Did the Soviets really build a "doomsday machine"?

Did the Soviets really build a "doomsday machine"?

Did the Soviets really build a "doomsday machine"?

Scrutinizing culture.
Aug. 31 2007 3:52 PM

The Return of the Doomsday Machine?

Please don't count on me to save the world again.

Nuclear bomb test
Nuclear bomb test

"The nuclear doomsday machine." It's a Cold War term that has long seemed obsolete.

And even back then, the "doomsday machine" was regarded as a scary conjectural fiction. Not impossible to create—the physics and mechanics of it were first spelled out by U.S. nuclear scientist Leo Szilard—but never actually created, having a real existence only in such apocalyptic nightmares as Stanley Kubrick's Dr. Strangelove.


In Strangelove, the doomsday machine was a Soviet system that automatically detonated some 50 cobalt-jacketed hydrogen bombs pre-positioned around the planet if the doomsday system's sensors detected a nuclear attack on Russian soil. Thus, even an accidental or (as in Strangelove) an unauthorized U.S. nuclear bomb could set off the doomsday machine bombs, releasing enough deadly cobalt fallout to make the Earth uninhabitable for the human species for 93 years. No human hand could stop the fully automated apocalypse.

An extreme fantasy, yes. But according to a new book called Doomsday Menand several papers on the subject by U.S. analysts, it may not have been merely a fantasy. According to these accounts, the Soviets built and activated a variation of a doomsday machine in the mid-'80s. And there is no evidence Putin's Russia has deactivated the system.

Instead, something was reactivated in Russia last week. I'm referring to the ominous announcement—given insufficient attention by most U.S. media (the Economist made it the opening of a lead editorial on Putin's Russia)—by Vladimir Putin that Russia has resumed regular "strategic flights" of nuclear bombers. (They may or may not be carrying nuclear bombs, but you can practically hear Putin's smirking tone as he says, "Our [nuclear bomber] pilots have been grounded for too long. They are happy to start a new life.")

These twin developments raise a troubling question: What are the United States' and Russia's current nuclear policies with regard to how and when they will respond to a perceived nuclear attack? In most accounts, once the president or Russian premier receives radar warning of an attack, they have less than 15 minutes to decide whether the warning is valid. The pressure is on to "use it or lose it"—launch our missiles before they can be destroyed in their silos. Pressure that makes the wrong decision more likely. Pressure that makes accidental nuclear war a real possibility.


Once you start to poke into this matter, you discover a disturbing level of uncertainty, which leads me to believe we should be demanding that the United States and Russia define and defend their nuclear postures. Bush and Putin should be compelled to tell us just what "failsafe" provisions are installed on their respective nuclear bombers, missiles, and submarines—what the current provisions against warning malfunctions are and what kinds of controls there are over the ability of lone madman nuclear bombers to bring on the unhappy end of history.

As for the former Soviet Union, the possible existence of a version of a doomsday machine is both relevant and disturbing.

In the Strangelove film, the Soviet ambassador tells the president and generals in the U.S. war room that the device was designed to deter a surprise attack, the kind of attack that might otherwise prevent retaliation by "decapitating" the Soviet command structure. The automated system would insure massive world-destroying retaliation even if the entire Soviet leadership were wiped out—or had second thoughts. As a result, some referred to it as the "dead hand" doomsday device.

It is Dr. Strangelove himself, the madman U.S. nuclear strategist played by Peter Sellers, who detects the flaw in this plan. After being apprised of the system's existence by the Soviet ambassador, and the likelihood of its being triggered by a U.S. bomber on an unauthorized mission to nuke its Soviet target, Dr. Strangelove exclaims:

Yes, but the ... whole point of the doomsday machine ... is lost ... if you keep it a secret! Why didn't you tell the world, eh?


In other words, a doomsday machine kept secret is no good for deterrence, only for retaliation by extinction.

Did the Soviets actually design a variation on a doomsday device and not tell us about it? And could an accidental or terrorist nuclear attack on Putin's Russia (by Chechens, for instance) trigger an antiquated automated dead-hand system and launch missiles capable of killing tens, maybe hundreds, of millions at unknown targets that might include the United States?

Up until Aug. 10 of this year, I would have thought these questions were best consigned to the realm of apocalyptic film fantasy. But on that day I came upon a startling essay in the London Times Literary Supplement. It was a review (titled "Deadly Devices") of a book recently published in the United Kingdom: Doomsday Men: The Real Dr. Strangelove and the Dream of the Superweapon by nuclear-age historian P.D. Smith of University College London. (It will be out in the United States in December.)

The TLS reviewer, Christopher Coker (who is on the faculty of the London School of Economics), asserted that the book demonstrates that "only after the Berlin Wall had been breached and ... the Cold War began to thaw did military analysts realize the Russians had actually built a version of the [doomsday] device. The details of this top-secret Soviet system were first revealed in 1993 by Bruce G. Blair, a former American ICBM launch control officer, now one of the country's foremost experts on Russian arms. Fearing that a sneak attack by American submarine-launched missiles might take Moscow out in 13 minutes, the Soviet leadership had authorized the construction of an automated communication network, reinforced to withstand a nuclear strike. At its heart was a computer system similar to the one in Dr. Strangelove. Its code name was Perimetr. It went fully operational in January 1985. It is still in place."


Wait a minute. Still in place?! How is this possible?

In the endnotes of Smith's book (which turns out to be an illuminating portrait of the Doomsday weapon concept and its cultural implications), I found a reference to a further description of the Perimetr system in a 2003 Washington Post op-ed by Bruce G. Blair, the former Minuteman ICBM launch control officer who first revealed the existence of the program. (When he wrote the op-ed, he was a Brookings fellow; he is now head of the World Security Institute in Washington, a liberal think tank.)

The op-ed offers a far more detailed and chilling picture of Perimetr than the brief mention devoted to it in the book and review:

Die-hard [U.S.] nuclear war planners actually have their eyes on targets in Russia and China, including missile silos and leadership bunkers. For these planners, the Cold War never ended. Their top two candidates [i.e., targets] in Russia are located inside the Yamantau and Kosvinsky mountains in the central and southern Urals.

Both were huge construction projects begun in the late 1970s, when U.S. nuclear firepower took special aim at the Communist Party's leadership complex. Fearing a decapitating strike, the Soviets sent tens of thousands of workers to these remote sites, where U.S. spy satellites spotted them still toiling away in the late 1990s.


Blair sources his information on these command bunkers to "diagrams and notes given to me in the late 1990s by SAC [Strategic Air Command] senior officers," men in charge of targeting our missile and bomber forces.

From them, he paints a Strangelovian picture:

The Yamantau command center is inside a rock quartz mountain, about 3,000 feet straight down from the summit. It is a wartime relocation facility for the top Russian political leadership. It is more a shelter than a command post, because the facility's communications links are relatively fragile. As it turned out, the quartz interferes with radio signals broadcast from inside the mountain.

A quartz nuclear-war mountain! Something phantasmal about it, like a satanic big rock candy mountain. But the quartz mountain melts in comparison with the Perimetr dead-hand system at Kosvinsky.


"Kosvinsky," Blair tells us, "is regarded by U.S. targeteers as the crown jewel of the Russian wartime nuclear command system, because it can communicate through the granite mountain to far-flung Russian strategic forces using very-low-frequency (VLF) radio signals that can burn through a nuclear war environment. The facility is the critical link to Russia's 'dead hand' communications network, designed to ensure semi-automatic retaliation to a decapitating strike."

Of course, there's a world of difference between a "semi-automatic" doomsday device and the totally automatic—beyond human control—doomsday device in Strangelove, something that Blair is careful to note. The Soviet facility does require a human hand for the final fatal push of the button. But Blair believes that the human brain behind that hand has not been programmed to suddenly turn peacenik. And the details of the device are far from reassuring.

"This doomsday apparatus, which became operational in 1984, during the height of the Reagan-era nuclear tensions, is an amazing feat of creative engineering." According to Blair, if Perimetr senses a nuclear explosion in Russian territory and then receives no communication from Moscow, it will assume the incapacity of human leadership in Moscow or elsewhere, and will then grant a single human being deep within the Kosvinsky mountains the authority and capability to launch the entire Soviet nuclear arsenal.

"Kosvinsky came online recently," Blair wrote in 2003, "which could be one explanation for U.S. interest in a new nuclear bunker buster."

Blair also suggested that the Bush administration's recurrent interest in funding the development of nuclear "bunker buster" bombs was at least in some respects designed to give them the capacity to destroy the dead-hand device buried deep in a Kosvinsky bunker, an argument that, if true, would suggest the dead-hand doomsday device was still thought to be operational. And perhaps you've heard something about its deactivation, but I haven't found any evidence of it.

Blair, who has written previously on the extremely rickety structure of presidential nuclear decision-making, believes that the current U.S. contingency plan is itself a "doomsday strategy":

President Bush's nuclear guidance doubtless instructs the Pentagon to plan the destruction of Yamantau and Kosvinsky, along with 2,000 other targets in Russia and hundreds more in China. But such targeting requires very high-yield weapons, typically 10 to 100 times more destructive than the bombs dropped on Japan in 1945. We are talking about a doomsday plan in which Yamantau and Kosvinsky are struck as part of an all-out nuclear exchange that would kill hundreds of millions of people.

There's some ambiguity in Blair's use of "doubtless": Does it imply that Bush's "nuclear guidance" includes only one all-out, 2,000-target response, or "merely" the capability of it? But shouldn't we know at least that in a genuinely "doubtless" way?

Blair's primary recent concern is not the prospect of a deliberate, ideological, Cold War-type nuclear war, but accidental war caused by the continued deadly presence of all-too-easily triggered Cold War arsenals. In four fascinating papers on the subject (all available online, and well worth reading), Blair describes the "launch on warning" bias built into our nuclear command structure, and foresees the possibility of a doomsday that results from our attempt to pre-empt their doomsday plan, all of which might be touched off by accident, mistake, or malfunction on either side.

Blair is not a wild-eyed Cassandra raising unsupported suspicions. Colleagues in his field regard him as a serious and cautious scholar raising real questions. Stephen M. Meyer, an expert on the Russian military at MIT, told the Times that Blair "requires of himself a much higher standard of evidence than many people in the intelligence community."

Blair's troubling papers, along with his book The Logic of Accidental Nuclear War, serve as a reminder that the illogic, irrationalities, and vulnerability to catastrophic error of our Cold War nuclear war command and control mechanisms were never resolved or fixed, just forgotten when the Cold War ended. His analysis suggests that during the Cold War, we may have escaped an accidental nuclear war by luck rather than policy.

It was Blair who pointed out, in congressional testimony, another continuing problem with nuclear launch posture, this one involving the much-ballyhooed "de-targeting"—a process by which the United States and the former Soviet Union purportedly reduced the risk of accidental nuclear war by insuring that their missiles were—after the fall of the Soviet Union—not still targeted at each other. Blair told Congress that, especially on the Russian side, detargeting was only "cosmetic and symbolic," and easily reversible, implemented in name only.

What drove Blair? I was particularly fascinated by one of Blair's other papers, his more personal "Nuclear Recollections," which might have been called "Memories of a Minuteman Missile Crewman," and describes his period of service in a missile silo at the Malmstrom, Mont., Air Force Base, hundreds of feet beneath the Great Plains.

Especially because I'd been there! Down in one of those silos, under the bleak landscape of the Great Plains (this one in Grand Forks, N.D.), interviewing missile commanders like Blair (for a Harper's story), only a few years after Blair resumed life aboveground and retired.

In the course of talking to Minuteman commanders down in their underground launch capsules, I'd glimpsed what they might be called upon to do. They had the ability to launch from their underground pods up to 50 missiles able to kill 200,000 or 300,000 people each. You do the math.

They certainly had, and it showed beneath their black-humored jokes about coming above ground after a nuclear war and finding "only huge mutant bunny rabbits alive."

They were, thank God, not automatons. As Blair points out, their training system was designed to turn them into automatic button pushers, but the ones I spoke to retained a sharp sense of skeptical individuality. About the gravity of their "mission": killing that many people. And about the sketchy mechanics of it.

One crew member even disclosed to me a flaw in the "command and control" "permissive action" system that was supposed to prevent a madman missile commander from launching his "birds" and starting an apocalyptic nuclear war all by himself. The flaw: the system's susceptibility to the "spoon and string" improvisation.

So much focus has been placed—in film, fiction, and nonfiction—on our supposedly "failsafe" barrier to a lone-madman launch. We'd been told that to launch a missile, two keys must be inserted simultaneously into their slots by two separate launch officers, and that the slots for the keys were located at a sufficient distance from each other that one madman couldn't, say, shoot the other crewman and then use both his arms to twist both the keys simultaneously.

But the missile crewmen I talked to told me they'd figured out a way to defeat that impediment with a spoon and a string. Not that they were planning to do it, but that they knew someone could do it.

You just shoot the other guy and "rig up a thing where you tie a string to one end of a spoon," he told me, "and tie the other end to the guy's key. Then you can sit in your chair and twist your key with one hand while you yank on the spoon with the other hand to twist the other key over."

American ingenuity! Can't beat it for finding a new way to end the world.

I always wondered if I should follow up on what happened after I published this information. (In a piece reprinted in The Secret Parts of Fortune, I assumed the flaw had been fixed somehow, and have long credited myself with saving the world. Kidding!)         

I actually turned down an invitation to lecture about such matters from the Air War College in Alabama * (because of my peacenik inclinations at the time), and assumed that if they read the article, they must have taken action to save the world from a lone madman with a spoon and string, to whom I'd in effect given instructions for an unauthorized missile launch that could destroy the world. (Hmmm, maybe I'd come close to destroying the world, rather than saving it. Sorry about that.)

But it's clear from Bruce Blair's "Nuclear Recollections" that the experience of holding the lives of tens of millions in his hands when he held those keys left a profound mark on him. I know that when the missile crewmen I was interviewing let me hold the keys, even twist them into the (deactivated) locks, that it had a profound effect on me. The keys to Kingdom Come!

And while I may have abandoned my responsibility for too long, I was grateful that Bruce G. Blair was still on the case, raising the right questions. In fact, he's devoted his subsequent life to raising the alarm about our flawed nuclear alarm and launch system, using what an actual missile commander learned about its dysfunctions and biases.

Blair's work continues and I think it's urgent, now that Putin's "nuclear bombers" are flying again, that Congress re-examine the whole issue and take seriously Blair's warnings about the variations of doomsday we still face.

Pay attention to Blair. You can't count on me to save the world again.

Correction, Sept. 5, 2007: This piece originally referred to the Army War College in Alabama. In fact, it's the Air War College that is sited in Alabama. (Return to the corrected sentence.)