Pearl Harbor after a Quarter of a Century
Harry Elmer Barnes
III:
Washington Should Not Have Been Surprised When The Japanese Attacked
Pearl Harbor
“Where [General George C.]
Marshall spent the rest of the afternoon and the night of the 6th [of
December] has never been determined in any final fashion. When examined
in the Joint Congressional Committee Investigation of 1945-1946,
although known for his excellent memory, Marshall contended that he
could not remember where he spent the night of December 6th, probably
the most significant, critical and exciting night of his professional
life, at least down to that time. . . . . During the Joint Congressional
Committee investigation Senator Homer Ferguson reported to his
colleague, Senator Owen Brewster, and to his research aide, Percy L.
Greaves, that a few days after Marshall’s attack of amnesia on the
witness stand, he overheard Marshall tell Senator Alben W. Barkley,
chairman of the JCC: ‘I could not tell you where I was Saturday night
(the 6th). It would have got the chief (Roosevelt) into trouble.’”
—Harry Elmer Barnes
1. The
Probable Place
of a Japanese Attack in the Event of War with the United States
No
item in the revisionist presentation of the causes and merits of the
second World War is better established than the fact that no top
military or civilian authority in
Washington
on December 7, 1941, should have been surprised at either the place or time of
the Japanese attack on the Pacific fleet at
Pearl Harbor.
The only element of surprise, if any, should have been over the damage
that the Japanese planes delivered to the fleet.
After the
Japanese had abandoned dependence on their Red diplomatic code, which
American cryptanalysts had earlier broken, American experts in the Army
Signal Corps, directed by Colonel William F. Friedman, had broken the
top Japanese Purple diplomatic code by mid-August, 1940, and for a year
and a half before Pearl Harbor Washington had been intercepting and
reading the secret Japanese diplomatic messages to their officials all
over the world. Less difficult diplomatic codes, such as J-19 and
PA-K2, were also easily read. Among other things, this breakthrough had
enabled the
Washington authorities to know that the Japanese peace offers were
sincere and not mere window dressing for sinister later designs of an
aggressive nature. The Japanese messages also revealed equally clearly
that if even extreme Japanese efforts to reach a peaceful settlement
with the
United States failed, the Japanese would go to war for self-preservation
and self-respect. We may first consider the extensive evidence that, if
Japan
did attack the United States, it would be where the American fleet was then located,
namely, at
Pearl
Harbor.
For
years before the attack on Pearl Harbor, naval maneuvers had been held
off the
island
of
Oahu
in
Hawaii
to test the feasibility of a Japanese attack on
Pearl Harbor.
The results were far from reassuring to the
United States, and were equally a definite warning of the danger and
practicability of a Japanese task force attack there. As early as 1932,
Admiral Harry E. Yarnell, one of our earliest air-minded naval officers,
made the first carrier-based task force test and he was able to execute
a surprise attack when operating only sixty miles off
Pearl
Harbor. These maneuvers were continued, and in 1938 a successful
air attack was launched from the carrier
Saratoga one hundred miles off
Pearl Harbor.
The Japanese task force in December, 1941, operated from over 200 miles
away. In April, 1941, General Frederick L. Martin and Admiral Patrick
N. L. Bellinger, commanders of the Army and Navy air forces respectively
at Pearl Harbor, described in detail the nature of a possible Japanese
air attack on
Pearl Harbor
which was uncannily identical with Yamamoto’s plan for the actual
Japanese attack a few months later. This was forwarded to the Army and
Navy headquarters in
Washington but no positive response or protective operation took
place.
Long
before Admiral Kimmel assumed command at Pearl Harbor in January, 1941,
it had become basic in Pacific naval strategy to accept the fact that if
the Japanese ever started a war with the United States they would first
strike our Pacific fleet, especially if based at Pearl Harbor, to
protect their flank before they could safely move large naval forces
south or north from Japan. This had been constantly emphasized to
Washington from the time of the assertions of General Hugh Drum in 1935
and of General George V. Strong in 1940, to the observations in 1941 of
Commander Arthur N. McCollum, head of the Far Eastern Section of Naval
Intelligence, the man who had probably the best informed conceptions of
the naval and diplomatic situation in the Far East, with the possible
exception of Colonel Otis K. Sadtler of the Army Signal Corps and
Colonel Rufus S. Bratton, Chief of the Far Eastern Section of Military
Intelligence.
Viewed
most generally, then, it had long been assumed that the Japanese would
not go to war with the
United States without first protecting their flank by trying to destroy
the American Pacific fleet, wherever it was stationed. It was also
clear that the American fleet would be both more inciting and more
vulnerable to a Japanese attack if stationed at Pearl Harbor, as
compared to its relative safety before the spring of 1940, when it had
been based on the Pacific coast of the United States, mainly at San
Diego. Admiral James O. Richardson, Kimmel’s able predecessor as
commander of the Pacific fleet, bitterly protested the fleet’s permanent
retention at Pearl Harbor, after maneuvers in the spring of 1940, and
labelled Pearl Harbor “a damned mouse trap” for the American navy.
Indeed,
it is certain that Richardson’s untimely removal as head of the fleet was brought about
by his determined resistance to what he considered the folly of keeping
the fleet at
Pearl
Harbor. Admiral Frank E. Beatty, a well informed authority, has
told me that it may also have been due in part to the animosity of Harry
Hopkins, who sat in on
Richardson’s
conferences with Roosevelt. Richardson
was annoyed by Hopkins’ interjection of his opinions into the debate and
understandably commented unflatteringly on
Hopkins’ lack of qualifications as an authority on naval strategy.
Added to
this generalized conception of our Pacific naval strategy centering
around Pearl Harbor was a precise statement from our Ambassador in
Tokyo, Joseph C. Grew, in January, 1941, that he had received a friendly
warning from the Peruvian Minister in Tokyo, which the latter had
obtained from several sources, one Japanese, to the effect that, if
Japan could not reach peaceful relations with the United States, it
would start war by a surprise attack on Pearl Harbor. After the
successful Japanese attack on December 7, 1941, the Washington
authorities, who were desperately trying to cover up their bad guessing
or actual guilt, tried to represent this warning as worthless hearsay,
but it was not so regarded by Ambassador Grew and some top Washington
officials in January, 1941, notably Secretary of the Navy, Frank Knox.
It
was not necessary, however, to rely on generalized strategic
considerations, however sound. From September, 1941, to December 7th
Washington authorities intercepted a considerable number of Japanese
messages between Tokyo and Honolulu that specifically and most obviously
indicated that, in the event of war between Japan and the United States,
the first Japanese move would actually be a surprise attack on the
Pacific fleet—that Pearl Harbor would be the target. These messages
came to be known as the “Bomb Plot” messages and consisted of requests
from the Japanese government in Tokyo to the Japanese consul-general in
Honolulu, Nagoa Kita, for detailed and specific information as to the
nature, number and types of vessels in the Pacific fleet at Pearl
Harbor, their location and movements, and other relevant information
connected with the American military establishment located there,
together with Kita’s replies to these requests. These requests from
Tokyo to Kita became more insistent, frequent and detailed as we
approach December 7, 1941.
The
first of these was sent in the J-19 Japanese code to Kita on September
24, 1941, and was decoded, translated and read on October 9th at
Washington. This requested very detailed information on the
composition, location, and operations of the Pacific fleet at
Pearl
Harbor. From this time onward,
Washington should have had no doubt that the Japanese were planning a
surprise attack on
Pearl
Harbor, if negotiations failed. The Bomb Plot messages clearly
pinpointed Pearl Harbor as the target of any Japanese surprise attack on
the
United States.
When
relations became more tense after the fall of the Konoye Cabinet in
October, 1941,
Tokyo ordered that these espionage reports from Kita should be
sent at more frequent intervals. On November 15th, Kita was ordered to
send his reports twice each week. On November 18th and 20th, orders
were given to inform
Tokyo in regard to all our warships and others anchored in areas
adjacent to
Pearl
Harbor. On November 29th, Kita was ordered to make his reports,
even though there had been no movements of the warships at
Pearl
Harbor.
No
such detailed or comprehensive reports, containing as they did grids and
coordinates, were demanded of any Japanese officials and spies at any
other American outpost or naval base anywhere in the world, not even
those on the Pacific coast. Those who have sought to minimize the
significance of these Kita Bomb Plot messages have pointed out that
Japanese spies were frequently detected making inquiries at leading
American naval bases but these were routine and trivial matters and not
in any way to be compared or rated with the Kita messages. All these
Bomb Plot messages were available to the appropriate top Washington
officials in the Army and Navy and to Roosevelt and Hull, and they
thoroughly established the probability that if the Japanese made any
surprise attack on the United States it would be at Pearl Harbor.
The
most crucial Kita report available in Washington before Pearl Harbor was sent to
Tokyo by Kita on December 3rd. He informed the Japanese
government that he had set up an elaborate system of window code signals
at
Lanikai
Beach which were easily visible to boats off the coast. From
this spot he would signal passing Japanese fishing craft and submarines
as to the nature and movements of the Pacific fleet. These boats and
submarines could then pass this vital information back to the Japanese
task force as it was nearing
Pearl
Harbor for the attack.
This
sensational and revealing message was intercepted at the army monitoring
station at Fort Hunt, Virginia, on the 3rd, was decoded by Naval
Communications in Washington before noon on the 6th, and was translated
and ready for reading and distribution before 2:30 P.M. on that day.
This finally confirmed the pin-pointing of Pearl Harbor as the place of the Japanese attack. Due to the fact that
the Kita message implied that the signals would end on the night of the
6th, this December 3rd intercept also clearly indicated that the
Japanese task force under Admiral Nagumo was moving on toward Pearl
Harbor and intended to organize off Oahu on the night of the 6th, and
make ready for the attack on Pearl Harbor the next morning. Hence, this
message not only made it clear that Pearl Harbor would be the place of
the Japanese attack but also revealed the time of this attack, unless
something happened to slow down or divert Nagumo’s expected arrival on
the 6th as anticipated.
How
far Roosevelt,
Hull,
and the top military brass in Washington were informed of the nature, contents and implications of
this vital and revealing Kita message that was available on the 6th has,
naturally, been the subject of much controversy. It was actually far
more revealing than the fourteen-part reply to
Hull and the Time of Delivery message as to the time, place and
certainty of an immediate Japanese attack on
Pearl Harbor.
If it could be proved that its contents had been known to the top
officials in Washington by early evening of the 6th, then their failure
to warn Kimmel and Short would appear to be far more culpable than that
connected with the replies to Hull that were not available until the
late evening of the 6th and the morning of the 7th, and even then did
not make the time and place of the Japanese attack on Pearl Harbor
anything more than trained and informed guesswork. Hence, there would
be every effort to indicate that no information about this Kita message
was available until after the attack.
Certain
of the important facts about the Kita message are established beyond any
reasonable doubt. Its interception on December 3rd has been described.
It was decoded some time between the 3rd and 6th and was given to the
translating section of Naval Communications for translating. This was
done by a Mrs. Dorothy Edgers, a competent expert on the Japanese
language, between 8:00 A. M. and
2:30
P. M. on the 6th. Her immediate superior was Yeoman Bryant, and the
chief of the section was Commander Kramer. Both of these men knew
during the time that Mrs. Edgers was working on the translation that she
regarded it as a very important document and that she gave it careful
attention. She was supposed to leave at noon, but was so much
interested in the document that she worked until after 2:00 P.M. to
complete and revise her translation. She handed it over to Yeoman
Bryant to discuss with Commander Kramer with respect to its distribution
to top-level civil and military officials entitled to receive such
material. While there is controversy over whether Kramer read the
Edgers’ translater [i.e., “translation”—A.F.] carefully, there is
little doubt that Bryant did so. The main dispute is over whether
Kramer distributed the message to at least a few key officials in the
Army and Navy on the afternoon of the 6th.
The
accepted legend is that when Kramer looked over the Edgers’ translation
after she left the office he found that it was so imperfect that it was
unsuitable for immediate distribution. The excitement that followed with
the arrival of the Plot Message and the Japanese reply to Hull, together
with Kramer’s responsibility for distributing the reply to Hull during
the evening of the 6th, made him decide to delay reworking the Edgers’
translation until Monday, the 8th, when it was too late to be of any
value in warning Kimmel and Short.
The
circumstantial evidence tends to support the probability that Kramer
read the Edgers’ translation well enough to recognize its great and
immediate significance and showed the message to some of the leading
officers in the Navy, and possibly in the Army, and was ordered by these
persons, who recognized its importance, to suppress it for the time
being. Mrs. Edgers was a competent translator and she remembered the
essential parts and the full implications of the message well enough so
she could describe the contents on the witness-stand some three and a
half years later without ever refreshing her memory by seeing the
document during that long interval. If she could remember the message,
it is likely that Kramer could have quickly grasped its significance.
He
was familiar with the Bomb Plot messages from the time of the first one
decoded on October 9th, the importance of which he was the first to
recognize. Since the reply to Hull was in English, Kramer did not have
to be busy translating this on the afternoon of the 6th and should have
had plenty of time to study the Edgers’ translation and call it to the
attention of his responsible superiors. Kramer was the most severely
intimidated of all the witnesses in the post-Pearl Harbor
investigations—to the extent of bringing on a nervous breakdown. Hence,
he was not likely to come clean in his testimony on the Kita message if
he did suppress it. He has declined to answer personal questions on the
matter since his retirement. Although Yeoman Bryant was present in the
room during the investigation he was not called to testify.
Hence,
we are likely to remain as much in the dark about documenting the
distribution of this final and most sensational of the Bomb Plot
messages as we are about Captain Kirk’s frustrations in regard to the
first one on October 9th. The best discussion of the controversy to
reach print is that by Commander Charles C. Hiles which was published in
the Pearl Harbor supplement to the Chicago Tribune on
December 7, 1966, the twenty-fifth anniversary of the attack on
Pearl Harbor.
Why
these Bomb Plot messages were not sent to Hawaii by the Washington
authorities, so they could be used by Kimmel and Short and enable them
to be prepared for the Japanese attack has never been adequately
explained. In naval headquarters at Washington, they were suppressed
chiefly by Admiral Richmond Kelley Turner, whose ignorance of the
details of the cryptanalytic set-up and operations at Pearl Harbor was
only exceeded by his arrogant self-confidence, and Admiral Stark backed
him up instead of keeping his promise to Admiral Kimmel to have him
fully and speedily informed on all such matters. At the army
headquarters the responsibility was mainly that of General Marshall and
General Sherman Miles, chief of Military Intelligence. Until December
3rd, most of these messages were intercepted by the Army Signal Corps
station MS5 at Fort
Shafter, General Short’s headquarters near
Honolulu,
and were transmitted to Washington for decoding, translation and reading. They were also
usually intercepted at several other monitoring stations in the
United States.
When
the first one was decoded, translated and read on October 9th, Commander
Alwyn D. Kramer, who was in charge of the translation work for the
Far East
section of Naval Communications, noted that this was a very significant
message that needed further study. It must have received such study for
Captain Alan G. Kirk, the able, forthright and experienced Director of
the Office of Naval Information at the time, insisted that the October
9th message must be sent to Admiral Kimmel. He was blocked in this
proposal by Admiral Turner, who was supported in this by Admiral Stark.
Frustrated and disgusted, Kirk left his post and sought the sea duty he
needed to become an admiral, and later rendered very distinguished
service in naval operations in
Europe.
The details of Kirk’s leaving for sea duty have been furnished to me in
person by Admiral Beatty, at that time chief aide to Secretary Knox.
It
is most unfortunate that Admiral Kirk was not thoroughly interviewed
after the war concerning the refusal of Stark and Turner to permit him
to transmit this first Bomb Plot message to Kimmel. I had arranged to
have this done in 1962 when Kirk was residing in
New York City. Being on the opposite side of the continent at this
time, I could not do it personally, but had arranged that a trained
interviewer and an expert on
Pearl Harbor
would carry it out. He delayed briefly to make more complete
preparation, and in the meantime Kirk was appointed American ambassador
to Formosa (Taiwan). Another student of the situation, without my approval,
took the chance of writing Admiral Kirk in
Formosa about the incident. It was hardly to be expected that
Kirk could give any detailed answer under these circumstances. He might
well have been expected to ignore the letter but he gave a courteous
reply, making no categorical denial and thus by indirection implying
that he may have been prevented from transmitting the information to
Kimmel. He soon retired due to ill-health, was then in no condition to
accept the request for an interview, and died soon after his retirement.
This ended the possibility of clearing up the October problem in any
final and definitive manner. That the situation in the Office of Naval
Intelligence was confused in 1941 is evident from the fact that by the
end of October there had been four chiefs of this organization: Captains
Anderson, James, Kirk and Wilkinson.
It
is certain that Turner was directly responsible for frustrating Kirk,
but there is no proof that this was the result of any conspiracy to keep
Kimmel in the dark. Turner was a very able but conceited officer, sure
of himself. His mind was mainly on the Atlantic, and so far as the
Pacific was concerned he still believed that the Japanese would attack
Siberia. He was unpardonably ignorant about
Pearl
Harbor intercepting facilities at the time, actually believing that it
had a Purple machine and was reading the Japanese diplomatic messages on
the spot. The main responsibility for Kirk’s frustration was, however,
that of Stark, who had promised Kimmel that he would transmit to him all
significant information about any possible Japanese menace to
Pearl Harbor,
and Kirk had been fully informed of this. Why Stark deferred to Turner
in this episode has never been cleared up. Admiral Beatty informed me
that Turner often dominated Stark in the matter of naval decisions.
Stark refused to clarify matters in a long interview with Percy Greaves
in mid-December, 1962. It has been alleged on good authority that an
attempt was made to falsify the Naval Directory for 1941 to indicate
that Kirk had left his post as Director of ONI before October 9, 1941.
Kirk
was succeeded by an able but far less experienced and more pliant
person, Captain Theodore S. Wilkinson, who may have feared to repeat
Kirk’s insistence with Turner. This has come to be known as “the
October Revolution” in the Office of Naval Intelligence. In any event,
both the Army and Navy Departments had these crucial Bomb Plot messages
at hand and if they neglected them, then it was no less than a criminal
neglect and it was an important factor in leading to the destruction of
the Pearl Harbor fleet. The place—and through the Kita message of
December 3rd even the probable time—of the Japanese surprise attack no
longer needed to be a mystery. Not only the Japanese inquiries as to
the fleet, facilities and supplies at Pearl Harbor but also the general
strategic logic of all the circumstances connected with the launching of
a Japanese war against the United States at any time, and especially in
1941, made it all but certain that the first drastic move would be
against Pearl Harbor. Japan did not need to be attacked to start a war,
as
Roosevelt
did. It needed to destroy the American Pacific fleet, and it
would be difficult to do this after war had started elsewhere. Admiral
Kimmel would then deploy and scatter his forces as they moved to Wake,
Midway and the Far East and they would never again be bunched up as they
were at
Pearl Harbor
in peacetime.
It should
also be emphasized that although the treatment of the Bomb Plot messages
in the preceding pages has stressed the role of the Navy in receiving
and handling them, the Army also obtained them, and alert officers
therein were impressed as Captain Kirk had been by the threat to Pearl
Harbor which they revealed. Colonel Rufus S. Bratton, chief of the Far
Eastern section of Military Intelligence, delivered the original Bomb
Plot message, decoded on October 9th, to Secretary of War Stimson,
General Marshall and General Leonard T. Gerow, chief of the War Plans
division of the Army. These messages were discussed by officers in
Military Intelligence and the Signal Corps and most of them recognized
the desirability of sending them to General Short at Fort Shafter, but
they were no more able to get past Marshall and do so than Kirk,
Wilkinson, Noyes and McCollum could get by Turner and Stark. Just as
Turner was the chief navy obstacle to getting the Bomb Plot messages
through to Kimmel, so Marshall constituted the main blockage in passing
them on to Short, although he could delegate the action to General
Miles, chief of Military Intelligence. Marshall was also the person
mainly responsible for the slow transmission of the Bomb Plot messages
from MS5 at Fort Shafter to Washington, compelling them to be sent by
the China Clipper, every two weeks, or by ordinary boat mail, when they
could have been sent at once by cablegram or RCA radiogram. In other
words, General Short was as much victimized as Admiral Kimmel in being
deprived of these vitally important Bomb Plot messages.
2. The
Time of the Japanese Surprise Attack
Washington
also possessed extensive and diversified advance knowledge of the time
when the Japanese would attack Pearl Harbor; by the morning of December
7th almost to the minute when the attack would be launched. The Kita
message, which had been prepared for distribution by 2:30 P.M. on the
6th, also indicated even more directly that the attack would in all
probability take place on the morning of the 7th.
On
November 5th,
Tokyo
informed the Japanese embassy at
Washington that negotiations must be satisfactorily concluded by
November 25th. Unknown to
Washington
was the fact that the latter was also the date that the Japanese task
force was getting ready to leave the Kurile Islands for
Pearl Harbor
if negotiations were broken off, but with orders to return if
negotiations were resumed. On November 14th,
Tokyo
informed the Japanese consul at Hong Kong that
Japan
would declare war on the United States and
Great Britain, if the negotiations with the
United States failed. On November 11, 15 and 16,
Tokyo repeated to the Japanese ambassador in
Washington
that the deadline for completing negotiations with the
United States was November 25th. On November 22nd this deadline was
extended to the 29th, but the Japanese embassy in
Washington
was then emphatically informed that
Tokyo meant business this time and there would be no further
extension of the deadline. After that “things are automatically going
to happen.” On November 27th and 28th,
Tokyo informed the Japanese embassy in
Washington
that
Hull’s
ultimatum of November 26th was entirely unsatisfactory and
Japan would not negotiate any further on that basis. Hull
himself had said on the 27th that he knew his ultimatum meant war and
that henceforth, affairs between the United States and Japan were in the
hands of Stimson and Knox, the Secretaries of War and the Navy, both of
course under the control of President Roosevelt. On November 30th,
Tokyo
informed Germany that negotiations with the
United States had ended.
Yet,
on November 27, 28, 30 and December 1st there was a succession of
messages from Tokyo to the Japanese embassy in Washington warning them
not to reveal that negotiations were over, but to indicate they were
being stretched out. This move was both a last ditch attempt at a
peaceful settlement and, if that failed, an effort to cover up the
actual nature of “the things that are automatically going to happen”
after negotiations had ended, which was really on the 26th. These
“things” were the departure of the Japanese task force from the Kuriles
to
Pearl Harbor.
There had been no effort whatever to conceal the extensive movement of
large Japanese convoys and task forces to the southwest Pacific, and,
hence, these were clearly not the things which were “automatically going
to happen.”
The
policy of sending extensive Japanese convoys and task forces southward
helped to distract responsible attention in Washington from a possible attack on
Pearl Harbor,
even if it should not have done so, and thus worked well for the
Japanese program. This was one reason why many top officials in
Washington seemed to neglect the traditional Pacific strategy in regard
to Pearl Harbor and the Bomb Plot messages and after early November
concentrated most of their attention right down to December 7th on the
probability of an attack in the Far East, either on the Philippines,
along the coast of southeast Asia, or on the British possessions and the
Dutch East Indies. A little thought should have been sufficient to
convince the Washington authorities that
Japan
would not be likely to make its first major onslaught in the
Far East.
Admiral Thomas C. Hart’s Asiatic fleet was so small that its
destruction would not protect the Japanese flank from a major and
immediate American naval attack.
The
most important factor in this distraction from Pearl Harbor was the
basic strategic plan for a possible Pacific war with Japan, Rainbow 5 (WPL
46), which had been drawn up by our military services on the basis of
the Washington joint staff conferences ending with that at Singapore in
April 1941, and confirmed orally by Roosevelt in May and June. This
plan was based on the assumption that, if war came with Japan, it might
start in the Far East as a result of American commitments to come to the
aid of the British and Dutch, even if there were no attack on American
ships or territory. This naturally helped to divert the attention of
top military brass in Washington from the traditional Pacific strategy related to an attack
on the Pacific fleet, wherever located, especially during the week
before
Pearl
Harbor.
Washington
had all the numerous intercepted Japanese diplomatic messages indicating
by November 27, 1941, that war was in all probability only a matter of
days away, but Kimmel and Short knew nothing whatever about any of this.
They did not even know that
Hull
had sent an ultimatum to
Japan on November 26th or that the Japanese had warned that
negotiations were to be ruptured on the 29th if no settlement had been
arranged. They obviously knew nothing about the details of the
Japanese-American diplomatic negotiations from August to November 26th;
Short was in the dark after July.
By
November 27th, war with
Japan
seemed almost certain, and it was expected in
Washington
official circles that it would probably come coincident with the
Japanese reply to Hull’s ultimatum. Since
Japan usually made its surprise attacks on a weekend, when
opposing forces were most likely to be relaxed and off-guard, some
Washington
authorities, including Roosevelt, thought that the attack might come on
November 30th, but that was too soon for Japanese plans, which were
centered on the task force’s departure from the Kuriles and its movement
toward
Pearl Harbor.
When the Japanese did not attack on the 30th, there was special
apprehension in
Washington that it might come on December 7th, but no plans were made
or steps taken to warn Kimmel and Short of this possibility.
On
November 19th, the Japanese announced in their J-19 diplomatic code,
which we could and did read, the setting up of a so-called Winds System,
which Japanese diplomatic officials and consulates could intercept and
learn of Tokyo’s intentions in the event of breaking off diplomatic
relations and going to war with the United States. The Winds signals
were as follows: “East Wind Rain” for the war on the United States;
“West Wind Clear” for war on Great Britain; and “North Wind Cloudy” for war on
Russia. This Winds system, as we shall see, was executed on
December 4th.
Evidence
of the approach of war became ever more apparent after November 30th.
On December 1st and 2nd it was learned that Tokyo had ordered its main
embassies, with the exception of that at Washington, to destroy their
main code machines, including Purple, and burn their documents. This
was a measure that usually precedes immediate war and is rarely, if
ever, otherwise if ordered on any such scale as in December, 1941. The
Washington Purple machine was to be retained until December 7th so that
Tokyo could keep in touch with the Japanese embassy and be able to send
Ambassador Nomura the reply to Hull, which would be the last “peaceful”
communication, even though it would also mean, in all probability, the
actual onset of war.
On
the morning of December 4th a Japanese message was intercepted at the
important naval monitoring station at
Cheltenham,
Maryland. This did not need to be decoded for it was written in
plain Japanese language and was transmitted in the Japanese Morse code.
This was done in order to enable Japanese officials, who were without
decoding equipment after the codes-destruction order, to be able to
understand this critical message. This intercept was the all-important
execution of the Winds system set up by
Japan on November 19th. It is known as the “Winds Execute”
message and the information therein revealed that war would be made on
the
United States and
Britain, but not on
Russia.
Later
on, when the frenzied effort was made to cover up the responsibility for
the failure of Washington to inform Kimmel and Short, there was a
desperate attempt made to deny that any Winds Execute message had ever
been received, and most—perhaps all—copies of it were destroyed. The
last copy ever seen was identified by Commander Laurance F. Safford when
Commander Kramer was assembling documents for the Roberts Commission a
week after the
Pearl Harbor
attack. But honest and courageous experts, notably Safford, chief of
the Security Division of Naval Communications, who had received the
intercept from Kramer after translation and handed it over for
distribution, stuck by the facts and demolished all efforts to repudiate
the authenticity of Winds Execute. Safford was able to list some
fourteen persons, including Admiral Thomas C. Hart, commander of the
Asiatic fleet, who said that they had seen the Winds Execute message or
had discussed it with a responsible official who had seen it. The Naval
Court of Inquiry, which met from July to October 1944, established
beyond any doubt that the Winds Execute message was received on December
4th. Colonel Otis K. Sadtler, acting chief of the Army Signal Corps,
not only testified that he had seen the Winds Execute message but said
that he regarded it as the most important intercept he had ever handled.
On it he based a forthright warning to
Marshall’s
subordinates, presumably with
Marshall’s approval. The investigations by the Army
Pearl
Harbor Board and the Clarke Inquiries indicated that the Army
authorities knew that the Navy had intercepted the Winds Execute message
on December 4th.
Some
of the apparent excitement and confusion which seemed to prevail in
Washington military circles on December 4th, when Winds Execute was
received, may have been due to the fact that this was also the day that
the Chicago Tribune published the implications of Rainbow 5, thus
revealing Roosevelt’s deception of the American people as to his war
plans and his promise not to go to war unless attacked. At least we
know that
Marshall was far more concerned about this vital leak than he was
about the reception of Winds Execute. Colonel John R. Deane was still
working on this problem when he saw
Marshall in his office at
10:00
on the morning of the 7th.
Winds
Execute was not only the first explicit assurance that Japan was going
to make war, but it also made it clear that the war would be declared
against the United States. All that remained to be revealed was the
moment of the Japanese attack, and it was expected that this would be
when the Japanese handed in their reply to Hull’s ultimatum of the 26th,
which turned out to be the case.
It
was not necessary to wait long. The final and decisive Kita Bomb Plot
message, sent to Tokyo on December 3rd, was intercepted at
Fort Hunt,
Virginia, on the same day. By December 6th it was decoded,
translated and available for distribution at 2:30 P.M. in the
communications section of the Navy Department in
Washington. This revealed that the Japanese task force was nearing
Pearl Harbor and was expected to arrive off
Hawaii by the night of the 6th.
On
the heels of processing the Kita message came the so-called Pilot
Message from Tokyo, which announced that Japan was sending to the
Japanese embassy in Washington its ominous and anxiously awaited reply
to Hull’s ultimatum. The Pilot Message was decoded, translated and
ready for distribution before mid-afternoon on the 6th and enabled the
Washington
authorities to know that the Japanese reply to
Hull was arriving, that negotiations were over, and that war
was now at hand. The whole fourteen-part message told little more than
this, aside from a summary of the negotiations. There was no doubt that
Japan
would attack the United States in a matter of hours. Short and Kimmel
should have been warned at once. This would have provided their last
fair and decent opportunity to take action to avert, evade or repel the
Japanese attack. Of course, they should and could have been warned
weeks earlier, and certainly by November 27th.
Some
thirteen parts of a total of fourteen in the complete reply to Hull came
in during the afternoon of the 6th, were decoded from the Purple code,
and were ready for distribution by that evening. Since the reply to
Hull was in English it did not need to be translated. A copy
was delivered to Roosevelt and Harry Hopkins at the White House about
nine o’clock that night. After reading it, Roosevelt acknowledged that
it meant war but he took no steps to order any warning sent to
Pearl Harbor.
As we shall point out later on, Roosevelt knew by the forenoon of the
6th, if not on the 5th, that the
United States was already at war with
Japan due to our commitments to the British and Dutch under ABCD
and Rainbow 5.
The
thirteen-part message was delivered on the evening of the 6th to the
available Army officers entitled to Magic and to Hull under the
supervision of Colonel Rufus S. Bratton, chief of the Far East section
of Military Intelligence of the Army, and to the appropriate Navy
officers, except for Stark, who was at a theater, by Commander Alwyn D.
Kramer of the Far East section of Naval Communications. A copy for
Marshall
was left by Bratton with Colonel Walter Bedell Smith, who was
Marshall’s
secretary and the man short of Roosevelt himself, most likely to know
where
Marshall was to be found. It was Smith’s duty to deliver such
messages to
Marshall.
The
final or fourteenth part, also in English, arrived during the night of
the 6th and was decoded by early morning on the 7th. It confirmed the
Pilot Message’s implication that negotiations between
Japan
and the United States were over, hardly news to
Washington. Following this fourteenth part of the reply to
Hull was another and far more important short message from
Tokyo, the crucial so-called Time of Delivery message. It
ordered the Japanese ambassador, Admiral Nomura, and his associate,
Kurusu, to deliver the full fourteen-part Japanese reply to Secretary
Hull in person at 1:00 P.M., Washington time, about 7:30 A.M. Pearl
Harbor time.
The
fourteenth part was intercepted, decoded, and ready for distribution by
7:30 A.M. and the Time of Delivery message by 9:00 A.M., if not earlier.
One authoritative report indicates that both were ready for distribution
before 7:00 A.M. They had been received before 5:00 A.M. This does not
make too much difference because Admiral Stark did not get to his office
before 9:00 and General Marshall was either on a horseback ride or
hiding out in some place. When these late intercepts were shown to
Admiral Stark; chief of Naval Operations, about 9:00 by Admiral Leigh
Noyes, chief of Naval Communications, Captain Wilkinson chief of Naval
Intelligence, and Commander McCollum chief of the Far East section of
Naval Intelligence, they pointed out to Stark that 1:00 P.M. in
Washington was about 7:30 in Pearl Harbor and that this could very well
mean that the Japanese would attack Pearl Harbor at 1:00 P.M. Washington
time. Stark had four hours remaining in which to warn Kimmel, whom he
could have reached in ten minutes or less by his fast naval transmitter,
but he ignored the appeal of Admiral Noyes, Captain Wilkinson and
Commander McCollum that he send a separate warning to Kimmel, and did
nothing for the time being beyond phoning to Roosevelt, who certainly
did not order him to warn Kimmel.
Marshall’s conduct on the morning of the 7th was even more
mysterious than that of Stark. According to the accepted legend,
supported by sworn testimony of himself and prominent army associates,
he abruptly left his office in the old Munitions Building on Saturday
afternoon right after he learned from the Pilot Message that the
Japanese reply to Hull was about to start coming into Washington, which
was exactly the moment Marshall should have settled down in his office,
warned Short of the prospect of immediate war, and spent the night with
him discussing the best manner of dealing with the imminent attack. Of
course, if he had given Short an honest and adequate warning on November
27th, there would have been no attack to discuss on the night of the
6th.
Where
Marshall spent the rest of the afternoon and the night of the 6th
has never been determined in any final fashion. When examined in the
Joint Congressional Committee Investigation of 1945-1946, although known
for his excellent memory, Marshall contended that he could not remember
where he spent the night of December 6th, probably the most significant,
critical and exciting night of his professional life, at least down to
that time. Later on, after his wife had gallantly refreshed his memory,
Marshall stated that he spent it at home with Mrs. Marshall, who
was recovering from an accident at the time. During the Joint
Congressional Committee investigation Senator Homer Ferguson reported to
his colleague, Senator Owen Brewster, and to his research aide, Percy L.
Greaves, that a few days after Marshall’s attack of amnesia on the
witness stand, he overheard Marshall tell Senator Alben W. Barkley,
chairman of the JCC: “I could not tell you where I was Saturday night
(the 6th). It would have got the chief (Roosevelt) into trouble.”
Continuing
the official legend, the next morning Marshall rose leisurely, had a
late breakfast with his wife, and then took a long horseback ride when,
for all he is alleged to have known, the Japanese could have already
attacked the United States. While washing up after his return from the
relaxing ride, he was summoned to his office by Colonel Bratton, who had
been greatly alarmed by the clear implications of the Time of Delivery
message relative to an attack on Pearl Harbor.
Arriving
at his office about 11:25, so the story goes Marshall allegedly read for
the first time the fourteen-part reply to Hull and the Time of Delivery
message, and decided to send a warning to Short, which he did at 11:50.
Despite more rapid means of communication that were available, it was
sent by Western Union to San Francisco and from there to Hawaii by RCA,
not marked urgent, and was not actually put on the wires until 12:17.
It did not reach Short until after the Japanese planes had returned to
their carriers, over 200 miles from
Pearl Harbor.
This delay in delivery did not, however, make too much difference,
since the message was sent far too late, even if telephoned, to have
given Short enough time to have taken effective steps to repel the
Japanese attack and it did not even suggest to Short that there was any
reason to expect that a Japanese attack might take place immediately.
Kimmel and Short thus remained entirely unwarned even after leading
experts in both Army and Navy Intelligence had concluded by around 9:00
A.M. that the fourteen-parter and the Time of Delivery messages meant
that the Japanese would probably attack Pearl Harbor at about 1:00 P.M.,
Washington time.
Another
much different version of
Marshall’s activities from mid-afternoon on the
6th to
noon on the 7th appears to be far closer to the truth than the
traditional legend and it is supported by persons of unimpeachable
integrity. This version also accepts the fact of the complete mystery
of Marshall’s disappearance from the mid-afternoon of the 6th to the
morning of the 7th and our lack of precise knowledge as to where he was
during all this time, when he should have been constantly in touch with
Short at Fort Shafter. But it does eliminate the horseback ride and
Marshall’s incredibly late arrival at his office on the morning of
the 7th. Colonel John R. Deane, then an aide of Colonel Walter Bedell
Smith, who was Marshall’s secretary at the time, has asserted that he
saw Marshall at his office at about ten o’clock on the morning of the
7th. Commander McCollum has twice stated once under oath that
Marshall came to Stark’s office with a military aide about 9:00
that morning. Marshall and Stark, along with others in Stark’s office,
notably Admiral Noyes, discussed the fourteen-part and Time of Delivery
messages, and formulated the message that was to be sent to Short by
Marshall. Admiral Stark asked, later that morning, that the
message sent to Short should also be handed on to Kimmel.
Marshall
delayed sending this message for nearly two hours after he left Stark’s
office, thus making it too late to enable Short to go on an alert that
might frighten off the Japanese attack, and did not hand over his
message to be sent to Short until 11:50. He further assured its late
delivery by refusing to use the quick methods provided by his scrambler
telephone connection with Short or the more powerful Navy and F. B. I.
transmitters which were offered to him. It was sent by Western Union at
12:17 from
Washington
to
San Francisco
and from there to Short at
Fort
Shafter by R. C. A., not even marked urgent, with the results
noted above.
Both
versions of
Marshall’s
conduct on the morning of the 7th agree upon the content of the
“warning” message Marshall finally sent to Short and make it clear why it has been
dubbed the “too-little-and-too-late” message. This is the message:
Japanese are presenting at one p.m. Eastern Standard Time
what amounts to an ultimatum also they are under orders to destroy their
Code machine immediately. Just what significance the hour set may have
we do not know but be on the alert accordingly. Inform naval authorities
of this communication.
Marshall
There
was nothing in the message to indicate any immediate emergency for Pearl
Harbor or that there was any knowledge or conviction at Washington that
the Japanese might be attacking Pearl Harbor within about an hour. Marshall deliberately deceived Short in telling that the
significance of the Time of Delivery message was unknown. Even Admiral
Samuel E. Morison admits this. Actually, when
Marshall
read over the fourteen-part reply to
Hull and the Time of Delivery message, his associates in his
office state that he exclaimed, “This means immediate war!” There was
no such interpretation, even by way of implication, in what he sent to
Short, and the same message was to be transmitted to Kimmel. That
Marshall was capable of sending a clear and incisive warning when he
wished to do so is shown by the message he sent to General Herron, the
commander of the Hawaiian District, on June 17, 1940. In what was
little more than a practice alert to impress the Japanese Marshall
ordered Herron to: “Immediately alert complete defensive organization to
deal with possible trans-Pacific raid.”
The
defenders of Roosevelt and Marshall have contended that
Marshall knew nothing of the fourteen-parter or the Time of
Delivery messages until he read these in his office after 11:25 on
Sunday morning. To accept this requires the utmost credulity, even
naivete. If he did not know about them, then this proves carelessness
and callous indifference to his official duties quite sufficient to
justify his dismissal from office as Army chief-of-staff. He needed to
be well informed about these documents just as much if he were to
practice clever deception as though he were doing his duty in getting
the facts and cooperating with Short in meeting the Japanese attack. Roosevelt would not have allowed him to get out of touch with Army
Intelligence or the Signal Corps. There is no reasonable doubt that
Marshall, informed of the Pilot Message, had arranged for receiving
these messages, had read them, and was fully informed by the time he
reached his office or Stark’s on Sunday morning whatever time that was.
This is the only reasonable explanation of why he sought an early
conference with Stark. Those who have sought to indicate otherwise are
better known for their proclivity to cover up the facts in this
situation than for their zest for revealing the truth. The only
reasonable motive for
Marshall’s disappearance would have been to make himself
inaccessible to those who might plead with him to send a warning to
Short and Kimmel.
Another
important qualification bearing on the validity of the traditional
account of
Marshall’s conduct on the morning of December 7th has been pointed
out by Commander Hiles. He made a very careful estimate of the time
which would have been required for
Marshall to have done all the things he is stated to have
accomplished in the twenty-five minutes between 11:25 and 11:50 in his
office on the morning of December 7th, and conservatively concluded that
it would have required at least two hours.
There is
no space here to go into the complicated question of where Marshall was
from mid-afternoon of the 6th until 11:25 on the morning of the 7th,
when he is represented as reaching his office in the Old Munitions
Building to examine for the first time the messages that had come in
from Japan during this interval. I sought to do this as well as
possible in the Pearl Harbor Supplement published by the
Chicago Tribune on December 7, 1966. It is obvious that
one must choose between the traditional legend and the statements of
Admiral McCollum, made during the post-Pearl Harbor investigations, and
before a luncheon at the Army and Navy Club in Washington on May 3,
1961, as reported in the notes of Admiral John F. Shafroth that were
twice checked and confirmed with minor revisions by Admiral McCollum.
They are also supported by the statement of Colonel (now General) Deane
that he saw Marshall at his office about 10:00 A. M. on the morning of
the 7th, and by a direct report to me by Professor Charles C. Tansill
who was present at the meeting on May 3rd.
It
obviously makes a great deal of difference, factually, whether Marshall
was off on a horseback ride during the morning of December 7th and did
not reach his office until 11:25, where he first saw the reply to Hull
and the Time of Delivery message, or whether he came to Stark’s
office shortly after 9:00 on the morning of the 7th, discussed the reply
to Hull and the Time of Delivery message with Stark, McCollum, Noyes and
Wilkinson, formulated there the warning to be sent by him to Short
before 10:00 A.M., but delayed sending it until 11:50 that morning.
Personally, I prefer to accept the statements of Admiral McCollum as
being far better supported by documentation and circumstantial evidence
and motivated only by a courageous desire to establish the truth. Most
of the documentation supporting the traditional story has been destroyed
or kept a close secret. On December 17th General Sherman Miles, chief
of Military Intelligence, prepared an honest account of what went on in
Marshall’s office on the morning of December 7th and showed it to
Marshall. It made the latter furious and he banished Miles to the post
of military observer in
Brazil and allowed him to stay in the service on condition of
making no further revelations. Later on, Marshall summoned the officers
who were acquainted with the facts to a room, locked the door, walked
around the room, shook hands with each of those present, and told them
that the facts relating to the events of December 6th and 7th and
associated developments must remain a secret with them “to the grave.”
One of those present decided not to have this situation on his
conscience until he reached the grave and revealed the facts to
Professor Tansill and myself. Whatever
Marshall
did on the morning of December 7th, it was all too late for any
effective warning of
Pearl Harbor.
It is
probable that even revisionist historians have erred in putting
exclusive emphasis on the Time of Delivery message as the sole or best
basis for deciding that the Japanese would attack Pearl Harbor on the
morning of the 7th. Intelligence experts like Bratton and McCollum did
this only by clever guessing and logical inference. But the Kita
message, which was ready in all essential parts for distribution by 2:30
P.M. on the afternoon of the 6th, left nothing to guesswork. Kita’s
complex system of signals to be passed back to the approaching Japanese
task force was to end on the night of the 6th, clearly implying that the
task force was expected to be arriving at Hawaii by the night of the
6th, organizing off
Oahu
and put in readiness for the attack the next morning.
Such
was the situation in
Washington. There was an impressive accumulation of evidence by the
morning of December 7th which made it certain that war with
Japan was coming in a matter of a few hours, with every
probability that the attack would be made on
Pearl Harbor.
Even as early as December first, it was probable that war was about to
start somewhere, and by December 4th it was certain that Japan would attack the
United States. Surely, by then, it was mandatory to warn Short and
Kimmel in clear and definite fashion. If they had been so informed on
the 4th they would have taken steps to go on an effective alert that
would have led the Japanese task force to turn back. It was not until
December 5th that Tokyo sent its vital radio message directing Admiral
Chiuchi Nagumo (who commanded the Japanese task force) to “climb Mount
Niitaka,” which meant that he was to proceed to Pearl Harbor with no
further delay or interruption unless negotiations were resumed.
Of
course, Short and Kimmel should have been told of the negotiations with
Japan in November, 1941, and warned when Tojo began to set
deadlines for the end of these negotiations, notably after he set
November 29th as the date when they must be settled unless “things were
automatically going to happen.” But neither Kimmel nor Short received
any warning whatever of an impending Japanese attack on
Pearl Harbor
until the Japanese bombing planes appeared over the fleet and the
military establishment about 7:50 on the morning of December 7th.
Forward to Chapter IV
Back to Chapter I and Table of Contents
Back to Barnes Page
