Debating Nuclear Deterrence

Hiroshima, after the bombing (

In his NYRblog post last week, “Is Nuclear Deterrence Obsolete?” Jeremy Bernstein asked whether the principle of deterrence continues to provide a valid ground for maintaining a nuclear arsenal. Following are a series of responses from several of Bernstein’s correspondents who have studied nuclear weapons closely, together with a rejoinder by Bernstein. We invite readers to submit further comments of their own.

Freeman Dyson
Institute for Advanced Study
Princeton, New Jersey

I like your piece but do not agree with the main conclusion. I agree that the nuclear deterrent does sometimes work to deter non-nuclear wars. But the deterrent does not abolish wars altogether. The question is whether the non-nuclear wars that we avoided were better or worse than the risk of nuclear war that we accepted. Opinions can differ about this. I think we are better off going to zero.

Pervez Hoodbhoy
Professor of Physics
Quaid-e-Azam University
Islamabad, Pakistan

Members of Bharatiya Janata Party holding up a dummy of India’s Agni-3 missile as a sign to take up a war with Pakistan,
Ahmadabad, India, May 30, 2002 (Siddharth Darshan Kumar)

You’re right about the uselessness of nukes, although I scarcely need convincing. Around here, as in India, nukes were wildly celebrated after the tests. But suicide attacks occur almost every day against army installations, mosques, public rallies, hospitals, etc. When it comes to terrorism, nukes provide no deterrence, no protection. Last night on television I repeated this dreary truth for the n’th time, drawing outrage and vilification from my opponents for the (n+1)’th time.

As for whether deterrence can work in conventional warfare: absolutely! Since 1971, Pakistan and India have only hovered on the brink of war but not had a real war. By definition, deterrence works because nukes induce terror—that’s the whole point. But what if there’s fear-fatigue? What if political leaders, generals, and even ordinary people start thinking that nuclear war is a mere abstraction? Witnessing events from rather close quarters, I can assert that terror disappears after countries loiter long enough on the brink without blowing each other up. So, by induction, if four South Asian nuclear crises (1987, 1990, 1999, 2002) ended with a whimper (rather than a bang) then the fifth one will as well. Right? That’s what really scares me.

Peter Zimmerman
Professor of Science and Security, Emeritus
King’s College, London

First, I believe that the Russians were deterred. I have some fairly strong evidence from my discussions with Velikhov, Sagdeev, and my counterparts on the Soviet delegation to the START-1 negotiations in 1984-1986 when I was first at the US Arms Control and Disarmament Agency. During the long lunches and dinners I attended two-on-two with Soviets, they clearly communicated that they felt that their government’s behavior, particularly in the darkest days when Stalin and Brezhnev were in charge, was altered by their knowledge that a conventional attack which really engaged the US’s interests had to be limited in scope or be carried out by proxy, else both countries would start to climb the nuclear escalation ladder. Neither side really thought that it could control escalation, so the best policy was to cut off the bottom rung by limiting violence to well below the WWII level. I consider this deterrence worked similarly in continental Europe as well as Korea. Vietnam was such a different conflict that I’m not quite sure what to say.

A further bit of evidence that the Soviet Union felt deterred is to be found in the unclassified writings of Soviet strategic and military thinkers over the years, and finally in Soviet embrace of the ABM Treaty once they came to understand stability. I do think nuclear deterrence kept the Soviets sober about Berlin on two crucial occasions—the Blockade and Airlift, and the building of the Wall—and possibly kept us from intervening during the 1954 riots in East Berlin.

(Gil Hanly)

The Iranians’ hysteria, expressed to me by a senior Iranian nuclear negotiator and very many Iranian journalists, about the Israeli nuclear stockpile convinces me that the Israeli deterrent has kept the Iranians and the Arabs from mounting a serious and broad-based military challenge to Israel’s existence (not just to its frontiers). We will see how Iran acts if it does acquire nuclear arms and how other states in the region react to Iran.

I once chatted seriously with Lord George Robertson, former British Defense Secretary and NATO Secretary General, on the subject of the UK nuclear deterrent. He told me that he grew up as essentially a nuclear abolitionist, but when handed the responsibilities of Secretary of State for Defense, he began to look at the problem very differently and concluded that British freedom of action was essential. Air Marshal Tim Garden, the Liberal-Democrats’ defense spokesman in the Lords, who flew V-Bombers early in his career, said much the same thing.

As the nuclear weapon approaches its sixty-fifth birthday, the question of the need for Hiroshima and Nagasaki will surely be addressed again. Revisionists will likely point out that Japan was defeated and would surely have surrendered if only certain conditions were met, among them retention of the Imperial system, not merely the emperor as constitutional monarch. What matters, of course, is not what an after-the-fact evaluation might demonstrate was true; far more important is what Washington could reasonably infer about the Japanese situation from its own intelligence sources. The Final Months of the War with Japan, Douglas MacEachin’s study of signals intelligence gathered in the spring and summer of 1945, shows that the Japanese forces on Kyushu grew from May to July to three times the number estimated by American officers planning the invasion.


The conclusion drawn was that Japan’s capability to resist remained formidable. Whether that was objectively true or not, it is what Washington knew when the decision to bomb was made. The attacks on those two cities demonstrated the frightfulness of nuclear weapons in ways that no tests in Nevada, Kazakhstan, or Eniwietok never could. And that “vaccination” remains the foundation of nuclear strategy.

Carey Sublette

I haven’t had any opportunities for kaffeeklatsches with former Soviet insiders, but have studied the issue of deterrence for my own edification for a long time. (How long? Well, I first learned of the imminence of nuclear annihilation during the Cuban Missile Crisis—an event that left me with vivid memories, an event that burst into public just 3 weeks after my fifth birthday. I had read Glasstone’s The Effects of Nuclear Weapons cover-to-cover by the time I was 11.)

My view is that deterrence is that once a substantial (and not very large) number of nuclear weapons are in the hands of an opponent, deterrence from behaviors that seem to genuinely risk their use is a natural and unavoidable occurrence of a sane leadership structure. It is simply a manifestation of innate risk-aversion, the caution that ensues when facing the prospect of calamity. The mutual restraint against direct confrontation that both sides exhibited in the Cold War supports this, I argue.

A British trident submarine, HMS Vanguard, approaching the harbor, Port Canaveral, Florida, 1994 (John Bouvia)

One complication is that during the 1950s, military thinking on both sides tended to lag behind the realities of the weapons themselves due to inexperience and novelty. Both sides integrated tactical nuclear weapons into regular military operational planning, with the expectation that there use would be inevitable in war. U.S. “massive retaliation” policy lagged behind Soviet capabilties in the late 50s after it had become obsolete.

The Cuban Missile Crisis puts this thesis to an interesting test. One could say that deterrence failed, since the fear of U.S. nuclear weapons did not prevent Khruschev from taking a gamble that might have precipitated a nuclear war. But it appears that Khruschev did not realize this would be the case. In his view, U.S. forward deployment of stategic weapons and Soviet deployment in Cuba were mirrors of each other (“fair play” you might say), and he did not appreciate how deeply ingrained the Monroe Doctrine is in U.S. thinking. Even so—the U.S. was NOT planning on a nuclear response, it was planning a conventional attack with the belief that a conventional assault on Cuba would not trigger a Soviet nuclear response, not realizing (until decades later) that Soviet tactical nuclear weapons had been introduced with field commanders having the authority to use them in the event of combat.

The way both sides backed away from the confrontation as each side’s “red lines” became apparent to the other shows that in the end deterrence remained potent. The real danger was not failure of deterrence so much as blundering in the dark—which remains the real danger of nuclear weapons today. Mistakes can be cataclysmic.

Another crucial case study is the Yom Kippur War. The Syrian battle plan is known today, and its objective was simply to recapture the territory lost 6 years earlier, stopping on the Jordan River. Achieving this objective would have been a crushing blow to Israel, but restoring the status quo of 4 June 1967 in no way represented any sort of existential threat to the state of Israel. Nevertheless, for the 20 day duration of fighting, the Yom Kippur War was a state of total war—a full commitment of resources—for the three states engaged, the only time a nuclear armed state has found itself in such an intense conflict since World War II.

Side note: the period during which the Syrians halted at the edge of the Heights, with no Israeli ground forces to block them was (I find, upon refreshing my memory) no more than 20 hours. The speculation that has been offered in various quarters over the years that this halt was deliberate and due to fear of an Israeli nuclear response is clearly wrong. The Syrian leadership had already set that concern aside in planning the attack—the attacking forces were under orders to advance to the river, they simply failed to do so when they had the opportunity.


This illustrates the impotence of nuclear deterrence in preventing conventional conflict—the Yom Kippur War was the most intense episode of combat in the world since the Battle of Kursk in 1943. The more recent example of the Kargil War, a limited conflict between two nuclear armed states, reinforces this point.

My idea of “Simple Deterrence” is essentially the same as the one Richard Garwin advocates in his 1977 paper on a “Second Nuclear Regime.” It is based on the simple policy that the only role of nuclear weapons is to deter the use of nuclear weapons. There is no secondary role at all. This principle then drives all other aspects of nuclear weapons policy: force size and structure, declared policy, bilateral and multilateral agreements, etc. The likely result of adopting this policy would be a nuclear arsenal quite similar to those envisioned by advocates of “minimal deterrence”, but this is merely part of the means of implementation rather than an objective. Fully realizing a universal simple deterrence policy will likely require offering multinational guarantees to all states against external non-nuclear existential threats.

Commander Robert Green
Royal Navy (Retired)

In 1969, as a 25-year-old Buccaneer nuclear strike jet bombardier-navigator on the British aircraft-carrier HMS Eagle, I accepted with a buzz of excitement and pride being trained to drop a ten kiloton WE177 free-fall tactical nuclear bomb on a military airfield on the outskirts of Leningrad. Thirty years later, visiting St. Petersburg for a conference about European security on the eve of the 21st century, I realized with a shock that the airport where I landed had been my target. That night on Russian television, I apologized to my startled and appreciative hosts. I had been prepared to carry out a mission that would have caused appalling numbers of indiscriminate casualties and poisonous effects from radioactive fallout for decades, while virtually destroying Russia’s beautiful ancient capital as collateral damage. I then told them I had learned that nuclear weapons would not save me, or them.

I’d like to inject some realism into the debate stimulated by Jeremy Bernstein’s blog “Is Nuclear Deterrence Obsolete?” First of all, nuclear weapons have no military use. Their uniquely indiscriminate, long-term health effects, including genetic damage, on top of almost unimaginable explosive violence, make them the most unacceptable terror devices yet invented—far worse than chemical or even biological weapons.

The Enola Gay and its crew, who dropped the “Little Boy” atomic bomb on Hiroshima

As I explain in my new book Security Without Nuclear Deterrence
it was the 1991 Gulf War that persuaded me to break out of the understanding of nuclear weapons I had been inculcated with in the military. From my experience in naval intelligence in the 1982 Falklands/Los Malvinas conflict—an earlier regional resource war where nuclear weapons were a factor—I feared that chemical-headed Iraqi Scud missile attacks might have provoked Israel to respond with a nuclear strike on Baghdad. This time, a de facto nuclear state was directly attacked by a non-nuclear state. Israelis, cowering in gas masks in basements as nearly 40 conventionally armed Scud attacks rained down, learned that their so-called ‘deterrent’ had failed. Meanwhile, the Irish Republican Army just missed wiping out the entire British War Cabinet with a mortar-bomb attack from a van in Whitehall. Where was Polaris? And what if they had tried using a crude nuclear device? This was a foretaste of the primary threat we now face from terrorism.

Nuclear deterrence is treated like a game between naïve, posturing political leaders—who, unlike myself, will never be assigned the task of actually “pressing the button.” They play down the apocalyptic effects of modern nuclear weapons, assuring their citizens that nuclear deterrence means they would never be used. Yet for nuclear deterrence to be credible, these same leaders must be prepared to order their use, in what would be state-sponsored nuclear terrorism amounting to irrational, posthumous revenge. That’s why all but a handful of states have rejected nuclear deterrence as undermining their true security.

A recent study published in Scientific American analysed what could happen if just 100 nuclear devices—half of one percent of the current US and Russian arsenals—were detonated over cities in South Asia in a regional nuclear war between India and Pakistan. In addition to millions of casualties, smoke from fires could encircle the globe and cause enough loss of sunlight and cold to trigger widespread famine threatening billions. So much for climate change.

This is why I am watching with growing frustration the game-playing at the UN conference on the Nuclear Non-Proliferation Treaty this month in New York. To judge from the discussions so far the prospects for continuing disarmament and slowing proliferation don’t look promising. No delegations from Israel, India or Pakistan are present: Israel is still allowed to get away with the irresponsible fiction that it doesn’t have nukes, while India and Pakistan reject the Non-Proliferation Treaty as nuclear apartheid between the nuclear haves and have-nots. Meanwhile, the US and its subservient allies gang up on an understandably paranoid and dangerous Iran, which hasn’t yet got any nukes.

When will British leaders stop abusing the Royal Navy, requiring it to do their dirty nuclear work in exchange for uncritical obeisance to the US? My one hope is that the black hole in the UK defense budget could yet force the new Conservative-Liberal Democrat coalition government to make virtue out of necessity, and abandon the Trident program—thus positioning Britain for new world leadership as the first recognized nuclear weapon state to break out from enslavement to the huge hoax of nuclear deterrence

Jeremy Bernstein

When I write about something like this I am keenly aware of Bohr’s admonition not to speak more clearly than one thinks. I am also aware that unlike some of you I am an outsider with no access to classified information—a perspective that I think has advantages and disadvantages. I also cannot say as Carey did that I was reading Glasstone at age 11. I was at that age exploring the deeper meaning of Uncle Wiggly. To the matter at hand.

Slim Pickens as Major T.J. Kong, in Stanley Kubrick’s film Dr. Strangelove

It is evident to me that deterrence is a binary relationship between a set of deterers and a set of deterees. Therefore having a “secret weapon” as part of deterrence is absurd. This is one of the paradoxes that Kubrick explores in Doctor Strangelove. The nuclear missiles on Cuba which were secret played no role in deterrence. One only wonders what would have happened if we had known. As I recall, the Indians and Pakistanis were about ready to go nuclear when Colin Powell explained to them what a nuclear war would mean. One of the dangers is that countries that are fooling with nuclear weapons now do not understand this. Powell told them what would happen if an atomic bomb was dropped on Karachi or Mumbai. In the Israeli situation one also has to take into account what having these weapons means to the psychology of the Israelis, who can show some self-restraint knowing that they have the bomb.

The British situation seems especially instructive. The notion of having an “independent deterrent” may sound sensible until one looks into what in the British case this actually is: one submarine wandering around submerged. When I read about this I kept thinking of the Beatles song, “We all live in a yellow submarine.” As part of their agreement to form a new coalition government, David Cameron and Nick Clegg have committed to re-examine this situation. Their stated concern seems to be to try to run the Trident program more economically. Their real concern should be whether to do it at all.

Subscribe or give a gift, The New York Review of Books

Give the gift they’ll open all year.

Save 55% off the regular rate and over 75% off the cover price and receive a free 2024 calendar!