Every dozen years or so a new book comes out about Pearl Harbor. Some of these books merely tell how the attack succeeded. The more interesting ones seek to explain why. Why was it possible for a far-off country to surprise the mighty United States and sink some of its most powerful warships? The puzzle has been deepened by the knowledge that the United States had been breaking some of Japan’s codes and reading some of its secret messages.

Orthodox historians argue that Japan had cloaked its attack in such complete secrecy that no form of intelligence then used by the United States could have penetrated it. Revisionists offer a different answer. The attack succeeded, they say, through treachery—at the highest level. President Franklin D. Roosevelt is the traitor. They argue that Roosevelt, eager to get America into World War II to save Britain and defeat Hitler, needed an enemy attack on American forces to unite the nation. To ensure that such an attack would succeed, he and his subordinates withheld intelligence from Admiral Husband E. Kimmel, commander-in-chief of the Pacific Fleet, and Lieutenant General Walter C. Short, commanding general of the Hawaiian Department. Roosevelt thus was responsible for the deaths of 2,400 Americans and the sinking of eleven warships to get his war.

Among the revisionists are some distinguished historians: Charles A. Beard, a former president of the American Historical Association, for example, and John Toland, a Pulitzer Prize-winning author. The newest, most ambitious revisionist author, Robert B. Stinnett, a former photographer on the staff of the Oakland Tribune, is not in their class. He has spent a decade and a half on Day of Deceit but has come up with the most irrational of the revisionist books.

Stinnett posits a conspiracy so immense as to dwarf anything the earlier revisionists had proposed. One theory, for example, required only that the chief of naval operations sneak into the office of a subordinate, find in his files an intercept revealing the coming attack, and destroy it. Stinnett contends, however, that many naval officers passed documents to Roosevelt and his advisers while keeping them from Kimmel; they then concealed their acts from congressional investigators and historians—until Stinnett unearthed the conspiracy. He maintains that newly released documents, new interviews with aging survivors, and government suppression of papers support his view. But he misreads the record, misunderstands intelligence, mishandles facts, and misdirects readers.

One expert on communications intelligence found twenty-three pages containing technical errors in the first third of Stinnett’s book before publication, but the author refused to correct any.1 Another concluded a detailed review of Stinnett’s book in the journal Cryptologia, by saying,

To those of us who are familiar with Japanese naval codes and communications procedures at the time, available documentation in the Pearl Harbor arena as well as the pertinent personnel and history of OP-20-G [the Navy’s communication intelligence center], it is abundantly clear that the book fails to prove any part of its massive revisionist conspiracy theory.2

Central to the surprise was the radio silence of the strike force. The Japanese, commanders and radio operators alike, say unanimously that they never transmitted any messages whatever, not even on low-power ship-to-ship messages. Except for Homer “Charlie” Kisner, an American intercept operator now near ninety, everyone else who was listening for Japanese messages says the same thing.3 And the naval communications intelligence summaries produced in Hawaii have only one statement to make about the Japanese aircraft carriers after November 26, when the strike force sailed: “Carriers are still located in home waters.” On December 3, in the last mention before the attack, the summaries say, “No information on submarines or Carriers.”

But Stinnett says the strike force did transmit at least one message and that US naval intelligence heard and located it. The US Navy did this by the well-known procedure of radio direction-finding. In this method, radio receivers determine the direction in which a transmitter lies by rotating and listening, just as people turn a portable radio to get the best reception. If two receivers report their bearings to a central listening post, it can draw lines on a map to locate the transmitter at the point where the lines intersect. Stinnett claims that bearings were taken from the Philippines and Alaska and that the fix or fixes were transmitted to Hawaii. He does not give a precise enough citation for researchers to find such reports in the million or so documents in the relevant section of the National Archives’ Record Group 38, containing materials from the Navy. But the Hawaii summaries contain no such fixes. Stinnett contends that they do, but he misleads the reader.

For example, he says that the intelligence summary of November 25 told Kimmel that “a large Japanese force of fleet subs and long-range patrol aircraft was heading eastward toward Hawaii from Japan.” He cites Box 41 in the records of the Pearl Harbor Liaison Office, Record Group 80, and claims it is a newly discovered document, although it, or a carbon copy of it, is photoduplicated on page 2629 of Part 17 of the Joint Congressional Committee hearings on Pearl Harbor. But the summary says not what he claims it does, only that “Fourth Fleet is still holding extensive communications with the Commander Submarine Fleet.” It refers in no way and in no place to any messages from any forces heading toward Hawaii. Moreover, the previous day’s summary placed the Fourth Fleet in Truk, far south of Japan and Hawaii. Stinnett also says that, according to a summary of December 5, the Japanese commander-in-chief, Admiral Isoroku Yamamoto, “originated several [radio] messages to the Carriers” (Stinnett’s brackets). The summary he cites mentions no such messages. His book contains many other such misstatements.


Moreover, Stinnett seems unaware that a single bearing does not fix a vessel’s location. The line of bearing from the Philippines runs not only through the Kurile Islands north of Japan, from which the strike force sailed, but also through the home waters of the Japanese fleet. So it cannot be said to have located the strike force.

Stinnett states that before Pearl Harbor the United States had broken the main Japanese naval code, the five-numeral system, later called JN 25 B. This codebook replaced Japanese words and phrases with groups of five digits, so that, say, “Tokyo” would become 43181; to these were added key digits, such as 52001, and the sum, 95182, was transmitted as the cryptogram. Deciphering the code, he says, enabled the United States to read messages from the Japanese naval command ordering the strike force to attack. As one example, he cites an intercept of December 6 beginning, “A special message on the occasion of the Declaration of War.” Stinnett thinks that it was broken before the Pearl Harbor attack the next day and should have been interpreted by the US as revealing the impending attack. The intercept is one of tens of thousands of recently declassified Japanese naval intercepts. They were read in 1945 and 1946 by American cryptanalysts awaiting demobilization. All bear such annotations similar to the one made on the December 6 intercept: “Navy Trans. 10/14/45.”

Stinnett refuses to accept this plain statement. He contends that the dates have been falsified as part of the conspiracy to cover up Roosevelt’s culpability. But the month-by-month reports of the Navy’s cryptanalytic center, declassified in December 1998, discovered by Stephen Budiansky and used by him in his new book, Battle of Wits, 4 tally the number of code groups deriving from the five-numeral system that the Navy recovered each month. They show that by December 7 only 3,800 of the code’s estimated 30,000 groups had been recovered—and many of the 3,800 stood for numerals. This was not enough to read more than fragments of each message and could not produce any significant intelligible messages. A wartime report gives the number of messages read in this code in 1941 as “none.” All this fits perfectly with the unanimous statements by navy codebreakers that no five-numeral messages were read before Pearl Harbor.

According to Day of Deceit, the Japanese order for war, “Climb Mount Niitaka December 8,” was transmitted in uncoded Japanese, even though the photocopy of the intercept shown in the book plainly shows the designation “JN 25 B” and even though the Japanese preceded the directive with “This despatch is Top Secret.” Stinnett then charges that “deceit took over” after the message was intercepted: Lieutenant Commander Edwin T. Layton, Kimmel’s intelligence officer, denied ever receiving it, and it “remained in locked vaults.” The intercept operator, he writes, was never called to testify by anyone investigating the Pearl Harbor debacle.

Layton was only one of the many involved in Stinnett’s Pearl Harbor conspiracy. Among the others who illegally withheld information from Kimmel were not only Roosevelt and Admiral Harold Stark, chief of naval operations, but also Admiral Walter Anderson, the battleships commander at Pearl Harbor; Captain Irving Mayfield, the intelligence officer for the 14th Naval District, in Hawaii; Lieutenant Commander Thomas Dyer, chief cryptanalyst in the Navy’s Hawaiian communications intelligence unit; and Lieutenant Commander Arthur McCollum, the Far Eastern specialist in the Office of Naval Intelligence in Washington.

The list of suspects also includes, among other specialists in communications intelligence, Lieutenant Commander Joseph J. Rochefort, head of the Hawaii communications intelligence unit, even though he often met with Kimmel, who initialed many of Rochefort’s summaries. Of Mayfield, Stinnett writes, “For the six days prior to the attack, Captain Mayfield supervised the handling of the coded radiograms. He had one paramount responsibility, to get the intercepts to Admiral Kimmel. His failure to do so has escaped scrutiny.” According to Stinnett, this huge conspiracy has remained secret until he uncovered it.


Stinnett rests his argument that Roosevelt wanted to provoke the Japanese into firing the first shot on a memorandum that he says he found in McCollum’s personnel file. Dated October 7, 1940, and addressed to Anderson, at the time director of naval intelligence, and Dudley W. Knox, chief of the ONIlibrary, it “suggested” giving all possible aid to China and embargoing “all trade with Japan,” among other proposals. “If by these means Japan could be led to commit an overt act of war, so much the better,” McCollum wrote. Stinnett does not mention or seem to realize that McCollum’s points about aid to China and the embargo of Japan reflected longstanding American policies in support of China and opposition to Japan’s aggression and fascism. Stinnett admits that he has no record of Roosevelt’s having seen McCollum’s document, but says that the fact that its eight points were put into effect—as most of them were—proves that Roosevelt did see it and follow it. But it is much more likely that McCollum was following national policy and adding his own view on the risk of war than that Roosevelt was taking guidance on American policy from a mid-level Navy officer.

Aside from its basic errors of fact and its tendentious interpretations, Day of Deceit is an extraordinarily sloppy book. Topics that should stand together are separated and information is occasionally repeated in the space of a few pages. Citations are often inadequate, and Stinnett sometimes contradicts himself. For example, after saying repeatedly that Kimmel was denied codebreaking intelligence, he says that such information was sent to him on “apparently the slowest boat in the Navy.” Moreover, in mentioning Roosevelt’s last-minute appeal to Hirohito, he writes that the American ambassador was not granted an audience—implying that the message never got to the emperor. He doesn’t say that it was in fact delivered in an audience granted by the foreign minister at 3 AM. He calls “F-2” McCollum’s “code name,” when it merely designated his office: OP-20-F-2. He says letters were engraved on the rims of the code wheels of a Japanese cipher machine: no code wheels ever existed. He claims that Rochefort’s description of the Japanese naval code called “AD” by the Americans “appears to be a cover story.” It’s not. He misreads the date of 15-5-41 on an illustrated document as December 5, 1941. These and many other blunders discredit his work.

It is further undermined by three of its assumptions. One is that the plot would work. Any of the naval officers who were more loyal to Kimmel than to Roosevelt could have slipped information to him. If the army lieutenant supervising a radar on the morning of December 7 had not told the two operators who spotted the approach of the Japanese airplanes to “forget it,” surprise would have been lost. And if Kimmel and Short had done their jobs despite less than perfect knowledge and driven off the attackers, the conspiracy would have failed.

Another erroneous assumption is that Roosevelt needed the destruction of major American forces to get into war. A Japanese attack on a few antiquated battlewagons left as bait would have achieved Roosevelt’s alleged goal. And although the deaths of men and the sinking of ships appalled Americans, it was the sneakiness of the attack while negotiations were still underway that enraged them.

Finally, Stinnett is simply wrong when he writes, “If Japan could be provoked into committing an overt act of war against the United States, then… mutual assistance provisions [of the Axis Tripartite Pact] would kick in.” They wouldn’t. The pact—it is in English in the original—reads in part:

Germany, Italy and Japan agree to co-operate their efforts on the aforesaid lines [the distribution of spheres of influence]. They further undertake to assist one another with all political and economic military means when one of the three Contracting Parties is attacked by a power at present not involved in the European War or in the Sino-Japanese Conflict.

The treaty did not require assistance if one of the parties mounted an attack on another nation. And in fact Japan had not attacked its old foe Russia when Hitler invaded the Soviet Union. So Japan could not be sure that Germany would fight on Japan’s side after Pearl Harbor. When Hitler, who was surprised by the event, declared war on December 11, he did not do so because he was complying with the treaty. His reasons have never been made clear and remain a subject of controversy among historians. Roosevelt could not be certain of his reaction any more than the Japanese were.

Stinnett does not see what nearly all other students of the attack agree on: Pearl Harbor succeeded because of Japan’s total secrecy about the attack. Even the Japanese ambassadors, whose messages the United States was reading, were never told of it.

The ultimate problem with his book, as with most conspiracy theories, is that it refuses to admit that it can be in error. Conspiracies certainly exist, but those who propound a theory must be open to evidence that can prove them wrong. Otherwise rational argument cannot take place. Unfortunately Stinnett is such a passionate believer in conspiracy that he is unwilling to consider the countervailing evidence. He is offering not a theory but a definition, which cannot be contradicted. It may be expected to have the same fate as all of the other Pearl Harbor conspiracy arguments made during the last fifty-five years.

This Issue

November 2, 2000