A Cryptologic Veteran’s Analysis of Day of Deceit

Reviewed by Philip H. Jacobsen, LCDR USN (Ret)

The author, Robert B. Stinnett, made a thorough search of National Archives files other repositories and contacted numerous personnel to justify his long held belief that President Franklin D. Roosevelt not only actively fomented war with Japan as a pretext to aid Britain in its fight with Hitler but that he purposely made Pearl Harbor an attractive target for the Japanese Navy.

Then (as the theory goes) after learning of the of the Japanese plan to attack Pearl Harbor, Roosevelt (through conspiracies continuing today) not only kept Admiral Kimmel and General Short from obtaining information on Japanese intentions to attack Pearl Harbor but ordered or had ordered actions that prevented those commanders from discovering the KIDO BUTAI and adequately defending Pearl Harbor from the expected attack by the Japanese.

Day of Deceit argues that Roosevelt was convinced the loss at Pearl Harbor must be of sufficient magnitude to overcome the isolationist views of the general public so that he could safely declare war on both Japan and Germany. Furthermore, after the attack on Pearl Harbor, Roosevelt through his co-conspirators (who apparently include General Marshall, Admirals Stark, Ingersoll, & Anderson, Captain Turner, and Commander McCollum and by implication Admiral Noyes, Captain Redman, Commander Rochefort, and many others), attempted to cover up his and his co-conspirators’ dastardly deeds. However, through Stinnett’s foresight, expertise and diligence, he was able to see through this monstrous conspiracy and its cover-up to reveal its details to us some 58 years later when all previous efforts by revisionist conspiracy theorists have failed and all the participants are dead and cannot defend themselves. Nevertheless, this book will sell well among rabid Roosevelt haters, many Kimmel and Short supporters, and dedicated conspiracy theorists.

Beginning in World War I with the nascent United States involvement in cryptology, followed by the Yardley Black Chamber, the book rapidly progresses into the beginnings of the Army involvement in cryptology after Secretary of State Stimson closed the Yardley Black Chamber. Mr. William Friedman, who headed the Army’s Signal Intelligence Service since 1929, in 1930 hired three mathematician assistants, Mr. Frank Rowlett, Mr. Solomon Kullback, and Mr. Abraham Sinkov.1 The four individuals functioned in a small secure space in the third wing of the old Munitions Building on Constitution Avenue, Washington, D.C. until the beginning of WW II.

In an effort to support his conspiracy theory, Stinnett came up with many new documents not generally known to be available. However, these documents do not add anything new to the question of who knew what and when. In his zeal, he misinterprets not only some of these “new” documents but comes up with radically new meanings for the plain words and characterizations of well accepted documentation already available in this Pearl Harbor arena. One of the centerpieces of his argument is an October 1940 memorandum by then Lieutenant Commander McCollum of ONI in response to the September 1940 signing of the Tripartite Pact by Germany, Italy and Japan and not as any blueprint for initiating war with Germany and Japan. McCollum recognized the danger to the western powers if Japan was able to connect up with Germany and Italy through Asia and suggested eight actions designed to contain Japan generally and to keep her from making such connection with its other Axis partners. Unfortunately, the book seizes on an off hand comment that is not one of the main points of the memo as the springboard for its conspiracy theory. That comment was if the eight proposed actions designed to contain Japan should by chance cause Japan to commit an overt act of war, so much the better. No proof of any official implementation of this mid-level memo is provided. Furthermore, Stinnett improperly ascribes McCollum’s office as “an element of Station US (by which he means OP-20-G), a secret American cryptographic center located at the main naval headquarters” in an effort to tie McCollum closer to OP-20-G than he actually was before WWII. A non-cryptologic fallacy of the book is the fact that Roosevelt had no assurance that Germany would declare war on the U.S. if the Japanese did attack Pearl Harbor thus negating any reasonable conspiratorial design to get the U.S. into war with Germany by forcing Japan to attack the U.S.

It is well established that the SRN series of Japanese naval messages in the National Archives were decrypted in 1945-46 and translated in 1946-47, but Stinnett incorrectly suggests they may only have been transcribed at those times and that these decrypts (or at least some of them) were available not only in radio intelligence centers in Washington, but Stations Hypo (Rochefort) in Hawaii and Cast on Corregidor. Among other things, the book misinterprets an article by Captain Pelletier in the CRYPTOLOG. Even though Pelletier is now dead, he also wrote in the NCVA History Book that all such JN-25B raw messages were two months old by the time he saw them in Washington and that no KIDO BUTAI transmissions while en route from the Kuriles to Hawaii were ever found before or after 7 December 1941. Further, the book fails to inform its readers that Rochefort and his Hypo personnel were only assigned to and only worked on the unproductive Flag Officer’s Code and not the main Japanese Fleet Code JN-25B as well as the fact that they were only given the go ahead to work on JN-25B a few days or so after the Pearl Harbor attack. As mentioned before, Stinnett also omits the well known information that JN-25B intercepts from Corregidor, Guam, and Station H were only forwarded to Washington by mail and took up to two months to arrive mostly by ship and rail. Thus, even Washington’s alleged 10 percent capability on JN-25B decrypts had not even begun to be applied to the November and December 1941 intercepts en route there while Stinnett maintains they were available to all commanders—except, of course, Kimmel and Short due to FDR’s co-conspirators.

The book implies more improprieties by the fact that Hypo had no assigned Japanese diplomatic intercept or decrypt authority until RCA President Sarnoff made available RCA cables from Honolulu beginning in early December 1941. Part of Stinnett’s overall conspiracy theory includes the allegation that Hypo only decrypted the administrative messages of these low level Japanese diplomatic messages provided by RCA before Pearl Harbor and did not decrypt the “bomb plot” messages until after Pearl Harbor.

Although Stinnett obtained definite information from Captain Whitlock that no significant JN-25B decrypts were made by Station Cast on Corregidor during the period in question, he disputes this fact and misinterprets other documents and sources as proof that Whitlock is wrong. Some Navy cryptologic veterans involved in this book have complained Stinnett gained their confidence by agreeing to tell their stories but ignored their version of events in favor of the monstrous conspiracy theory finalized in the book. Admiral Layton terminated his interview with the author, most likely when he learned where the book was going. It should be noted that it took OP-20-G some 14 months to read the much simpler JN-25A system that was used from 1 June 1939 to 1 December 1940. The book misleads its readers by not revealing there were two distinct codes, the earlier JN-25A and its much more complicated successor JN-25B used during the period in question and refers to them collectively as “Code Book D” or “5-Num code.” Thus, the final successes of JN-25A are improperly imputed to JN-25B which was not read to any significant extent until March 1942 when the first published decrypt is found. The ever-increasing requirements to provide Japanese diplomatic decrypts and translations during 1941 took most of the time of Navy cryptographers so that few people at both Washington and Station Cast were assigned to work on the new version of the Fleet Code, JN-25B. In addition, JN-25B used about eight additive cipher books up through 4 December 1941 further delaying the effort to read any significant amount of this new and far more complicated code and cipher combination.

Stinnett and his sources are apparently not aware that Japanese naval shore broadcast stations transmitted simultaneously on a number of frequencies covering their communications area and it was up to the ships in their communications zone (or U.S. intercept operators) to choose the best frequency on which to copy such broadcast. Thus, the deduction that because an intercept operator copied one message in the 12 MHz. range part of one day and 16 MHz. on part of a different later day means the ship or force has moved further away from the shore station is patently incorrect. These Tokyo broadcast transmitters were active on several of their assigned frequencies simultaneously and the 16 MHz. frequency had long been used by the Tokyo broadcast as a daytime frequency.

Stinnett often claims carriers or fleet units must have transmitted on high frequencies when they are only seen in the headings of messages on fleet broadcasts. He does not tell his readers that many ships are tied up at docks and have landline or cable communications available to them so they do not have to use radio and the original transmissions of such messages will never be heard by foreign intercept operators. In this regard, he maintains that Admiral Yamamoto’s messages sent (while tied up at a Kure dock) to the Pearl Harbor attack force and other ships on the Tokyo broadcast violated radio silence when, in fact, the radio silence imposed then only meant that ships (or aircraft) are not permitted to transmit by high frequency radio, not that messages to these units cannot be sent by fleet broadcasts or that fleet units or commands that have landline, cable, or other approved facilities available to them cannot use them.

Apparently, Stinnett did come up with records to substantiate Hypo’s summaries about the carrier AKAGI being active on the air on 26 and 30 November 1941. However, there is no documentation that any high frequency direction finder (HFDF) fixes were available to Hypo on such transmissions. The single line bearings reportedly obtained by Corregidor’s old DY-2 HFDF went by the island of Honshu as well as the Kurile Islands and the former location with acceptable HFDF variations was within Hypo’s previous general determination of carrier locations. According to the book, a possible cross bearing from Dutch Harbor was found in that station's November monthly report that did not reach Station H until after 7 December and for some reason was not reproduced in the book. No documentary evidence was shown that such bearing was actually transmitted to Station H or subsequently forwarded to Rochefort at Hypo except a general statement as to routine forwarding by a Dutch Harbor operator.

Although the book claims more carrier and carrier commander transmissions were made after 26 and 30 November, this information is apparently due to a misinterpretation of the TESTM reports from Corregidor to Station H and a misunderstanding of traffic analysis procedures identifying call signs appearing in broadcast and point to point messages sent by shore communication stations. The single TESTM report provided in the book first lists the transmissions heard and their bearings and such bearings are mainly on unidentified call signs. Then, any fleet level call signs identifications made from the traffic analysis of message headings in shore station transmissions by Station Cast are given. In his enthusiasm to support its conspiracy theory, Stinnett apparently assumes that the latter call sign identifications by traffic analysis of shore station transmissions actually represent high frequency radio transmissions by such fleet units and commanders. Layton, Pelletier, and Whitlock, among others, deny such transmissions were ever received. One wonders why Stinnett did not reproduce the other two TESTM reports upon which he relies to make his specific allegations to clarify his identification and deductive processes, especially since the one page reproduced does not support his allegations.

Gross misinterpretations of two decrypts and translations in the SRN series at the National Archives make up the other parts of the book’s centerpiece of its conspiracy theory. In an effort to give some credence to its allegation of a massive conspiracy, the book contradicts the plain meaning on the face of translations of these two decrypted messages, established Japanese naval communications practice, and standard decryption procedures. These messages were reported on long ago by Frederick D. Parker in Cryptologia, Vol. 15 (4) p. 295. However, Parker fully reported that JN-25B was being decrypted at best on a 10 percent basis in Washington and those November and December 1941 raw messages discussed were en route to Washington D.C. so that they were not available to be worked on until long after the Pearl Harbor attack. The glaring omission in the book of this vital “unavailability” information is instructive.

The first decrypt refers to naval spy Suzuki who was sent to the First Air Fleet on business to be picked up on 23 or 24 November at Hitokappu Wan (Bay). It is abundantly clear from the document that Hitokappu Wan is spelled out letter-by-letter in five numeral code groups of JN-25B because there was no two- or three-letter coded geographic designation available for this remote location (like AF for Midway Island.) Nevertheless, the book baldly claims, without any substantiation, that the words Hitokappu Wan were sent in plain language while the rest of the message was sent in code, an incredible absurdity. No other examples of plain language inserts within a high level Japanese naval coded message were ever claimed or reported. No one else has had the temerity to make such a ridiculous assertion when confronted with the JN-25B code designation on the face of the decrypt and no reference to a plain language insert in the decrypt.

The second gross misinterpretation contained in the book is that Yamamoto’s famous message of 2 December 1941 only referred to as “Climb Mount. Niitaka 1208” may have been sent in plain language. If so, it implies Rochefort knew of these two plain language “busts” by the Japanese and therefore is part of the conspiracy for not reporting them in his summaries. For this strong implication, one Japanese historian is cited saying the message was sent in the clear while Yamamoto’s biographer is identified as saying the message was encoded in a five numeral code (JN-25B). Captain Pelletier in the Naval Cryptologic Veterans Association History Book confirmed this message was sent in JN-25. To show the extreme lengths the book will go to conjure up his implication of conspiracy, it omits the fact in the narrative that this message labeled SRN 115376 by the National Archives had a cryptographer’s reference below the heading clearly showing that it was encoded in JN-25B. Furthermore, Stinnett does not clearly point out to his readers that “Climb Mount Niitaka” was prefaced by the words,

“This dispatch is Top Secret. This order is effective at 1730 on 2 December #10.”

Can you imagine the Japanese sending a Top Secret message in the clear and depending on a transparent underlying meaning for security? Except for battle tactical reports during the war, the Japanese seldom used plain language and even then preferred tactical codes. These are only a small part of the omissions, errors and misinterpretations contained in the book to try to make its revisionist conspiracy theory seem plausible to the uninitiated.

The book also resurrects the old allegations of Robert D. Ogg, a seaman in the 12th Naval District Intelligence office, and disregards Ogg’s recorded interview by then Commander Newman that he only plotted two very closely parallel bearings from California stations 100 miles apart. Stinnett now says Ogg had prewar information on Japanese warship transmissions in the Kuriles with HFDF bearings by Dutch Harbor in spite of Ogg’s original transcript to the contrary.

The old and thoroughly repudiated hearsay report of dead Dutch codebreakers’ prewar determinations that Japanese carriers were in the North Pacific en route to Hawaii are regurgitated by the book. Only now it has the Dutch putting them in the Kuriles instead of the North Pacific. Stinnett also repeats Parker’s reporting of the tanker SHIRIYA moving eastward from the Bonin Islands in a 1 December 1941 message (SRN 115398) to Destroyer Division 7 with the KIDO BUTAI that was intercepted on the Tokyo broadcast. Again, he does not tell his readers that this JN-25B message was only decrypted in 1945-46 and translated in 1946-47 and that the raw intercept was en route to Washington DC in the U.S. postal system on 7 December 1941.

To further its revisionist conspiracy theory, the book argues that government censors are still withholding the publication of decryptions (and translations) of hundreds of vital Japanese naval messages whose secrecy is a part of this monstrous conspiracy. Stinnett points to missing Station Message Serial (SMS) numbers and missing versions of original transmissions by fleet units and commanders (supposedly on high frequency radio) that appear on shore station fleet broadcasts to naval ships and point to point circuits. However, the book does not mention that after the war, Navy analysts discovered that about 7,000 Japanese naval messages per month were forwarded to Washington from Corregidor, Guam, and Hawaii from July to December 1941. During the expanded intercept coverage of WWII, an OP-20-G official estimated that the U.S. intercepted 60 percent of Japanese naval traffic. Therefore, far more than 10,000 messages were probably sent over the airways by the Japanese Navy per month in the months before Pearl Harbor and less than 60 percent were actually intercepted. Thus, the missing SMS numbers and original transmissions could be accounted for by missed intercepts and transmissions originated by landline, cable, visual means or even hand-carried to shore radio stations. In fact, there was a cable office at Hitokappu Wan available to fleet units to send messages to Tokyo without transmitting on high frequency radio.

Again, in 1945-46 analysts decrypted those intercepts from the Pacific that were available in Washington. A total of 26,581 messages in seven different crypto systems were intercepted between 5 September and 4 December 1941. Between 15 March 1946 to 20 August 1947, OP-20-G analysts and linguists from ONI undertook the study of these 26,581 post war decrypts and only 2,413 were considered important enough for full translations. Of these, only 188 were isolated as pertaining specifically to the events of 7 December 1941. This information contradicts Stinnett’s assertion that government censors are withholding disclosure of hundreds of vital decrypted and translated messages in furtherance of the alleged conspiracy by President Roosevelt and many top and middle level government officials. Those 2,413 messages that were translated in this period are available in the SRN series and no other decrypts or translations are available for this period of time.

To his credit, Stinnett does recognize that the Winds Execute message (a favorite revisionist conspiracy allegation) was never sent. He also recounts Secretary of War Stimson’s blatant attempt to reverse the Army Board of Inquiry’s determination that Marshall was in dereliction of his duty as to his Pearl Harbor actions. Stimson sent attorney Clausen around the world to obtain new affidavits countering the witnesses’ previous testimony of Marshall’s neglect to act on Purple decrypts. However, Stinnett omits the fact that Clausen also tried to place the blame for not fully informing Hawaiian commanders on Navy cryptologic officers. The latter effort is also part of the aim of this book, but its shot is far wide of the mark.

To those of us who are familiar with Japanese naval codes and communications procedures at the time, available documentation in the Pearl Harbor arena as well as the pertinent personnel and history of OP-20-G, it is abundantly clear that the book fails to prove any part of its massive revisionist conspiracy theory. In fact, the expansion of prior revisionist conspiracy theories to include so many new allegations of wrong-doing by Roosevelt and his mid and high level co-conspirators plus a continuing cover-up makes its enormous conspiracy theory a complete impossibility.

In conclusion, it is still clear that no U.S. official knew beforehand of the Japanese plans to attack Pearl Harbor or discovered that the KIDO BUTAI was on its way to Hawaii for such an attack in spite of this latest in a series of revisionist conspiracy theory books.

Afterward

A few of the more gross errors noted after researching the actual archives documents:

Further research confirms reports by numerous high ranking Japanese officials who participated in the Hawaiian Strike Force that programs of Japanese radio deception activities were carried out from major naval bases of Sasebo, Kure and Yokosuka to deceive the U.S. Navy's radio intelligence organization that the carriers were still in home waters. The actual Station C Corregidor TESTM HFDF bearing reports from 13 November through 4 December 1941 show the AKAGI’s bearings remaining between 026° and 030° even though the KIDO BUTAI first transited from the Inland Sea to Hitokappu Bay in the Kuriles and thence across the North Pacific Ocean to Hawaii. If the Corregidor bearings on the radio deception transmissions using the AKAGI’s callsign for 27 and 30 November and 4 December had been valid, they would have been 041°, 048°, and 051° instead of remaining between 026° to 030°. The reports of “carrier” transmissions on 26 November and the AKAGI on 30 November intercepted by Station H, Heeia, Oahu, Hawaii [but without HFDF bearings] were obviously part of the same Japanese radio deception program emanating from Sasebo (027° from Corregidor) and Kure (030° from Corregidor).

Official OP-20-GYP-1 reports verify that zero decrypts of JN-25B were made prior to the Japanese attack on Pearl Harbor. All the early JN-25B decrypts are listed in numerical order with Station Hypo, Pearl Harbor making the first decrypt in January 1942.2  In addition, Commander Rudolph Fabian, the Officer-in-Charge of Station C Corregidor, testified before a Congressional committee about breaking JN-25B before the war.

“We were in the initial stages, sir. We had established liaison with the British unit at Singapore. We were exchanging values both code and cipher, but we had not developed either to the point where we could read enemy intercepts.”

Stinnett dismisses Captain Whitlock’s confirmation of this no decrypt testimony. The fact that Station C located the AKAGI off Corregidor on 8 December 1941 based direction finder bearings on more radio deception activity is further evidence that they had not broken JN-25B and relied only on traffic analysis and direction finder bearings for their reports.

Stinnett blatantly misconstrued Station H’s COMINT Summary of 25 November 1941 that reported that Vice Admiral Inoue, CinC Fourth Fleet in the mandated islands, was observed in “extensive communications” with many entities like Commander Submarines, Commander Carriers, Juluit, and other mandate island bases as evidence that Nagumo violated radio silence and that U.S. Navy stations obtained bearings and fixes on such phantom radio transmissions. However, it was Inoue who was “observed” in the extensive communications not Nagumo. In the parlance of the times, the word “observed” meant these communications noted were only in the form of addressees of messages seen mostly on the Tokyo Fleet broadcast and not original radio transmissions as Stinnett alleges. Had Rochefort intended to describe extensive communications of Nagumo such an entry would have been under the heading of Combined Fleet and he would have specified “heard transmitting” instead of “observed.” Thus, Stinnett completely turns the summary upside down to support his predetermined conspiracy agenda. Stinnett also erroneous states that this COMINT Summary of 25 November covers the Japanese naval activity of 26 November 1941 when Nagumo departed Hitokappu Bay due to the time difference of the International Date Line. However, all U.S. Naval radio intelligence logs, messages, supervisor’s reports, and COMINT Summaries covering Japanese naval activities used the Tokyo [-9] time zone of their target to avoid confusion. Thus, this summary reported Japanese naval activity for 25 November when Nagumo was at Hitokappu Bay not 26 November 1941 as Stinnett claims.

Stinnett also claims some 129 violations of radio silence during a 21-day period which he implies is from mid-November on. The figure of 60 actual radio transmissions by Admiral Nagumo being intercepted is ridiculous. Only a few were seen on the Tokyo broadcast and were not original radio transmissions. The messages were undoubtedly filed while in port by messenger, blinker or landline. None of these alleged transmissions by Nagumo were during his transit from Hitokappu Bay to Hawaii. The same for the 40 messages allegedly sent by radio by KIDO BUTAI carrier commanders and units. Stinnett even makes the absurd claim that 25 messages sent on the Tokyo fleet broadcast by Yamamoto and other commands and ships were violations of radio silence. The reason for using the broadcast method of transmission from shore stations is to maintain radio silence by not requiring ships and commands to use their transmitters to receipt for messages.

Footnotes

1According to Mr. Rowlett in his book, The Story of Magic, it was called Signal Intelligence Section. The Section was part of the Army Signal Corps, and the immediate boss was actually Major Crawford, USA, who reported directly to the Chief Signal Officer.

2See Stephen Budiansky’s article, Too Late For Pearl Harbor, in the December 1999 issue of U.S. Naval Institute Proceedings and my article, Foreknowledge of Pearl Harbor? No!: The Story of the U.s. Navy's Efforts on JN-25B.