NYT, 5 April, 1996 Pentagon Spy Agency Bares Some Dusty Secret Papers By Tim Weiner Washington, April 4 -- The National Security Agency, the Pentagon spy service that eavesdrops on global communications, said today that it had declassified more than 1.3 million pages of secret documents, some from before World War I. All the declassified material is more than 50 years old, older than the agency itself, and represents a tiny fragment of the billions of pages of Government documents that have been kept secret on the grounds that their release would damage national security. Agency officials were at a loss to explain why these documents, now at the National Archives, had remained secret for so long. Among the documents declassified today is a January 1919 memorandum from CoL A. W. Bloor of the Army, a commander in the American Expeditionary Force in France, explaining the origin of the "code talkers," American Indian soldiers who spoke in their native tongues to confound enemy code breakers in World War I and World War II. Their languages were largely unwritten and largely unstudied by foreigners, and so constituted an instant code translatable only by the speakers. "The German was a past master at the art of 'Listening In,' " on radio transmissions, the memorandum says. "It was therefore necessary to code every message of importance and coding and decoding took valuable time." Then, Colonel Bloor wrote, he remembered that he had a company of Indians in his regiment who among them spoke 26 languages or dialects, and that "there was hardly a chance in a million" that the Germans could translate them. David Hatch, the National Security Agency's historian, said Choctaws, Navajos, Comanches, Winnebagos, Pawnees, Kiowas and Cherokees served as code talkers. In World War II, he said, the Marine Corps used more than 400 Navajos as communicators in the Pacific campaign. That story has been popularized by Hollywood films, documentaries and books. Mr. Hatch said he could not explain why the documents stayed secret for so long. The agency's archives run into the billions of pages, and the agency, loath to disclose anything concerning codes, has only begun to consider declassifying documents in the past four years. "We have so many pages and we've only been at it for a few years," Mr. Hatch said. "The interesting thing to me is that this is coming out. What was known only to insiders is now becoming known to historians and outsiders." ----- WSJ, 5 April, 1996 Secret Cables of '43 And the Hiss Case May I offer a distinction that may clarify a point in Eric Breindel's March 14 editorial-page piece "New Evidence in the Hiss Case?" The matter deals with the newly released Soviet cables dated from 1943 to the early Cold War, and intercepted and solved by the National Security Agency and its predecessors in a project called Venona. As Mr. Breindel states, "The single most interesting document in the new Venona batch is a March 30,1945, Washington-to-Moscow report on an agent whose cover name was 'Ales.' The cable was decrypted on Aug. 8, 1969, and the NSA glossary ... explains that 'Ales' is 'probably Alger Hiss.' " A distinction must be made between the test of the cables and the identification of the individuals mentioned in the text only by code name. The cables were cryptanalyzed. The internal cross-checks in this work make the likelihood of their being incorrectly solved all but zero. This means that the code names are almost certainly right. But the determination that a particular code name represents a particular individual did not come from cryptanalysis. It came from FBI field investigations. I have no reason to question their accuracy, but they stand on a different basis than codebreaking. This is why NSA qualified the Ales=Hiss identification with a "probably." David Kahn, Great Neck, N.Y. (Mr. Kahn was scholar in residence at the National Security Agency in 1995.) -----