got the information from Commander Morton Seligman.  Seligman, an otherwise exemplary officer who had been headed for great things, found his Naval career stopped cold; Johnston narrowly avoided being brought up on espionage charges and the whole thing was swept under the rug, in hopes the Japanese military wouldn't notice.

     Given the complete lack of love between the Tribune's Colonel McCormick and President Roosevelt dating back long before the war, FDR probably regretted having to officially ignore the incident.

     What do you know, Imperial Japan wasn't reading the Chicago papers.  They never found out we were reading their mail, or at least a fair amount of it.  They might not have believed it; feeling secure behind a difficult language for outsiders to learn and an encryption scheme that added complications to a mechanical system similar to Germany's Enigma, they apparently never considered any risk to their secure communications other than losing code books to the Allies.

     To this day, the Tribune considers their report quite a wonderful scoop; but it could have delayed the war's end by a couple of years if Japan had realized we were reading over their shoulder the whole time.

     Commander (on retirement, Captain) Seligman served ashore until 1944, and never uttered a word of complaint or explanation about it for the rest of his life.


Comrade Misfit said...

According to an article in Proceedings, decades ago, Johnson was sharing a stateroom with Seligman. It is quite possible that Seligman left some papers on his desk and Johnson did what reporters will do.

Merle said...

Johnson should have been hung - and Seligman should have been court martialed......

Anonymous said...

Fishy. The XO of a carrier doesn't have a roommate. Lexington didn't arrive in the war till a year after Battle of midway.

Roberta X said...

Anon, it was USS Lexington (CV-2), not USS Lexington (CV-16). The older Lexington was at the bottom of the Pacific by the time of the Battle of Midway.

The accounts I have read aren't completely clear to me, but Johnston had acted heroically aboard the older Lexington during the Battle of the Coral Sea. Seligman was XO of that carrier. After the battle, they were both being transported aboard some other ship, headed for Hawaii or the U.S. mainland and it appears that was when the leak occurred. The precise manner of the leak has never been made clear. Johnston was not, at that time, an officially accredited reporter and thus had not promised to maintain secrecy of any uncleared material that might have come his way. You can make up your own mind about how he found out.

Roberta X said...

"Johnson should have been hung - and Seligman should have been court martialed......"

Thereby confirming to any enemy spy who might have been watching that Japanese codes had been compromised. Say our spy's working for Germany; it's a short step from learn IJN codes aren't as secure as they think to wondering if Enigma is as uncrackable as believed, at which point they get a lot more serious about crypto and suddenly we have a lot more U-boat trouble and the War runs even longer. If Purple decrypts shortened the war by two years, what did reading German comms save?

Merle said...

Nt if it was done in secrecy......

Matt said...

Ala WEB Griffin, put the offenders in St Elizabeth's mental health hospital in DC for the duration of the war. No habeus corpus if FDR said so. Problem solved.

Roberta X said...

"...put the offenders in [a] mental health hospital [...] for the duration of the war..."

Because that's totally not a violation of the clear intent of the First, Fourth and Fifth Amendments, hey?

Just because some past President, Congress or court did a thing, that doesn't make it right, moral or even legal. At one time, a plurality of 'em owned slaves and were totally okay with women not voting or owning property in their own names. That stuff's still wrong and it was even when it was legal. It's not even okay "becuase there's a war on." Better read over your copy of the Constitution; it's real short of secret wartime codicils and exclusions. The people who wrote it were rightfully wary of such things.

The Feds did the pragmatic thing and the Constitutional thing in the case I blogged about. The only sensible course was to shut it up and hope the Axis powers hadn't noticed. The (probable) source of the leak paid a heavy price; the reporter got what he was due, which wasn't much in the way of punishment because he hadn't promised to keep his mouth shut. That was the Navy's screw-up, and Seligman's, too, though he might not have known.

markm said...

"If Purple decrypts shortened the war by two years, what did reading German comms save?"

Probably somewhat less than that.

First, no decryption helps with information that wasn't sent in a way that could be intercepted. E.g., the fleet that attacked Pearl Harbor not only kept complete radio silence, but during the preparations and training in a northern Japanese harbor, they didn't even use the landline telephones. Couriers shuttled back and forth from Tokyo to that harbor, carrying papers in locked briefcases. Our cryptographers could tell us nothing about that plan because there were no transmissions to intercept. Likewise, the Germans kept the plans for the 1944 Ardennes Offensive off the air and caught us by surprise in what became the Battle of the Bulge.

The Japanese put complete trust in their Enigma-style coding machines, aside from occasional concerns that a machine and codebook might have been captured. They used the same codes across an entire department or command, and went six months or more without changing them, each time causing a gap of a few weeks in the decodes. (This was usually just before a major operation - but sometimes so close to it that we had already decoded the battle plan. And so although we couldn't read the messages flying around the Japanese fleet at Midway, counter-plans made weeks before had two US carriers waiting in ambush and the Yorktown hastening to join them. I think that one battle accounted for most of the two years saved in the Pacific war.)

By comparison, the Germans generally trusted their coding machines, but made decoding much more complex out of sheer Germanic OCD thoroughness. They added more rotors and a plug-board to the basic Enigma design. They had every network use a different set of machine settings, and often differently-wired rotors, so there were many codes to be cracked at the same time. And they changed codes often. The result was that decryption was often a month behind intercepting the messages, and it was rare for something like the position of a U-boat, the target of a bomber raid, or the time and location of a counter-offensive to come through before the info was out-dated.

Another factor: German message traffic was much heavier, but most of it went by landlines and could not be intercepted. It was also much easier to call Rommel (for instance) back from Africa to Berlin for a face-to-face meeting than to have a Japanese commander come back to Tokyo from Burma or the Solomon Islands. So the Japanese often sent out the complete plan by radio, while the Germans used land-lines and couriers to send the plan, and only sent updates by radio. Even if we'd had perfect up-to-the-minute decryption, we'd get disjoint pieces of the whole picture; e.g., we'd see orders to U-boats at sea but have to guess at the underlying plan to attack a convoy, or see last-minute changes to the Afrika Corps orders without knowing what the original orders were.