Pearl Harbor: Final Judgement

4 WHO WAS TELLING THE TRUTH?

To explain how I approached my investigation, let me tell a story about the professor of law who began his class by writing the numbers 2 and 4 on the blackboard and asking: “What’s the answer?”

Some students shouted “six.” Others said “four.” Some said “eight.” Whereupon the professor shook his head with a look of sorrow. “None of you have given me the answer, which is: What is the problem?”

In other words, if I was going to proceed as an independent prosecutor and find out what had really gone wrong at Pearl Harbor, I’d damn well better focus on the hole in the doughnut. I suspected what the problem might be, but I still wasn’t sure as to its scope, so I began rereading the hearings of the Army Board. Then I began mixing in the unknown quantity—the Magic intercepts, which had been denied the Board until the end of its investigation.

This resulted in a three-page memorandum with twenty-five questions that needed answering with a mixture of flexibility and discretion. But what the reader should know is that these were my “cover” questions that I spoke of earlier. Knowing how the Army, and the Navy, really works, I submitted the memo to General Cramer; within twenty-four hours, I had on my desk, a “Memorandum for Major Henry C. Clausen, JAGD” that outlined the type of questions I would ask and pursue at my discretion, only now the questions were those of Cramer, the Judge Advocate General, himself. Needless to say, they gave me the opportunity to lift up the corner of any carpet I might find to see if there was dirt hidden underneath.

But now for the hole in the doughnut. The more I read the record of the Army Board hearings, the more convinced I became that there were four primary questions that needed answering. I went over them with Colonel Hughes in the privacy of our vaultlike office, simplifying them to essentials. The questions were these:

(1) What had Pearl Harbor known about Japanese intentions before the attack?

(2) What had Pearl Harbor done with this information before the attack?

(3) What had Washington known about Japanese intentions before the attack?

(4) What had Washington done with this information before the at tack?

If I could find the answers to these seemingly simple queries, my investigation would be a success. And now that I had my agenda fixed in terms of knowing the road I would travel as an independent prosecutor, I was ready on January 24, 1945, to depose my first witness.

Before I begin my parade of evidence, it is only fair to warn the reader that, as in a real-life trial, evidence frequently comes before the court and jury in unusual, unexpected ways. From the very start, the reader will be presented with conflicting statements. I will try to point out the most glaring examples of such conflict during the course of my narrative. And at the end, in my role as independent prosecutor, I will summarize the case to avoid confusion.

Colonel Hughes came up with the suggestion that I begin with Brig. Gen. Kendall J. Fielder, who at the time of Pearl Harbor was a colonel serving as head of G-2 (Intelligence) in the Army’s Hawaiian Command. At the moment, he was working in the Pentagon. (Hughes had one of the best ears for picking up rumors and knowing who was where, doing what and to whom that I could have asked for.) So I telephoned General Fielder, whom I recalled from watching him testify before the Army Board as being a rather slick, fast-talking public-relations type of fellow, and asked if he’d be willing to come to my office in room 4D852 for further discussion.

“I’d be delighted to,” said Fielder. “Would half an hour be soon enough?”

“That would be perfect, General. Thank you.” My goodness, I thought, imagine a general’s being so polite. The grapevine must be working overtime.

Colonel Hughes and I rushed to secure all the Top Secret documents in our office, which was about twenty feet by fourteen, and Hughes then excused himself so that Fielder might feel more comfortable answering questions without a witness present. Why, one might ask, did we go to such lengths to make our office sterile? Well, I had a hunch that General Fielder, despite his having been the top Intelligence Officer for the Army at Pearl Harbor, had never been cleared for Top Secret Magic. What I needed to know was how he could have done his job properly, and kept his commanding general informed about what we knew about the Japanese intentions before the attack, without having had access to this type of information.

I couldn’t let Fielder know that this was one of the major points of my interrogation. That isn’t how a good lawyer works. First he needs to know everything he can before he starts asking questions. Then he has to employ the Socratic method, probing gently here and there, until he has laid the groundwork and can ask the all-important question. You win your cases by studying your books and working papers, and by thinking hard at your desk before you go into the courtroom. That’s when you figure out what you know and what you don’t know, what the answers of the witness might be, what you should ask the witness, and when you should stop asking questions. I was prepared for Fielder.

As it worked out, however, Fielder wasn’t ready to be deposed.

He turned aside many of my questions, asserting that he would have to review his records, which were locked away in Hawaii—and since nearly four years had passed since the attack, he could no longer remember the specific details—and this would entail considerable research on his part. Meanwhile, he addressed certain other points, including these:

Commenting on a message he had sent to Washington on September 6, 1941, asking the G-2 office not to send him any more intelligence bulletins, because they duplicated what the Navy was already distributing, Fielder claimed that he was referring only to counterintelligence matters. He claimed his message had had “nothing to do with combat [intelligence].” Personally, I found his comment odd, because that was not what his original message said.

The message from Fielder contained three paragraphs. The first said that the Summaries of Information sent regularly by Army (G-2) Washington actually originated with Naval intelligence in Pearl Harbor and already had been given Fielder’s office. The second paragraph said that the “cooperation and contact” between the Office of Naval Intelligence, the FBI and Fielder’s own G-2 office at Pearl Harbor “is most complete.” In conclusion, Fielder was telling Washington to stop sending Summaries of Information, “to avoid duplication of effort.” In other words, Fielder was reassuring Army (G-2) Washington that everything was fine with his operation, that he was getting everything he needed to do his work properly from the FBI and Naval intelligence.

I suspected that this message was misleading, if not an outright lie. The Navy had withheld information from the Army Pearl Harbor Board until we went to the mat with them, as I earlier told Secretary Stimson. The question in my mind now was what the Navy had withheld from Fielder that he didn’t know about. In turn, this led to the question as to whether or not Army intelligence at Pearl Harbor had done its job.

My suspicions were even further aroused when Fielder next said “he was not very well acquainted with” Comdr. Joseph J. Rochefort of Naval Intelligence at Pearl Harbor and that he had “maintained no liaison with him.” Also that he was unaware of any liaison arrangements between Commander Rochefort and Col. Edward Raley of the U.S. Army Air Force. (At that time, the Air Force was under Army control.) He also stated that he had no knowledge of the Top Secret Magic material. When I asked Fielder if he had known about the Navy’s intercepting such traffic, he said he understood this was being done, but that “it was talked about in whispers” and only on this basis did he know that we had broken the Japanese codes.

As I took all this down for the record, I could not help wondering at the significance of what Fielder was revealing, not about himself and how he ran his shop, but about the Army’s system for handling these exceedingly important and sensitive matters.

How could he be G-2 for the commanding general at Pearl Harbor when he neither had training for the job nor was cleared for Top Secret Magic information? Some people in the Pentagon referred to this Top Secret Magic as “boogie-woogie.” Now I could see why they called it that, because there was no apparent rhyme or reason in how it was disseminated. Even by January 1945, when I was interviewing Fielder and he had been promoted to brigadier general, he still hadn’t been cleared for Magic. Because of this, I couldn’t show him the decrypts I was using in my investigation. Instead, I had to paraphrase everything and ask him questions such as: “Were you ever aware of such and such a message of this date that said this …?”

Again, Fielder parried my questions by saying he would have to check his records in Hawaii. (The reader may not understand my questions at this point, but they will become clear later on.) Fielder “understood” that his G-2 estimates of October 17 and 25, 1941, had been circulated to the Hawaiian Chief of Staff and to General Short. He also “thought” he had talked to General Short concerning an all-important warning from G-2 Washington of November 27. He claimed he did not see the Navy Intelligence Bulletin of December 1, 1941. He claimed he did not see a warning from the Military Attaché in Melbourne, Australia, of December 5-6, 1941. He stated that General Short had “assumed” that the Navy was conducting long-range reconnaissance by plane to warn Pearl Harbor of a surprise Japanese attack. He also said that he “didn’t remember” a message, numbered 519, from G-2 Washington of December 5, 1941, that instructed the G-2 Hawaiian Department to contact Commander Rochefort immediately. As Fielder put it, the December 5 message “might have come in as routine,” but that didn’t mean he had to see it.

Once again, I suspected Fielder wasn’t being completely honest. As a result, my research into who had seen this message number 519—or “old 519,” as I called it—would take up a considerable portion of my investigation.

And there I left it. Since Fielder didn’t have access to his files and notes, I wasn’t going to get a signed affidavit from this interrogation, but my typewritten memorandum for the files would be incorporated in the final affidavit that I would take later, when I went to Pearl Harbor, making sure that Fielder was there, too. Nevertheless, Fielder had given me information of a very disturbing nature. My investigator’s instincts were aroused, and I was worried.

The next morning, instead of going to the Pentagon, I drove out Connecticut Avenue to the Navy’s code-breaking complex not far from Chevy Chase Circle. There, I took the affidavit of Navy Capt. Joseph J. Rochefort, who had been the Combat Intelligence Officer in charge of the Combat Intelligence Unit at Pearl Harbor. This was actually a field unit of the home office in Washington, and it was attached to the Commandant of the Fourteenth Naval District in Pearl Harbor, not the Pacific Fleet, which was commanded by Admiral Kimmel. (As it turned out, this designation of assignment would be important later on.)

Rochefort gave me no idea of the strain he had been working under. I did not know that he had been recalled to Washington because he had dared to disagree with his masters in Washington about their analysis of the Japanese battle plans before the battle of Midway. It was only when Rochefort and Capt. Edwin T. Layton, the Fleet Intelligence Officer, convinced Adm. Chester Nimitz, the new commander of the Pacific Fleet, as to the true meaning of the Japanese decrypts before Midway that our Navy fliers were able to score a stunning victory that miraculously changed the course of the entire Pacific war. Rochefort’s independent thinking had proven the analysis of the strategic situation by the Office of Naval Intelligence in Washington to be totally wrong. This had infuriated the Washington command, who then punished Rochefort by bringing him back from his all-important combat command in Hawaii to give him an insultingly unimportant assignment. (It was only when Admiral Layton published his book, And I Was There, which made the front page of the Sunday New York Times on November 17, 1985, that Rochefort’s accomplishments were recognized officially. The Navy posthumously awarded him the Distinguished Service Medal, which the President presented to his widow.)

Anyway, Rochefort appeared totally calm and collected to me, the epitome of how an intelligence officer should look and act. He spoke with a benign smile. His words were precise and exact. Thus, I was flabbergasted when Rochefort calmly began to contradict Fielder’s statement to me of the day before.

“My opposite number in the Army at Pearl Harbor was Col. Kendall J. Fielder, G-2, Hawaiian Department,” said Rochefort. “In the fall of 1941, arrangements were made between Col. Fielder and myself for liaison and exchange of intelligence information pertaining to our functions on matters of personal concern.… I had discussions with him and Edwin T. Layton, Fleet Intelligence Officer, at my headquarters. Thereafter, including the period of 7 December 1941, we maintained most cordial and close relations, meeting informally and frequently, and carried out these arrangements.”

Great Scott—who was telling the truth? I reminded myself that both Rochefort and Fielder were intelligence officers, which meant they both knew how to lie convincingly. I wouldn’t make any conclusion now about the conflicting statements, but I promised myself that I’d get to the bottom of it all sooner or later.

As an aside, when I say that I was flabbergasted by what a witness might be telling me under oath, that doesn’t mean that I ever allowed the witness to see my surprise. Good lawyers have poker faces. We hear so many conflicting bits of testimony during the course of a trial that we have to keep a tight check on our emotions. When a wise lawyer appears angry in a courtroom, for example, he’s pretending to be angry. If he isn’t pretending, then he’s lost control and, most likely, his case. So while Rochefort continued his statement, I kept my face blank.

According to Rochefort, his duties from the fall of 1941 up to December 7 did not include the gathering of information or intelligence from Japanese political or diplomatic sources. He knew this was being done in other places by joint units of the Army and Navy. (What Rochefort was really saying was that he was concentrating on another aspect of Magic: breaking the Japanese naval codes for flag officers.) I noticed a flicker of emotion behind Rochefort’s eyes, however, when I showed him the Magic decrypts I had brought with me to the interview. He picked out five of them, saying that he knew their substance before Pearl Harbor, and he wrote his initials on the back of each one.

For reasons of security, when I produced my report back in 1945, I could not attach the Magic decrypts to an individual’s affidavit so the reader could easily understand the secret part of the interrogation. This made it terribly difficult, if not impossible, for the historian. But for the reader of this book, I can say that the full text of the five messages that Rochefort acknowledged having known about before Pearl Harbor can be found in the appendix on pages 322 (SIS 25432), 322–323 (SIS 25392), 335 (SIS 25545), 335–336 (SIS 25787), and 339 (SIS 25640). As for the significance of these messages, let me explain briefly.

The first two messages constituted the basis for the so-called Winds Code, which has been a constant source of debate and irritation for half a century. On November 19, 1941, Tokyo used a low-grade code called J-19 to advise its diplomats in Washington to be on the alert for a special radio broadcast in the “case of an emergency (danger of cutting off our diplomatic relations) and the cutting off of international communications.” (Emphasis added) Should this happen, a special warning would be added to the middle of the daily Japanese-language shortwave news broadcast.

Thus, if all normal forms of communication were cut off, the diplomats in Washington should listen to Tokyo shortwave radio. If they heard the words “east wind rain” in the middle of the news broadcast, they would know that relations with America were in danger. If they heard the words “north wind cloudy,” they would know that relations with Russia were in danger. Similarly, if they heard “west wind clear,” they would know that relations with Great Britain were in jeopardy. The two messages also said that when the diplomats heard these Winds Code words, they were to “destroy all [their] codes, papers, etc.” In other words, if all normal communications were cut, and the Japanese diplomats around the world heard these messages broadcast by Tokyo, they were to destroy all their secret codes, their code machines and other means of secret communications in preparation for war.

The other three messages had nothing to do with the Winds Code. The third message was sent in the Purple code from Tokyo to Washington on December 1, but was not translated in Washington until December 5. This message told Washington to “discontinue the use of your code machine [Purple] and dispose of it immediately.” Special emphasis was given to taking the machine apart and breaking up its important parts. Washington was also ordered to burn its machine (Purple) codes.

The fourth message advised the diplomats that if they had to destroy their codes, they should contact the Naval Attaché in that particular embassy and use the special chemicals the Attaché possessed to ensure that vital documents and equipment were properly destroyed. This message was sent by Tokyo on December 1 in the Purple code and was translated in Washington the same day.

The fifth message was also sent in the Purple code from Tokyo to Washington, on December 2. Translated on December 3, corrected on December 4, it gave specific instructions to the Japanese diplomats in Washington to burn all their Purple codes, plus all the other codes the embassy possessed. Only one copy each of two special codes were to be retained for future use. Washington was also told that it was to “stop using at once one Purple code machine unit and destroy it completely.” (Emphasis added) Meanwhile, a cipher specialist named Kosaka should have been returned to Tokyo on a Japanese ship that left the United States on November 28.

These last three messages were of extreme significance.

When a nation prepares to launch an attack and go to war, one of the most indispensable steps it takes is to make sure that its codes and code machines cannot be captured by the enemy should the enemy retaliate and raid an embassy for intelligence reasons. Washington knew from reading these messages that war would have to break out, with Japan attacking somewhere in the Pacific.

Therefore, the Navy in Washington alerted Kimmel on December 3 by sending two advisory messages that paraphrased the intercepts I have described above. The first message read as follows:

HIGHLY RELIABLE INFORMATION HAS BEEN RECEIVED THAT CATEGORIC AND URGENT INSTRUCTIONS WERE SENT YESTERDAY TO JAPANESE DIPLOMATIC AND CONSULAR POSTS AT HONGKONG, SINGAPORE, BATAVIA, MANILA, WASHINGTON AND LONDON TO DESTROY MOST OF THEIR CODES AND CIPHERS AT ONCE AND TO BURN ALL OTHER IMPORTANT CONFIDENTIAL AND SECRET DOCUMENTS.

The second message to Kimmel read:

CIRCULAR TWENTY FOUR FORTY FOUR FROM TOKYO ONE DECEMBER ORDERED LONDON, HONGKONG, SINGAPORE AND MANILA TO DESTROY MACHINE. BATAVIA MACHINE ALREADY SENT TOKYO. DECEMBER SECOND WASHINGTON ALSO DIRECTED DESTROY, ALL BUT ONE COPY OF OTHER SYSTEMS, AND ALL SECRET DOCUMENTS. BRITISH ADMIRALTY LONDON TODAY REPORTS EMBASSY LONDON HAS COMPLIED.1 (Emphasis added)

Because the Navy paraphrased these warning messages to Kimmel, they differ somewhat from the messages that I carried in my bomb pouch, which I showed Rochefort.² Both the Navy warning messages to Kimmel are exceedingly important, but the second is the more important. The first implies that Tokyo has ordered its embassies and consulates around the world to burn their codes. But it is the second that says without equivocation that Tokyo has ordered its consulates to destroy their codes and code machines. Once code machines were destroyed, there could be no turning back potential Japanese attacks. The consulates could no longer communicate effectively with Tokyo. War had to follow; it was inevitable.

This meant that Rochefort had seen vital intelligence warnings to Kimmel indicating that Japan was preparing to attack within a matter of days, or hours. It was imperative that I learn what had happened to this information within the Naval command at Pearl Harbor and how this information was passed to Short and his Army command.

As for the significance of the earlier Winds Code messages, from what I could determine, the three messages from Tokyo ordering the Japanese diplomats around the world to destroy their codes and code machines made a Winds Code broadcast unnecessary, since the required orders had already been given. I will speak about this later on. What I now needed to know was how Rochefort had communicated all the information about the Winds Code and the warning messages about codes and code machines from Washington to Colonel Fielder. (For the most authoritative analysis of the Winds Code, see the Appendix, The “Winds Code,” on pages 447 through 470. No one can read this report and believe the claims advanced by Captain Safford.)

I recalled that Fielder had told me earlier that he was not well acquainted with Rochefort and that the two men had no real liaison about such information. So you can imagine my surprise when Rochefort continued his testimony to me by saying: “It was my practice to give Col. Fielder all the information of importance in which the Army and Navy were jointly interested and which came to my knowledge in the course of my duties. This was done so that Col. Fielder and I could keep abreast of intelligence developments in our common interests.”

Rochefort cited examples to buttress his positions, not knowing that Fielder had never said anything about these things to me. First, Rochefort maintained that the Navy at Pearl Harbor had begun to monitor the Tokyo shortwave radio to see if it could intercept the so-called Winds Code. No such interception was ever accomplished, but he claimed he had discussed these abortive intercept attempts with Fielder. Next, Rochefort claimed that on December 4 or 5, he had given information to Fielder that the Japanese consul in Hawaii was destroying his secret papers (which differs from saying codes and code machines). Lastly, Rochefort said he had also shared this information with Robert L. Shivers, the FBI Agent in Charge in Honolulu, plus Rochefort’s head office in Washington.

Upon leaving Rochefort’s office, I pondered the unique command structure at Pearl Harbor as it existed in early 1941, when plans were drawn up for the defense of the Hawaiian Islands. The Army and Navy had agreed upon a method of coordination—called the Joint Coastal Frontier Defense Plan, Hawaiian Coastal Frontier—that was based on the concept of mutual cooperation. For those of us who were basically civilians in life and heart, when we joined the military in World War II, the concept of mutual cooperation between the various services was something that made us laugh or curse. If you train dogs or people to be fighters, when there is no enemy to fight, they compete against each other, and orders for mutual cooperation be damned.

Both the Army and the Navy had paid lip service to the concept prior to 1941. Plans were made during peacetime for every potential wartime contingency based on everyone’s cooperating with everyone else. But things hadn’t worked that way. If I was already finding conflicting stories about mutual cooperation among the worker bees, the men who were really doing the labors of the day in the field at Pearl Harbor, I guessed there might be even bigger conflicts ahead. And with these grim thoughts racing in the back of my mind, I returned to the Pentagon to prepare a paraphrase of Rochefort’s affidavit. (The rules were such that Rochefort’s comments had to be sanitized and safeguarded under wartime security regulations.) Knowing it would take some time for Sonnett to review Rochefort’s paraphrased affidavit, I continued my review of the historical records. Before my own investigation, there had been five major inquiries into Pearl Harbor:

(1) The Roberts Commission began on December 18, 1941, ended on January 23, 1942, and produced a 2,173-page report.

(2) The Hart Inquiry commenced on February 12, 1944, ended on June 15, 1944, and yielded a 565-page report.

(3) The Army Pearl Harbor Board, on which I had served, began on July 20, 1944, ended on October 20, 1944, and produced a 3,357-page report.

(4) The Navy Court of Inquiry started on July 24, 1944, concluded on October 19, 1944, and produced a 1,397-page report.

(5) Lastly, the Clarke Inquiry had two sets of hearings, September 14–16, 1944, and July 13 to August 4, 1945. This investigation yielded a 225-page report. (I read the latter portion as I concluded my own report.)

From these thousands of pages of testimony, I deduced that I would not need to take affidavits from Admiral Kimmel or General Short. They had spoken fully and freely before these commissions and hearings. There was nothing further they could add to my investigation except more qualifying and self-serving statements that would muddy the waters. (As I said previously, a good lawyer knows when to quit asking questions.)

In terms of how poorly the concept of command of mutual cooperation had worked before the attack of Pearl Harbor, I was particularly struck by the testimony of Comdr. William E. G. Taylor, a Naval Reserve Officer, who, before Pearl Harbor, already had two years of war experience in England. As a fighter pilot, Taylor spent one year with the British Navy and a year with the Royal Air Force, during which he had access to and learned the air warning system that helped win the Battle of Britain, both ashore and afloat.3

According to Taylor, the Army had asked for the services of a specialist such as himself. He had been assigned to special duties to assist the Army “in an advisory capacity” to set up a radar early warning system for Pearl Harbor. His time was spent with various staffs in Hawaii trying to work out the liaison between the aircraft warning systems and the various commands. According to Taylor, although there was no permanent radar installation per se, the five mobile sets in use by the Army before the attack on Pearl Harbor were “adequate to do a fair job of early warning.” The problem was that “the communications between the fighter-director officers’, or controllers’, positions, and the fighter aircraft were totally inadequate to control fighters more than five miles off shore.”

The control and operation of the shore-based radar was the responsibility of the Army. But, General Short claimed, it did not have enough trained personnel to operate the equipment twenty-four hours a day. This was a complete contradiction of the Roberts Commission Report. Then, to save the equipment, General Short had ordered that it not be operated continuously. Said Taylor: “I feel, and felt then, that these stations should have been operating twenty-four hours a day, and the air warning system fully manned.”

Among his attempts to get the air warning system operating properly, Taylor called a meeting on Monday, November 24, copies of the minutes of which were circulated the next day to the appropriate Army and Navy commands. Everyone present at the meeting agreed that the interservice operational structure was inadequate; measures needed to be taken immediately to correct the situation. But, as Taylor explained, “very little was done as a result of this conference. We managed to complete our communication lines. We were not able to have either the Army or the Navy agree on an aircraft identification system. We were not able to get men to man the information center. We were able to get no more personnel and the information center remained as it was on 24 November. The fact that the radar stations were shut down, except for the period of 4:00 A.M. to 7:00 A.M., made it impossible to continue to train plotters and operators for more than three hours a day, which was not enough. That fact alone did more to slow down the development of the information center than anything else.”

When asked if he had made a request to the proper Naval authorities for Navy liaison officers to work with the Army on an air warning system and an information center, Taylor said he had made a verbal request for this from “Commander-in-Chief, Pacific Fleet, staff, and Com 14’s chief of staff [Capt. J. B. Earle], and also Admiral Bellinger.” But, said Taylor, “The reply was in all three places there were no liaison officers available.”

Had an information center been fully operational, the absence of these Naval officers would have prevented the communications center from operating properly. For without officers from both the Army and the Navy to identify which planes belonged to which service, the information center could not discriminate between friend and foe.

There were efforts on the part of the Army to make the system work, but only by two junior officers, both captains. “The two of them worked very hard and tirelessly,” said Taylor, “but they did not have enough force [clout] to get what they needed from the various commands to get the station [information center] operating.”

As it turned out, the radar station was operating on the morning of December 7, albeit only by radar operators who were being trained and who picked up the signals of the approaching Japanese planes some one hundred miles away from their designated target. At approximately the same time, however, a flight of Army B-17 bombers was supposed to be arriving from the West Coast. “But without some method of identifying the planes that came in, no one could have told whether the planes were friend or foe, and therefore no action would have been taken,” explained Taylor. “Without the Army and Navy working together … it would have been assumed that [the planes coming in] were friendly. The information center is set up with its Army and Navy liaison officers for the single purpose of identifying the planes that are coming in.”

As a result of the Pearl Harbor attack, said Taylor in his testi mony, from the time after the bombs started falling, enough staff was supplied, and radars were on continuously, so any incoming plane was immediately identified; those that were not identified were declared hostile, and Hawaii’s air-raid sirens were sounded.

Ah, me, I thought. This is a perfect example of the problems of command by mutual cooperation. I have always believed in the maxim proposed by Gen. Ulysses S. Grant, who declared: “Two commanders on the same field are always one too many.” What had been needed at Pearl Harbor was a unified command under the control of one man who understood the problems and had the authority to correct them. The Navy turned off the radar on its ships when they entered the port of Pearl Harbor, because (1) the radar didn’t work in port, since the anchorage was surrounded by high hills; and (2) the radar caused havoc with local civilian electronic communications on shore. In other words, the Navy turned off its early warning safety equipment, which was as effective as the Army’s radar, not understanding that the Army’s radar was often not operating. There must have been something wrong with a system that created such mistakes and inefficiency.

This was the point that Commander Taylor was trying to make in his testimony. The Navy’s lawyers, however, were interested only in showing that it had been the Army’s responsibility for making sure the land-based radar, on which the fleet depended, was working.

I wondered why the Navy hadn’t demanded that the Army have its radar working. An admiral who made that demand, and forced the issue, thereby ensuring the safety of his fleet with an adequate early warning system, probably would have ended up a hero. For if the early warning system and identification center staffed by Army and Navy personnel jointly had been in place as Commander Taylor envisaged, someone most likely would have figured out that when the Army radar showed more than 130 planes approaching Pearl Harbor on that fateful morning, the images had to represent something other than a small flight of bombers expected in from the West Coast.

You can see the problem: Because mutual cooperation was the order of the day, the blame finders in the two services only wanted to discover if their opposite service had let them down. No one really wanted to focus on whether or not there might be something wrong with the overall system in which they were involved.

Another question that I would have to investigate was: What did Pearl Harbor know about the Japanese intentions before the attack on December 7? Pulling together all the various items of intelligence that had been sent Pearl Harbor from Washington, and what Pearl Harbor had acquired from its own activities, was an arduous task. More importantly, the material had to be arranged in chronological order lest the investigator lose track of what he was working with. Slowly but surely, the list began to take shape,4 and I created a memo that bore the following heavy-handed title:

INFORMATION MADE AVAILABLE TO GENERAL SHORT FROM WAR DEPARTMENT AND OTHER SOURCES OF THREAT OF WAR WITH JAPAN

AND

OF THREAT OF SURPRISE ATTACK BY JAPANESE ON PEARL HARBOR ARRANGED IN CHRONOLOGICAL SEQUENCE

Some of the items were long, others short. Each item was given a separate page. All in all, there were fifty-six pages of information, beginning with the fateful letter from Secretary of the Navy Frank Knox to Secretary of War Stimson of January 24, 1941, which said: “If war eventuates with Japan, it is believed easily possible that hostilities would be initiated by a surprise attack upon the fleet or the naval base at Pearl Harbor.… The dangers envisaged, in their order of importance and probability, are considered to be: 1) air bombing attack; 2) air torpedo-plane attack; 3) sabotage; 4) submarine attack; 5) mining; 6) bombardment by gunfire. Defense against all but the first two of these dangers appears to have been provided for satisfactorily.…”

This correspondence was sent Hawaii, and by itself, the list was devastating in terms of attacking the claims of Admiral Kimmel and General Short that they did not have enough information to be prepared for a Japanese attack. In statistical terms, nearly 25 percent of all the warnings received by Pearl Harbor during 1941 arrived between December 1 and December 7. If anything, this sudden flood of messages should have alerted the Hawaiian commanders of the ever-increasing urgency of the dangers now facing them, as previously predicted.

Even more important, however, were the answers I was seeking: Who at Pearl Harbor had seen which Magic messages, and what had they done about them?

Another question I was going to have to solve was what happened in Washington that Saturday evening and night of December 6, 1941, and during the Sunday morning hours just before the Japanese attack. There has been so much written about this subject, and so much of it is confusing, that the only way I can tell this story is by asking the reader to bear with me and follow the trail that I followed through my investigation.

The next affidavit I took was that of Col. Moses W. Pettigrew, Military Intelligence Service (MIS). From early August to December 7, 1941, he was the executive officer of the Intelligence Branch, G-2, at the War Department in Washington. More importantly, from November 1939 to August 1941, he had been the assistant to Col. Rufus Bratton in the Far Eastern Unit of G-2. (This was important, because Bratton was going to be a key figure in my investigation.) Pettigrew’s duties consisted of reading various Magic intercepts of the diplomatic radio messages from Japan to its consulates and embassies around the world. They were in a variety of codes, such as Purple, High Level Diplomatic and J-19.

Pettigrew recalled seeing the so-called Winds Code messages that Tokyo sent out around the end of November. Then, on December 5, someone whom Pettigrew could not recall showed him an intercept of a message from Tokyo indicating that the relations between Japan and the United States were in danger even though regular communications had not been cut. He understood the intercept to mean that “anything could happen,” so he prepared a secret cablegram for dispatch to Hawaii addressed to the Army G-2 there [i.e., Fielder], saying:

CONTACT COMMANDER ROCHEFORT IMMEDIATELY THRU COMMANDANT FOURTEEN NAVAL DISTRICT REGARDING BROADCASTS FROM TOKYO REFERENCE WEATHER

It was this message that became the infamous cable number 519. Because of statements made to him by various people in the Navy, whose names he could not recall, Pettigrew was sure that the Army in Pearl Harbor possessed the same information that he had received in Washington. Pettigrew also understood from these same Navy people, again nameless, that the Army in Hawaii had everything the Navy had in terms of decrypted intercepts. The Navy in Washington also falsely claimed to Pettigrew that Commander Rochefort was “monitoring and receiving these intercepts and breaking and translating the codes [in Hawaii] as well as Washington.” As Pettigrew understood the situa tion, this was being done to save time, and by Hawaii’s exchanging these decrypts with ONI in Washington, the translations could be checked: one message against another.

Pettigrew was a frustrating witness. He was one of those “I can’t recall” people. I doubted that he was telling me everything he knew, because he kept saying that he couldn’t recall things that he obviously should have remembered. He wasn’t going to point a finger of responsibility at anyone, but I was sure that in his own way, the so-called Army way, he was going to tell me whom else I should talk to, and what compass direction I should follow on my journey. Pettigrew was so typically military, and so afraid to criticize his superiors in any way. And so wrong in his understanding of what was actually going on at Pearl Harbor. He knew far more than did the G-2 in Hawaii. But Pettigrew didn’t realize that Colonel Fielder wasn’t cleared for Top Secret, and didn’t know anything officially about the interception of the Japanese codes.

Pettigrew said that on December 5, he took the draft of message number 519 around the office for approval. Colonel Bratton was one of the people who signed his initials on it, giving approval for it to be sent over the signature of General Miles. According to the outgoing message number—519—on the copy of the message that I showed Pettigrew, he stated that the message had been sent to Hawaii, but he had no knowledge of any action taken on it by the Hawaiian Department. He recommended that I interview personnel in the War Department Message Center, check the records in the Signal Corps files in Hawaii and see one Col. Carlisle Clyde Dusenbury, who also had been an assistant to Colonel Bratton. Without saying so, Pettigrew was sending me to a previously unknown source, a man who would prove to be one of the most important players in explaining what went wrong at Pearl Harbor.

After Colonel Pettigrew had signed his affidavit and left my office, I reviewed what he had told me. Instinct told me I had learned something important. Knowing how the Army grapevine worked, I wanted to strike while the iron was hot, so I immediately called Colonel Dusenbury to set up an appointment for the next day.

If things went as they usually did, I suspected that Dusenbury would run to his boss, General Carter Clarke, who would give him some briefing and instructions. I already had an earlier run-in with Clarke about how I would conduct my investigation. He always tried to keep his finger on everything, which is what made him such a good intelligence officer, and I had gone to him and said: “Listen, I want to be very sure you fellas don’t tap my phone, and do not plant a bug in my office. Because if you do, the most astonishing things will happen. It’s going to be very, very disastrous for you personally.”

Clarke had promised faithfully that he wouldn’t do anything of the sort. I don’t think he ever did bug me, even though that’s what intelligence officers are supposed to do. I believe I really scared him; he knew I didn’t give a damn about having a career in the Army. And to protect the integrity of my investigation, I’d sound every alarm in the Pentagon if I found out that Clarke was messing around.

That doesn’t mean that Clarke didn’t know what Dusenbury was going to tell me, however, or that Clarke knew that Dusenbury had been kept under wraps since Pearl Harbor.

I cannot recall what Dusenbury looked like. But I remember that in giving his story, he stayed calm and collected the entire time, although he committed professional suicide before my eyes. He first explained how, during the course of the war, he had been Assistant Director of Intelligence, Southeast Asia Command, at Kandy in Ceylon, and was now on temporary duty in Washington. I decided to move slowly and began walking him through some of the material I had learned from Colonel Pettigrew the previous day.

For several months prior to, and after, December 7, 1941, Dusenbury had been serving as the Executive Officer of the Intelligence Group, Military Intelligence Division, in Washington; his duties were administrative, but also included the drafting and approval of outgoing messages. Dusenbury recalled cable number 519 of December 5, which Pettigrew said he had sent to Pearl Harbor. In fact, Dusenbury thought he had written the first draft of it. He recalled the reason for sending it as being twofold. First, there was danger indicated to the United States; second, “there was believed to be a lack of confidence by Edwin T. Layton, Fleet Intelligence Officer, as to Kendall J. Fielder, G-2, Hawaiian Department.” (This was the first confirmation I had received that there had been some form of trouble between Layton and Fielder.)

Dusenbury understood that in early 1941, the Navy had about four or five hundred people working in Hawaii monitoring, breaking and translating “the Japanese diplomatic codes.” He was wrong about the diplomatic codes, but correct in his belief that Commander Rochefort had been notified about our intercepting Tokyo’s messages about the Winds Codes, plus other information, and Dusenbury believed that message number 519 would cause Fielder to go to Rochefort and get clued in.

Dusenbury had gained his background knowledge by working as assistant to Colonel Bratton since first reporting for duty in Washington in August 1940. For the first four or five months, Colonel Bratton received from the Signal Corps the translated Magic intercepts of Japanese diplomatic messages—about fifty to seventy-five of them a day—of which about twenty-five would be sorted out for distribution. Bratton would deliver these messages to Colonel Harrison, the aide to Secretary of War Stimson; John F. Stone, secretary to Secretary of State Cordell Hull; Col. Ralph G. Smith, Executive Officer of the Assistant Chief of Staff, G-2; plus Col. Walter Bedell Smith, the Secretary of the General Staff, and Col. Thomas T. Handy, or Col. Charles W. Bundy, in the War Plans Division. The decrypts that were not circulated were destroyed immediately. Those that were circulated were placed in binders and delivered to the recipients, who signed a receipt for the binder if it had to be held for a few hours. When the binders had been read and returned to Colonel Bratton, their contents and any receipts that had been signed for them were also burned. Only one copy of the decrypts that had been circulated was kept in a secure central file in G-2, but there was no record of who had seen the messages.

Bratton and Dusenbury began alternating the job of assembling and distributing the intercepts around February 1941. By December, Dusenbury was doing almost everything himself. (Bratton, apparently, was going to be sent to the General Staff College in preparation for a combat command and promotion to brigadier general.) But, said Dusenbury, no record of the deliveries that he made “is now available in G-2,” and “none was kept because of Top Secrecy requirements.”

Heavens, I thought, what a sloppy way to run a railroad. I couldn’t run a business without keeping records. I’d go bust if I tried it, and all that talk about Top Secret requirements: that’s just bunk, to cover your flanks. I’d better keep pushing. Something worse is hidden here.

The question in my mind was what had happened in Washington during the evening and night of December 6, and in the predawn hours before Pearl Harbor. As I explained earlier, the Army and the Navy had been dividing the work of decoding and translating the intercepted Japanese diplomatic messages. (It is important to keep in mind the responsibilities of the two services to decrypt and translate on the odd and even days in terms of the question: Did the system work?)

Early on December 6, said Dusenbury, it became apparent and ominous in Washington that something of great importance was happening. The Japanese were making their reply to the latest American proposal that had been set forth during diplomatic negotiations, which were aimed at peacefully halting Japanese aggression in the Pacific. By noon, the Army and Navy knew from a so-called separate pilot message that Tokyo was sending a fourteen-part message to its diplomats in Washington, asking them to keep the information secret for the time being. As to when it was to present the fourteen-part message to the U.S. government, Tokyo said that the timing would be wired in yet another message.

All the testimony I had read to date showed that both the Army and the Navy had collaborated on translating the decrypts of the first thirteen parts of the fourteen-part message, with the Navy doing the majority of the work. This didn’t make sense to me at the time, because it wasn’t the Navy’s day for this task. Copies of the first thirteen parts were ready for distribution by nine o’clock that Saturday evening. (The fourteenth part of the message would be the conclusion of what Japan planned to do.)

Between nine-thirty and ten P.M. that December 6, the first thirteen parts of the message were given by a Navy courier to an assistant naval aide at the White House with the request that they be handed to the President at the earliest possible moment. They were presented immediately to President Roosevelt, who read them with his adviser, Harry Hopkins. The aide who delivered these first thirteen parts to the President had testified that, after both men had read the messages, “the President then turned toward Mr. Hopkins and said, in substance—I am not sure of the exact words, but in substance—‘This means war.’

“Mr. Hopkins agreed, and they discussed then for perhaps five minutes the situation of the Japanese forces, that is, their deployment.”

Once the President had finished reading the decrypts, the same Navy courier took them to the Wardman Park Hotel, where they were read by Secretary of the Navy Frank Knox. The courier then went to the home of Rear Adm. Theodore S. Wilkinson, the head of the Navy’s Intelligence Division, who was hosting a dinner party that included Rear Adm. John R. Beardall, the naval aide to the President, and Gen. Sherman Miles, the Chief of the Army’s Military Intelligence Division. All of these men read the thirteen-part message. The Navy courier then returned to the Navy Department at about one A.M. and, since the fourteenth part of the message, the conclusion, had not been received, went home to bed.

Meanwhile, another Navy messenger delivered copies of the thirteen-part message to the homes of Rear Adm. Royal E. Ingersoll, the Assistant Chief of Naval Operations, and Rear Adm. Richmond Kelly Turner, the head of the all-important Navy War Plans Division.

As for the distribution of the first thirteen parts of the message within the Army’s chain of command, Col. Rufus S. Bratton had testified to the Army’s Pearl Harbor Board 5 that he delivered the thirteen parts in the following order: First, he delivered them to Col. (later Lt. Gen.) Walter Bedell Smith, the Secretary of the General Staff, in a locked bag to which Marshall had the key. Bratton swore he told Smith that the bag contained very important papers, and that General Marshall should be informed of this at once so that he could unlock the bag and read its contents. Bratton continued, saying that he handed another copy of the thirteen parts to General Miles in Miles’s office, and that he discussed the message personally with Miles. Bratton also claimed that Miles had done nothing about the message so far as he knew. Third, Bratton testified, he delivered a copy to Col. Charles K. Gailey, Jr., for Brig. Gen. Leonard T. Gerow, the head of the War Plans Division. Bratton further testified that he gave another copy to the watch officer of the State Department for delivery to the Secretary of State, completing the distribution to everyone who was supposed to see the material by ten-thirty that evening of December 6.

I had been wondering about the apparent discrepancy between Bratton’s testimony, in which he claimed to have delivered the material to General Miles in his office, when I already knew that Miles had been having dinner that evening with Admiral Wilkinson. That was a delivery Bratton could not have made. Nor could he have discussed the Magic material with Miles, as he claimed. So I began prodding Dusenbury about his recollections.

Suddenly, my investigation flipped topsy-turvy.

Dusenbury remembered the night of December 6 all too clearly. As he told me his story, I could see his long and carefully nurtured Army career flying out the window.

The intercepts, “consisting of fourteen parts”—and it is crucial that Dusenbury told me fourteen and not thirteen parts—began coming in on the night of December 6 while he was on watch. Colonel Bratton was also on duty; he remained until about half of the message had been received, “whereupon he left and went home at about 9 P.M.” Dusenbury stayed alone so that Bratton could get some sleep. The fourteenth part of the message, in which the Japanese declared a break in diplomatic negotiations with Washington, arrived “about 12 that night.” Dusenbury solemnly swore he was so sure of this that on his affidavit, he scratched out his first words “11 or,” leaving the hour of 12, midnight, shining forth like a beacon.

But once the fourteenth part had been received, Dusenbury went home—without distributing any of the fourteen-part message to anyone! All fourteen parts of the intercepts were in his possession, but he did not distribute them. He finally began distribution the next morning, Sunday, December 7, at about nine, and when he gave the copy for the Operations Department to Col. Thomas T. Handy, Handy read the message and said to Dusenbury, “‘This means war,’ or words to that effect.”

I asked Dusenbury why he hadn’t delivered the material the night before, as Bratton had ordered. His reply has haunted me ever since: “I did not wish to disturb the usual recipients who were probably at home asleep, as I did not see the implications of immediate hostilities [in the messages]”!

Oh, Good Lord, I thought. All the earlier testimony about the thirteen-part message was nonsense. Dusenbury hadn’t understood the significance of the fourteen parts. His failure to deliver them had put him in the Army’s doghouse forever—he ended the war with the same rank he started it, a fate worse than death for a West Pointer—while Colonel Handy, who first read the messages, and immediately understood their import, ended the war as a full General and the Army’s Deputy Chief of Staff. But the real significance of Dusenbury’s failure to carry out orders was that he lost Washington a vital nine hours that could have been used to alert Hawaii and prevent Pearl Harbor.