Jeffrey LewisLife After the RRW & Virtual Swords

Shortly before I left for the holidays, Congressional Appropriators provided “no funds for the Reliable Replacement Warhead (RRW)” pending “a new strategic nuclear deterrent mission assessment for the 21st century.” The bottom line — no new warheads without a new posture — appears to command bipartisan support among the appropriators.

The RRW is, for all intents and purposes, dead.

Administrator Tom D’Agostino and the rest of NNSA have to be asking themselves: Now what?

Our friend John Fleck points to one answer in the Albuquerque Journal, noting similarities between a 1990 paper and D’Agostino’s remarks on 18 December:

In January 1990, as the Soviet Union collapsed, a pair of defense industry consultants wrote a paper outlining a new approach to meeting military needs in the post-Cold War world.

Rather than a pipeline constantly churning out new weapons, Ted Gold and Rich Wagner wrote, the United States should develop the industrial research and manufacturing capability to build weapons if needed.

We do not need a huge arsenal, they argued. Instead, we could deter future enemies merely by showing that we have the capability to build new weapons when we need them. The essay was titled “Long Shadows and Virtual Swords.”

Fast-forward to Dec. 18.

In a Washington, D.C., news conference, the man in charge of the U.S. nuclear weapons design and manufacturing process seemed to be echoing Gold and Wagner.

“Because our nuclear weapons stockpile is decreasing, the United States’ future deterrent cannot be based on the old Cold War model of the number of weapons,” said Thomas D’Agostino, head of the National Nuclear Security Administration. “Rather, it must be based on the capability to respond to any national security situation, and make weapons only if necessary.”

I created a text version of the Gold and Wagner paper because I can’t find it anyway on-line. (It probably has more than a couple of typos from the OCR recognition software — feel free to e-mail corrections.)

I really think this is the only argument that NNSA has going for nuclear weapons programs, including whatever stockpile work will come after RRW. I never got around to flagging the idea, even after Joe Martz made a pretty decent case to the San Francisco Chronicle‘s Jim Sterngold. Martz, speaking to Fleck, aptly argued “My work becomes the deterrent, not so much the products of my work.”

I may just be a sucker for the “virtual swords” thing, having got my start in Washington working for Mike “Virtual Nuclear Arsenals” Mazarr. But it seems to me that, at some point, we need a bipartisan consensus on what the labs are supposed to do in post-arms race world. And that requires a vision of what it is that nuclear weapons do in that world.

Now, don’t get me wrong — a “virtual swords” concept should not be an excuse to fund an infrastructure better sized to a nuclear weapons stockpile of 10,000 than 1,000 (see the Modern Pit Facility). And my politics are not those of Gold and Wagner. But I can see how prudent investments in our defense industrial base, most importantly the people, can provide a hedge that enables deep reductions in our bloated nuclear stockpile that could safely number in the hundreds, rather than thousands, of weapons.

I would argue that NNSA officials failed to secure Congressional support for a variety of multi-billion dollar initiatives — including funding for the Modern Pit Facility, Robust Nuclear Earth Penetrator and Reliable Replacement Warhead — precisely because these programs were conceived, articulated and implemented as part of a stockpile that looks liked a smaller version of the Cold War stockpile, instead of a stockpile based on the reality that much of the deterrent benefit from our nuclear stockpile is existential in nature.

It seems to me that fact — that the deterrent benefit accrues through the weapons existence and is robust across disparities in the technical details — forms to core of my answer to Cheryl Rofer’s excellent challenge to bloggers to articulate a new nuclear posture.

Update: Joe Martz has some interesting comments at LANL: The Rest of the Story.


  1. Haninah (History)

    When the RRW debates were still in full pitched mode a year or two ago, NNSA was making the argument (and a somewhat plausible one, in my opinion) that the only way to train new competent nuclear warhead manufacturing personnel – not just physicists, but engineers and technicians – was by manufacturing nuclear weapons. They argued that these types of skills could not be developed without real experience. What I want to know is, what has changed, either in reality or in people’s reasoning, to take that argument off the burner?

  2. Ed E. Torr (History)

    The second link in your Update: is broken.

  3. Chuck (History)

    Has the text of the Gold and Wagner paper been moved or removed? I get the infamous 404 error that site is not found.

  4. Jeffrey Lewis (History)

    Whoops, sorry.

    I fixed the link, but here it is:

  5. Andy (History)


    That same argument was successfully used to keep nuclear submarine programs “afloat.” Closing down production on submarines would mean a loss of talent and knowledge that might be difficult or impossible to replace at a future date.

    I don’t know if such claims in either case are valid, but if they do hold merit, it seems that a creative solution to mothball knowledge is possible were resources provided.

  6. John Fleck (History)

    Haninah – I can’t speak for NNSA people directly, but my assessment based on the discussion over the last few weeks is that they continue to believe that RRW is important, for the reasons you describe. But they also recognize the reality – that they need to come up with a “plan B” that is robust to the uncertainty about whether RRW goes forward, and (more importantly) that a large new nuclear manufacturing center is not in the cards.

    In other words, this plan is a realistic approach to the fact that they are not likely to get what they really want.

  7. Jeffrey Lewis (History)


    One thing that “changed” was that the design competition achieved most of those goals.

    (Also, we continue to manufacture pits, albeit at a very slow rate, so that argument isn’t germane to RRW.)

  8. Mark Gubrud

    Back around 1990 I was way ahead of Gold & Wagner, predicting virtual arsenals and arguing the need for, yes, virtual arms control. Because sooner or later we have to go beyond the idea that virtual weapons are stabilizing in comparison with actual weapons, to consider whether they might not be destabilizing in comparison with nothing at all.

    A relentless virtual arms race leads, by the same logic as a real arms race, to the predeployment of plans, plants, robots and whatnot to quickly startup rapid mass production of real weapons, and once again whoever puts the meanest machine in place and activates it first gains a preemptive advantage.

    Well, since nature abhors a vacuum, nothing at all can’t be stable, either. But the answer can’t just be a virtual arms race for virtual arsenals. To move beyond the inherently unstable peace of deterrence we need to build an active structure of peace assurance, which must include verified arms control to ensure the absence of breakout preparations.

  9. Geoffrey Forden (History)

    It was my understanding that one design in the “design competition” was so simple and so robust that there would never be an issue about it being made, no matter how the worker safety or environmental laws changed. Since they rejected that (in favor of a design they think will need to be redone every twenty years or so), they cannot be serious about any of the issues that “require” an old hand being present. Nor, in my opinion, do they really care about our nuclear deterrent or our national security. Otherwise, they wouldn’t be playing games with it in order keep weapons designers in jobs.

  10. Mastermind (History)

    Nuclear bombs were successfully built back in 1945 and then improved during the 50’s and 60’s using the technology of those periods. Given this, how hard would it be, really, to rebuilt this capability using 21st century technology if it is needed. Not hard, by my estimation.

    US mid-70’s nukes were highly optimized for size/weight so they could be MIRV’ed onto a single missile. That need is no longer relevant. Therefore, I suspect that a reasonably good US defense contractor could probably rebuilt a credible nuke arsenal if necessary at some future date. Don’t let the nuke designers song and dance that “it’s an art that could be lost” fool you.

  11. Kelsey D. Atherton (History)

    […]”…that the deterrent benefit accrues through the weapons existence and is robust across disparities in the technical details”[…]

    quoted @

  12. Nuk-U-Lar (History)

    Dear Mastermind,

    Could you please qualify the following opinion.

    “Not hard, by my estimation.”

    On what do you base your estimation? How many years experience do you have in either the DOE or DoD complex? Have you ever participated in getting a conventional weapons system type classified? Have you been involved in quality control with a domestic defense contractor? How about doing war reserve (WR) work for DOE?

    It all sounds easy until YOU have to reconstitute the ability to do plutonium work in a foundry. Try getting that EIS worked out in less than a decade.

    I thought the idea of virtual swords was that you had a ready ( almost immediate ) capability to respond. I assume that means a short period of time spent using facilities that are already in place.

    I agree that any one of the top weapons contractors could manufacture useful nuclear weapons. But if they had to start from scratch today, they would need quite a long time before they could begin to do the final system integration and product delivery.

    Its not so much an art as a statement of where we are after a sixty year history of development. You lose too much ‘tribal knowledge’ and you have to spend too much time reinventing it before you are overtaken-by-events.

    I find it amazing that Congress is willing to spend so much money to safeguard foreign nuke material and pay to keep former enemy designers busy doing something so they won’t go work for a rogue nation or organization and our DOE/NNSA is trying to economize on our complex. Lots of staff are wondering how they suddenly became expendable after devoting much of their career to a dead-end topic such as nuclear weapons.

  13. Mark Gubrud

    If we lost some of the “tribal knowledge” and if as a result a reconstituted nuclear weapons design and production complex would take some time to gear up and would initially not produce as refined a product as the old masters were making “after a sixty year history of development”, well, would that be such a bad thing?

    Might the time lags, costs and technical difficulties of restarting production not be an important stabilizing factor in any program of phased reductions leading to the final abolition of nuclear weapons?

    Or, if nuclear abolition seems an extreme idea, is it less extreme to hold it necessary or even desirable to be able to pick up right where the Cold War left off, without a hitch, at any time? Is it reasonable to fear being “overtaken by events” while we still have thousands (or even dozens) of perfectly functional nuclear weapons, or even in some hopeful future scenario where we have effected a verified and assured global abolition?

    What events, and how would they overtake us?

    After all, abolition cannot ever mean that we are more than a short step from reconstitution of what ought to be a sufficient deterrent. Remind me, how many nuclear weapons does it take to ruin your whole day?