Richard L. Garwin
IBM Fellow Emeritus
Thomas J. Watson Research Center
P.O. Box 218
Yorktown Heights, NY 10598-0218
(914) 945-2555
FAX: (914) 945-4419

January 7, 1997
(Via Email to letters at

Letters to the Editor
The New York Times
229 West 43rd Street
New York, NY 10036

Dear Editor:

NUCLEAR ARMS BUILDERS MOVE FROM CREATORS TO CUSTODIANS (01/06/97) features the big ticket items rather than the affordable basics. Having worked on U.S. nuclear weapons since 1950, and participated in several reviews for the Department of Energy in the last three years, I regret the "either-or" focus on the controversial, to the neglect of the truly essential. For more than 20 years I have been advising Presidents and the nuclear weapons establishment that, if desired, our nuclear weapons stockpile could be maintained reliable and safe for decades or even centuries, by technical surveillance and occasional remanufacturing to original specifications, without the need for nuclear explosion testing. That is still my view.

I have also argued that U.S. and world security would be served by the elimination of nuclear explosion testing, and in September 1996 the United States signed just such a total ban, which owed a lot to the leadership of the Clinton Administration.

The basic core of a reliable and safe nuclear stockpile is assiduous and perceptive surveillance as has been conducted for many years, together with not only the ability to remanufacture but the actual remanufacturing of weapons so that we will be able to observe the remanufactured weapons over the years. In a 10/28/96 letter to Secretary of Energy Hazel O'Leary, requested by the Department of Energy, Sidney D. Drell and I noted

The DOE decision document of 12/19/96 quoted this letter, but beginning with "we must not only maintain a cadre of first-class weapon scientists and engineers ..." thus perhaps implying that our support of the SSMP was based entirely on the first goal of "a safe and reliable nuclear weapons stockpile."

If one accepts the "two basic commitments" (as I do) then much of the program of new instruments and capabilities follow. However, to commit at this point every detail of that future program would prevent us from incorporating in future years approaches that might do a better job or do the same job at less cost.

Other aspects of the stockpile maintenance program deserve attention as well in the interests of economy and efficiency. For instance, U.S. nuclear weapons will need a supply of tritium beginning in the year 2005 or 2010. There is pressure from DOE contractors to build an accelerator to produce this additional tritium at a program cost estimated by DOE at $5.1 billion. To produce tritium in U.S. commercial reactors is estimated at $1.4 billion. But to buy the necessary tritium from Russia (in sufficient quantities so that one could build and begin to operate in a timely fashion the requisite replacement domestic tritium production capacity in case of default) is likely to have a program cost of only $0.5 billion. And tritium atoms are identical, whether produced in Russia or in the United States.

U.S. security is advanced if other nations do not test nuclear weapons. Britain, China, France, and Russia have made the judgment in signing the Comprehensive Test Ban Treaty that they will be able to maintain their own nuclear weapon stockpiles safe and reliable by their own version of a remanufacture and surveillance program, with their own version of advanced facilities at a more modest level. The test ban is essential if we are to retain the support of the rest of the world against the spread of nuclear weapons.

My own technical judgment is that it would be irresponsible to use the increased "science-based understanding" to design and build new versions of nuclear weapons, without nuclear testing, with a resultant decrease in confidence in our nuclear weaponry. But that understanding can provide enormous confidence that surveillance and necessary remanufacture will maintain our enduring stockpile of current-version nuclear weapons safe and reliable for many decades or even centuries, without nuclear testing.

Sincerely yours,

Richard L. Garwin