July 24, 2000

The Myth of Perfect Nuclear Security

By RICHARD RHODES

ADISON, Conn. -- When the Los Alamos National Laboratory opened for business in April 1943, the irony of volunteering to live behind barbed-wire fences was not lost on the Jewish physicists who had escaped from Nazi Germany and would now lead the development of the world's first atomic bombs. Understandably, they found the implication that they might be disloyal to the United States offensive.

Despite aggressive security, which extended to mail checks, telephone tapping and covert surveillance, two Manhattan Project scientists independently passed the detailed plans for the Fat Man plutonium implosion bomb to the Soviet Union late in the war; the first Soviet atomic bomb was a close copy of the American bomb.

Paradoxically, the transfer of the plans by Klaus Fuchs, a German, and Theodore Alvin Hall, a young American, delayed rather than accelerated the Soviet weapons program. Russian scientists under Igor Kurchatov independently designed a weapon twice as powerful and half as large as Fat Man, but Joseph Stalin and Lavrenti Beria, who was the head of the Soviet bomb program as well as the K.G.B., distrusted the very scientists they relied upon and insisted on building the cruder but "proven" American version.

Today, Los Alamos is once again defending itself against charges of lax security. The laboratory's security system certainly needs a tune-up, but fears of espionage are overblown. Worse, the remedies being proposed threaten a more fundamental kind of national security: the ability to attract and keep a staff of talented, dedicated scientists and engineers.

When it comes to developing nuclear weapons, there are no real secrets anymore. The technology reached maturity decades ago. Building a bomb is a matter of money, industrial resources and bad judgment; every nation that has undertaken to build an atomic bomb has succeeded on the first try. The only real differences today from nation to nation concern skill at making weapons smaller and more efficient, a technological disparity that more robust delivery systems can compensate for.

Weapons scientists know this fundamental truth, even if senators don't. It's the reason scientists chafe at overzealous security restrictions -- and the reason some of them devise ways to work around these restrictions. As long ago as 1970, a Defense Department task force on secrecy judged it "unlikely that classified information will remain secure for periods as long as five years."

Thus, the indictment of Wen Ho Lee, a Chinese-American physicist at Los Alamos now awaiting trial, for downloading old weapons-design files onto an unsecured computer has already damaged security more than it can possibly protect it.

Mr. Lee was charged with security violations only because the Federal Bureau of Investigation was unable to document its working assumption -- that he had passed weapons secrets to China. But what would he have told China's scientists that they don't already know?

Mr. Lee is not the first government employee to find secure computers frustrating and inconvenient. John Deutch, former head of the Central Intelligence Agency, also loaded classified information onto his unsecured home computer. (He, however, has not been subjected to arrest and extended incarceration for doing so.)

Moreover, blame for the other Los Alamos security scandal -- the temporary misplacement of two hard drives that almost certainly contained information about the electronic locks that prevent unauthorized use of nuclear weapons -- certainly deserves to be shared.

In 1993, President George Bush authorized a cost-saving program that eliminated accountability for merely "secret" material. And the Pentagon last year blocked a proposal by former Energy Secretary Hazel O'Leary that would have declassified many of the department's files, while securing the most sensitive nuclear weapons information.

Under Ms. O'Leary's proposal, the hard drives would have been upgraded from secret to top secret, which would have prevented them from being absentmindedly carried away and forgotten. Whoever wanted them would have had to sign them out and in.

Why was the proposal blocked? The Pentagon cited "substantial" costs.

Before we straitjacket our national weapons laboratories further with security restrictions like polygraph tests, which have already lowered morale and discouraged recruitment, we should consider their past record of accomplishment -- a better measure of loyalty than any number of mandatory polygraph tests. Across nearly 60 years, their work has contributed in major ways to our security and technological advantage.

Ten years after the physicist J. Robert Oppenheimer opened Los Alamos, he was subjected to a vindictive security hearing for the sin of advising his government not to waste its limited nuclear resources on a crash program to develop a hydrogen bomb it did not yet know how to build. One of his defenders before the security board, I. I. Rabi, the plainspoken Nobel laureate and Columbia University physicist, listed Oppenheimer's outstanding contributions to United States security -- first of all, the bomb itself -- and then asked the exasperated question I would ask of the present-day detractors of the Los Alamos lab: "What more do you want -- mermaids?"

Richard Rhodes is the author of "The Making of the Atomic Bomb," "Dark Sun" and, most recently, "Why They Kill."

Copyright 2000 The New York Times Company