Pearl Harbor in Retrospect

"Wherever or whenever Washington may have thought the Japanese cat would probably jump, Hawaii's primary mission was to meet it there if it came. Yet both the Army and Navy commands there acted as if there were no chance of a Japanese overseas attack on them. What they actually did and did not do, simply spelled 'It can't happen here.'"

The situation that existed in Hawaii in November and early December, 1941, presented a much less difficult problem to the commands there than had many of the countless studies and maneuvers by which the Army and Navy had for years been indoctrinated. Often our assumed situations envisaged a Japanese attack following a very short period of strained relations, or none at all—"out of the blue." Frequently our naval force, assumed to be in Hawaiian waters, consisted only of locally based submarines and air squadrons, supported by an unreinforced garrison. And rarely, if ever, did the problems offer the lead of specific warnings or directives from the War and Navy Departments.

The Pearl Harbor attack, on the contrary, followed a long period of strained relations. We had progressively imposed economic pressure on Japan. They had sent one of their leading diplomats (through Hawaii, incidentally) as the crisis approached in the long-drawn-out conference in Washington. The major part of our battle fleet was in Hawaiian waters. Our Hawaiian garrison had been materially reinforced. And lastly, well before the situation broke, the Army and Navy commanders there had received from their chiefs directives which contained clear warnings of the possible outbreak of hostilities at any time. Had such a situation been suggested as the basis of a theoretical war game or maneuver, it would probably have been rejected out of hand. For it would have been held to present no problem for solution: the answer would have been too obvious—full alert and the activation of approved plans.

I have said that the War and Navy Departments' dispatches contained clear warnings of possible hostilities. I think the record will bear me out. Let's look at it.

As early as July 25, when we froze Japanese assets, the Pacific commands, including Hawaii, were informed of it by the War and Navy Departments "in order that you may take appropriate precautionary measures against any possible eventualities." On November 24, a joint Army and Navy dispatch pointed out the possibility of Japanese "surprise aggressive movements in any direction." On November 27 the Navy Department sent another dispatch beginning: "This is to be considered a war warning"—not much doubt about that. On the same day the War Department sent another one, over General Marshall's signature, and Military Intelligence followed it up with a message to G-2's.

The Marshall dispatch read in part: "Japanese future action unpredictable but hostile action possible at any moment. If hostilities cannot, repeat cannot, be avoided the United States desires that Japan commit the first overt act. This policy should not, repeat not, be construed as restricting you to a course of action that might jeopardize your defense. Prior to hostile Japanese action you are directed to undertake such reconnaissance and other measures as you deem necessary but these measures should be carried out so as not, repeat not, to alarm civil population or disclose intent. Report measures taken."

This dispatch has since been criticized as a "do-don't" order. In its drafting, Military Intelligence had no part, and I have no personal concern with the "do-don't" controversy. But aside from a certain obscurity about not disclosing intent, the "don'ts" were nothing to get excited about—don't start a war; don't alarm civilians. Those were old Army policies. The importance of the message lies in what it was meant to convey and what it did convey—to others.

It was drafted under the personal supervision of the Secretary of War, who had in mind that "defense against an attack by Japan was the first consideration." General Marshall later defined the dispatch tersely as "a command directive for alert against a state of war." It is indeed difficult to miss a clear war warning in the phrases of the message itself. The Philippines, Panama, and the West Coast received the same or a very similar dispatch, had no doubts about its intent, and acted accordingly. Only Hawaii, the vital pivot in a Japanese war, thought that such a warning had but slight local application.

Wherever or whenever Washington may have thought the Japanese cat would probably jump, Hawaii's primary mission was to meet it there if it came. Yet both the Army and Navy commands there acted as if there were no chance of a Japanese overseas attack on them. What they actually did and did not do, simply spelled "It can't happen here."


It has since been implied that the reason Hawaii was not on the alert was that Washington thought the Japanese would not attack there. That suggestion points up very neatly the crucial issue. For the opposite was true—Washington thought the Japanese would not attack Hawaii largely because it believed Hawaii was alerted and prepared.

That was, admittedly, an assumption, but it was so fundamental an assumption, based on so many years of indoctrination, as well as on issued orders, that it was not questioned by anyone in Washington, from the President down. For guns don't shoot or planes fly by themselves.

General Marshall, testifying before the Congressional Committee, was asked: "Did the President of the United States, in your opinion, have a right to assume that the commands in Hawaii were properly alerted on the morning of December 7?" The General replied: "I think he had every right to assume that." Washington was surprised by the Pearl Harbor attack, but not nearly so much as it was by what the enemy encountered there.

For the War Department had actually cut into the seed corn of air power to strengthen Hawaiian defense. What could be spared only at great detriment to other commands, in planes, in antiaircraft-artillery, and in radar equipment, went to Hawaii. The Philippines were left, for that period, practically helpless, and even the vital artery through Panama was neglected—until Hawaii was given what could be had. Then, and only then, about August, did Washington begin to build up Panama and the Philippines. Even so, when Pearl Harbor came, Hawaii was much better equipped for defense than either of the other two great outposts or our own West Coast itself.

But Hawaii had lowered its guard to "alert against sabotage" on land and "condition 3" afloat. On this point, General Marshall testified: "I never could grasp what had happened between the period when so much was said [in Hawaii] about air attack, the necessity for antiaircraft, the necessity for planes for reconnaissance, the necessity for attack planes for defense and the other requirements which anticipated very definitely and affirmatively an air attack—I could never understand why suddenly it became a side issue."

It has been suggested that the high command in Washington neglected an essential of Hawaiian security, unity of command. That was not neglect: unity was a practical impossibility in those pre-war days. For years Hawaii had been the Exhibit A of both the proponents and the opponents of unified command. Three weeks before Pearl Harbor the last effort to attain unity failed. Only an executive order or war could have imposed it. After the war, with all of its lessons behind us, it still took two years to make possible unity of command (and of other functions) through merger of the Army and the Navy—if indeed they have merged!

Admiral Halsey, in his book, lays the blame for Pearl Harbor on Congress, for its failure to appropriate funds necessary for adequate protection. That is rather farfetched, since the issue clearly turned, not on what the Hawaiian commands had, but on what they did and did not do with it. They were not even planning to use it—then. And Admiral Halsey's description of the radical measures he himself took to forestall an air surprise, on his cruise to Midway, just before the attack—his famous "shoot first and argue afterwards"—comports strangely with the almost total unpreparedness in Hawaii to shoot at all.

It has also been pointed out that Military and Naval Intelligence estimates on the future course of the World War, made in the months preceding Pearl Harbor, fail to mention a possible Japanese attack there. That is quite true. But suppose we had told Hawaii that the Japanese might attack—would that have been news? Why—for what primary purpose was the garrison there, and against what single potential enemy? Why was the fleet there? What did our long-strained relations with Japan mean to the key of the Pacific? And what of our years of emphasis on Hawaiian security in the strategy of a Japanese war? Does a guard, as the opposing team goes into a huddle, have to be told: "This play may come through you"? I willingly assume my share of responsibility for the omission of the obvious.

Emphasis on the danger of sabotage has also been advanced to explain why Hawaii was not alerted against an overseas attack. But actually how strong was that emphasis? True, subversive activities had been mentioned in the message to G-2's of November 27, which read: "Japanese negotiations have come to practical stalemate. Hostilities may ensue. Subversive activities may be expected. Inform commanding general and chief of staff only. Miles." Heaven knows subversion was an issue in Hawaii, with its large Japanese population; but the crucial sentence in the message was certainly "Hostilities may ensue." So far as Hawaii was concerned, warning of Japanese hostilities meant the possibility of attack; and that danger, however problematical, exceeded all else in importance.

There was also a dispatch sent in duplicate by the Chief of the Air Corps and the Adjutant General as a precaution against sabotage of planes. But none of these dispatches superseded or modified the War Department's warning order, nor were they so interpreted by any command that received them. Even Hawaii regarded them only as supporting the decision it had already made to go on alert against sabotage only.

The War Department has been pilloried for failure to tell Hawaii that its alert did not meet either the situation or the intent of General Marshall's order. That order had required a report of action taken. The Hawaiian command reported: "Alerted to prevent sabotage. Liaison with Navy"—nothing more. The War Department did not reply. Admittedly, this was a serious oversight, for which senior officers have assumed responsibility. But to what extent does the War Department's failure to reply justify the retention of Hawaii's inadequate alert up to the time of the attack? On that matter I must again speak as an outsider, since it was not a function of Military Intelligence to check the readiness or any other disposition of United States forces, nor did I, or anyone else in Military Intelligence, see General Short's cryptic report.

Secretary Stimson says: "My initials show that the report crossed my desk, and in spite of my keen interest in the situation, it certainly gave me no intimation that the alert order against an enemy attack was not being carried out." For General Short did not say that he was alerted against sabotage only. He did not say that he considered the possibility of immediate war as solely an internal and not an external threat to his command, and give his reasons. Least of all did he imply any request for confirmation of so extraordinary a decision as an alert which ignored the possibility of attack by Japanese armed forces.

But Washington's failure to grasp the situation in Hawaii and correct it goes beyond the incident of General Short's report. Nine days elapsed before the attack. Why was Washington unaware, during those critical days, that Hawaii was out of line?

The answer lies in the system of decentralization of command which had for years prevailed in the Army and the Navy. General Marshall used to say that the War Department was a very poor command post. It was hardly one at all, save on the highest plane of general directives. So far as Hawaii, Panama, and the Philippines were concerned, the War Department was the office which assigned their personnel, provided their material, and gave them the general directives under which they operated. Within that scope those outposts were independent commands.

The message to G-2's of November 27, quoted above, was not an order. Rarely did the Chief of Staff issue orders to the Pacific outposts. The fact that the "command directive" of November 27 went out over his name added emphasis, if such were needed, to that crucial dispatch.


When war came, our "magic"—the breaking of Japanese codes—paid enormous dividends. It materially aided us in concentrating those slender means by which we won the Battle of Midway, the turning point in the Pacific war. It has always seemed to me that we were extremely lucky in keeping the vital secret of "magic" within an already fairly large group in Washington, and wise in rigidly limiting it to that group and the Philippine commands until we were actually at war. I well remember a day when a copy of one of the "magic" messages was missing. The Secretary himself got into that fracas, and I finally ran down the message in the possession of a person unauthorized to have it—in the White House!

After the war, "magic" supplied what appeared to many as definite indications of what the Japanese had been planning at Pearl Harbor. We had those indications before the attack. Why, then, didn't we foresee it? The question hinges on selection, guided by hindsight. If one reads first the end of a good detective story, and then starts in at the beginning and reads through, it is easy to pick out the real clues from those which would have led to other deductions. It is not so easy if one takes the clues, true and false, as they come. The "magic" intercepted and translated by us in the six months before Pearl Harbor, if printed in book form and type, would make several normal volumes. There was no lack of clues—a broad field from which to select, after the event, those which seem to point to that event and to that only.

There were many "magic" messages showing Japanese interest in conditions existing in Hawaii, largely requests for information of military value. Some concerned the location of anchorages of our warships in Pearl Harbor, by limited sections of that area, their arrival and departure, and so forth. These messages were primarily of naval interest, and the Navy apparently took them to mean two things: first, that Japanese spies there were looking down our throats—a deplorable condition which the Army and Navy had known for thirty years or more; and second, that the Japanese were planning an attack on our fleet, by air or submarines or both.

But since the fleet might eventually be a deterrent to them, it would have been strange indeed had they not made plans to attack it if they could. We ourselves had plans for contingencies far less obvious than that. It is difficult to believe that any senior Army or Navy officer in Hawaii would have found it news had he been told by Washington that the fleet was under close Japanese espionage and the subject of aggressive planning. Indeed, a naval officer would have replied that the fleet had no intention of meeting a major attack at their moorings in Pearl Harbor!

The Hawaiian commands later complained that this "magic" information was not transmitted to them—this in spite of their failure to react to the authoritative warning orders sent them when the situation was commonly known to be far more critical. By comparison, it may be noted that General MacArthur, who had access to "magic," could not later identify the more important "magic" messages; he apparently took no action on them, but alerted his command for war on Washington's warning orders.

There were two "magic" messages of a different type which have subsequently been held to have been signposts, had we so read them, to Pearl Harbor. The first was a message of November 22 from Tokyo to their ambassadors in Washington, saying: "After that [November 29] things are automatically going to happen." The other message, of November 30, was from Tokyo to the Germans, informing them of the danger of sudden war "through some clash of arms."

The statement that "things are automatically going to happen" after November 29 simply meant that certain preliminary movements of forces and supplies, without which operations in modern war are impossible, would, on that date, be irrevocably committed. Such commitments would, for instance, be essential to the Japanese southern advance, then apparently (and actually) in preparation. The point is that Japan had a wide choice of victims; and considered as a clue to the Pearl Harbor attack, the statement that something would happen automatically somewhere after the 29th is pretty thin.

Presented by

Join the Discussion

After you comment, click Post. If you’re not already logged in you will be asked to log in or register with Disqus.

Please note that The Atlantic's account system is separate from our commenting system. To log in or register with The Atlantic, use the Sign In button at the top of every page.

blog comments powered by Disqus