Curatorship vs. Stewardship

Jonathan I. Katz

Department of Physics, Washington U., St. Louis, Mo.

[my last name]


[A heavily edited version of this appeared in the Bulletin of the Atomic Scientists, November/December 1995. It has also circulated informally under the title "The Case Against Science-Based Stockpile Stewardship".]

Nuclear weapons were developed under the intense pressure of a World War, and matured during the Cold War. It has been axiomatic that the end of the Cold War, and the absence of a major threatening adversary since then, means that the future task of the nuclear complex will be to exercise stewardship over existing weapons and materials, maintaining confidence in them, preserving the physical and human resources needed to remanufacture them if necessary, and remaining capable of resuming nuclear design, manufacture and testing with minimum delay should the international situation require.

I suggest that it is better to describe the future task as curatorship than as stewardship, and emphasize the distinction between these two concepts. In stewardship the human resources required to design and develop weapons are maintained, with skills honed on classified and unclassified experiments conducted at facilities such as the National Ignition Facility (NIF) and in hydronuclear tests. In curatorship these facilities are not built, and design and development skills are allowed to atrophy; only those skills required to remanufacture weapons according to their original specifications are preserved. The purpose of this note is to argue that curatorship is preferable to stewardship. The chief nuclear danger in the present world is that of proliferation, and stewardship will exacerbate this danger, while curatorship will mitigate it while preserving our existing nuclear forces.

What is the value of the capabilities preserved by stewardship, but not by curatorship? These are the retention of a cadré of technically and scientifically competent people who understand nuclear weapons design, developing their understanding and skills on weapons-related experiments. Such people are needed in order to design new weapons, or to modify old ones. However, nuclear weapons are not well enough understood to permit the development of new weapons, or the modification of those we now possess, without tests at substantial (multi-kiloton) nuclear yield. Despite fifty years of experience, including large numbers of tests at full nuclear yield, we do not have sufficient confidence in our design tools. It is unlikely that any future work without nuclear testing could give us that confidence.

The value of maintaining stewardship of our design capabilities can only be the enhancement of confidence in the answers of the design cadré when new and unanticipated questions arise. The history of such questions shows that even with skills honed by tests at full nuclear yield, designers frequently could not give confident answers to unanticipated questions for which no empirical data existed. There are too many matters of nuclear design and effects which are understood only empirically, and empiricism cannot be confidently extrapolated to answer new questions.

Forgoing stewardship would imply forgoing facilities such as NIF and the research they would perform. Although popular among scientists who would be their likely users, and at a Department of Energy whose mission is disappearing, their importance as scientific research is slight and the chance of practical benefit is small. The obligation under the Nonproliferation Treaty to share the peaceful benefits of nuclear research with non-nuclear states would therefore be met by curatorship; stewardship would lead to a continuing concern that supposed benefits had not been shared.

Against the limited value of stewardship must be opposed its costs. The construction and operation of NIF and related facilities would not be cheap. More important are the consequences for the present and future danger of proliferation. NIF will bring together the weapons and unclassified communities. People will rub elbows, share facilities, collaborate on unclassified experiments and communicate their interests and concerns to each other. Information and understanding will diffuse from the classified to the unclassified world, without any technical violation of security. The desire to achieve renown and career success by publication in the open literature will diffuse from the unclassified to the classified world.

Inertial (chiefly laser) fusion has similarly brought its classified and unclassified communities into intellectual and geographical contact over the last 25 years. The consequence has been the declassification of many nuclear weapons concepts and information. It is common knowledge that there is a great deal of physics in common between inertial fusion and nuclear weapons. The unclassified inertial fusion community has reinvented weapons technology, and the classified community has pressed successfully for declassification of formerly classified concepts, some applicable to inertial fusion and some not so applicable. Weapons-related terms and concepts which were unknown outside the classified world until a few years ago are now widely discussed (fortunately, not always correctly) in the open literature. This process would continue at a NIF, which would provide a facility and funding for the unclassified world to rediscover nuclear weapons physics and (implicitly) to develop the understanding and computational tools required to design weapons. This reduction of the barriers to proliferation of both fission and thermonuclear weapons is not in the national interest.

Stewardship and curatorship also differ in the matter of hydronuclear experiments; they might be part of stewardship, but would certainly not be included in curatorship. Again, the value of these experiments to maintaining confidence in our nuclear forces would be slight, because the configuration tested in a hydronuclear experiment is materially different from that in an actual weapon. Their chief value is as tests for nuclear design codes; without tests at multi-kiloton yield neither hydronuclear experiments nor codes tested against them will be sufficient basis for adopting a new weapons design or materially changing an existing one.

As the nuclear force ages, it is inevitable that confidence in it will erode. I have argued that, given a nuclear test ban such as the US has accepted de facto, even successful stewardship cannot remedy that erosion of confidence when faced with the first unanticipated questions of the form ``Has a certain component deteriorated to the point of making the weapon unreliable?'' or ``Does a newly discovered oversight in design, materials, or construction reduce our confidence in a weapon's proper operation?''. The cadré of experts preserved by a program of stewardship would not be able to give confident answers to these questions, because expertise cannot substitute for experiment, which must include tests at multi-kiloton yield. The fount of confidence is the original testing program, which can be drawn upon only by faithful remanufacture to the original tested specifications. Curatorship is sufficient to make possible this degree of confidence, and stewardship can do no better.

In the present world the nuclear threats are hardly threats at all: Russia, whose confidence in its forces will erode just as ours does, and which is not an enemy country in any case; China, which is similar, but with a much smaller force; and potential proliferators. Our confidence in our forces will always be overwhelming in comparison to that of a proliferator which has never conducted a nuclear test.

The world may change, and more serious threats may arise in the form of a country, either a present nuclear power or a proliferator, which is hostile, aggressive, and has a large force of tested nuclear weapons and delivery systems. In such a world neither stewardship nor curatorship would be adequate, and nuclear weapons development and testing would become a living enterprise again. It would then become necessary to reconstitute a nuclear design and testing capability; this is asserted to be more rapid from a base of stewardship than from one of curatorship, because stewardship promises a greater base of technically expert people on which to build. However, it is not necessarily true that stewardship would permit faster and better reconstitution; perverse effects, such as an excessive influence of experience obtained under the different parameter regimes of NIF or hydronuclear tests are also possible. The basic concepts and the heritage of design and testing experience are preserved in classified libraries, permitting rapid reconstitution in either case. The development of fission weapons in 1943-45 and of thermonuclear weapons in 1951-53 from no previous base at all, once the basic inventions had been made, is evidence of the likely speed of reconstitution.

I believe that curatorship, as defined here, should be considered as a serious alternative to stewardship. The decision between them should be based on their respective costs, contributions to the danger of proliferation, and the degree of confidence in our existing forces which each affords.

For additional details on nuclear proliferation click here


Jonathan Katz
Fri Apr 30 13:28:42 CDT 1999