'People, ideas, machines' II: catastrophic thinking on nuclear weapons

Western theories re Soviet 'rationality' and nuclear strategy/ deterrence were revealed post-91 to be overconfident & wrong... Few realise... The rot of the UK nuclear enterprise...

‘Would it not be wondrous for this whole nation to be destroyed like a beautiful flower?’ General Anami trying to persuade the Japanese leadership not to surrender after Hiroshima and Nagasaki.

‘[Payne’s book is an] extended reminder of just how fortunate we probably were to survive the Cold War. He rubs our noses in the unarguable fact that nuclear deterrence is all theory, and much of it is not very convincing theory at that… When we consider the Soviet/Russian and U.S. strategic nuclear postures as complex somewhat interdependent systems, especially in their full political, military, human, social, and technological contexts, it is hard to resist the judgment that we are fortunate to be here today. Complex systems have ‘normal accident’ rates.’ Colin Gray, leading scholar of nuclear strategy.

‘There was no defense against our own preconceptions.’ Kissinger, re US blunder over Egypt in 1973.

‘The combination of physics and politics could render the surface of the earth uninhabitable… [Technological progress] gives the appearance of approaching some essential singularity in the history of the race beyond which human affairs, as we have known them, cannot continue.’ John von Neumann, one of the 20th Century’s most important mathematicians.

‘Politics is always like visiting a country one does not know with people whom one does not know and whose reactions one cannot predict. When one person puts a hand in his pocket, the other person is already drawing his gun, and when he pulls the trigger the first one fires and it is too late then to ask whether the requirements of common law with regard to self-defence apply, and since common law is not effective in politics people are very, very quick to adopt an aggressive defence.’ Bismarck, 1879.

‘People, ideas, machines — in that order!’ What Colonel Boyd shouted in the Pentagon for years…

Los Alamos: Von Neumann, Feynman, Ulam; below, the first picture of a nuclear explosion

Introduction

This blog is about ideas on nuclear weapons. With NATO helping Ukraine kill Russians and ideas like ‘no fly zones’ bandied about, the most important thing for those with influence on events is to understand crucial issues about deterrence, escalation, and WMD including all-out nuclear war. Regardless of what happens in Ukraine, these issues will be critical over Taiwan too. Pundit warriors and politicians are already banging the war drums for the next one.

In the Cold War America based its nuclear strategy on an intellectual framework that was false.

It defined standards of ‘rationality’ then concluded the Soviets would not use nuclear weapons in many scenarios. There was a governing tautology: rational leaders would be deterred otherwise they would be irrational. Given this tautology, more vulnerability improves ‘stability’ (e.g submarine launched weapons), while better defence is ‘destabilising’ (e.g missile defence).

The Cold War was won. The West concluded ‘we were right’. Many in the world of policy concluded: there is a reliable theory of nuclear strategy that allows us to send carefully calibrated signals, like ‘escalate to de-escalate’. You can see this false confidence in many politicians, journalists and academics over the past month. E.g Professor Elliot Cohen’s calls for America to attack Russian forces because he’s confident Putin is bluffing.

After the 1991 collapse some scholars went to talk to those actually in charge in Russia. They read documents. They discovered that we’d been wrong in crucial ways all along.

Actually the Soviets planned early and heavy use of nuclear weapons in many scenarios including outbreak of conventional war in Europe.

The theoretical basis of some of the west’s analysis, such as game theory from the likes of the economist Schelling, had been disastrously misleading. More important (I think) was the development of a theory that encouraged leaders/strategists to ignore an eternal lesson of history: one story after another of people risking death in ways opponents or observers thought ‘irrational’, ‘crazy’. (Remember the Jews’ response to Pilate’s severe and credible threats to kill them — they lay down and turned their necks to him and said do it! And it worked!) The actual motives and thinking of specific leaders was discouraged in favour of calculating balances of weapons. Bureaucracies focused on what can be counted and calculated rather than imponderable questions with no clear and comforting answers. The former was comforting. The latter seemed paralysing and/or nightmarish. Bureaucracies naturally gravitate toward the former unless very strong counter-pressures apply.

Not only were we wrong about that, but, because of how incentives work in the policy world, after 1991 nuclear weapons retreated to being a very niche subject. Very few senior politicians and officials now have studied these issues. Many of them, and influential journalists, believe we have reliable theories of ‘nuclear strategy’ and ‘nuclear escalation’.

After 1991 we also discovered that at various points we had come much closer to the brink than had been realised at the time because of accident and confusion, the famous ‘fog of war’ applied to nuclear.

For example, after the Cuban missile crisis JFK and RFK span many myths to journalists about JFK going ‘eyeball to eyeball’ and prevailing. These myths were powerful. You see them echoed in many screams over the past few weeks for a no-fly zone over Ukraine. The myths were false. (A former Russian foreign minister repeated some of these myths days ago, clearly not having read about what actually happened in Cuba.)

We now know that below the sea a Soviet commander ordered nuclear weapons to be fired. Because of a fluke (to do with rank), he needed the agreement of Vasily Arkhipov. Arkhipov did not consent. The nukes were not fired. The crisis was, barely, resolved peacefully. People read the news and the Kennedy spin. The world moved on. We did not learn about this extraordinary moment under the sea for years.

This is a recent book if interested in details:

In 1983, the Soviet early warning system thought it had detected a US nuclear first strike. Fortunately the commander on duty, Petrov, had a hunch it was a false alarm and kept quiet instead of escalating. (He was punished for not filling in the forms properly — cf. my blogs on incentives in bureaucracies!) There were other similar incidents where early warning systems went wrong and nearly caused disaster. (After leaving No10 I was asked what I’d like people to study and chose this subject.)

So:

1. We operated with bad theories (e.g Schelling) and bad specific information about Soviet plans (e.g they won’t use nukes in scenario X).

2. Many wrongly think these theories were good and provide a rational, effective approach to managing deterrence and escalation.

3. Systems for early warning nearly failed disastrously many times, are inherently prone to error, and physical security of nuclear bases is repeatedly exposed by Red Teams to be compromised (sometimes laughably, such that planes are parked unguarded overnight because nobody realises they have nukes on board). We have not taken all this nearly as seriously as we should have either.

4. Combining 1-3 together clearly has potential for the sort of misunderstandings that are normal in history and have often caused wars but could now cause nuclear war. 1-3 ought to be core knowledge for senior politicians who ought to spend considerable time learning about such things. They do not.

Further, the more you think ‘Putin made a terrible blunder in invading Ukraine, he’s lost the plot, isolated by covid fear, the institutions around him don’t work, he’s fed lies by sycophants’ — which is the standard view in London and DC today — the more sceptical you should be that simplistic ideas from the Cold War about ‘rationality’ and deterrence would work as planned.

It’s easy to imagine how in the next few weeks 1-3 above combine disastrously. For example…

• NATO sends in supplies via Poland.

• Russian forces strike the convoy, intending to strike on UKR soil but an error happens and it’s a mile inside Poland. Some Poles are killed. In the fog of war there are claims and counterclaims about what happened. (Maybe Putin truly believes it was in UKR not Poland and western governments/media are lying. Maybe it happens bang on the border and there’s genuine room for doubt about what happened. Or a drone whacks something and its location is disputed because it wondered off course, which happens all the time.)

• The media goes crazy: Russian war crime, Putin worse than Hitler, NATO must act NOW. Remember, the media is totally and utterly unreliable on Russia. It has both ignored many awful aspects of the Putin mafia state for 20 years and invented nonsense about it. While individual journalists can be honest, you cannot rely on any serious corporate and generally enforced journalistic standards. Now MSM largely parrots Ukraine claims and bigshots retweet ‘Ghost of Kiev’ and video game footage as if it’s real. (E.g in order to suppress the Hunter Biden email story (in order to avoid helping Trump), practically the entire US mainstream media spread the fiction that the emails were ‘Russian disinformation’. They had former CIA people purporting to stand this up. Now everyone quietly, not on CNN, has to admit they were authentic and not ‘Russian disinformation’. This sort of disinformation from our own media is routine. You only know about it if you actively search for such things which few do. Graduates are the biggest suckers.)

• Emergency summits are called.

• Nuclear forces step up in readiness on both sides, or at least seem to (such signals have been misinterpreted before, e.g 1983 when a classified wargame was misinterpreted).

• And in this atmosphere you have the sort of thing that happened with Petrov, say an apparent nuclear launch from a submarine giving only minutes to decide response.

• What do they do? What does the UK PM, desperate to change the news from his lawbreaking now playing across the BBC again, do?

Below are some notes from Payne’s The fallacies of Cold War deterrence. All bold is added by me. I will also blog shortly on his followup book, The Great American Gamble, and The Kill Chain. The previous blog on defence is here.

The largely unseen rot of the UK nuclear enterprise

I’ve written about how our institutions deal with WMD-style crises many times. In 2017, I wrote about some of these issues in the context of Taiwan:

I will also post some notes on stuff connecting ideas about advanced technology and strategy (conventional and nuclear) including notes from the single best book on nuclear strategy, Payne’s The Great American Gamble: deterrence theory and practice from the Cold War to the twenty-first century. If you want to devote your life to a cause with maximum impact, then studying this book is a good start and it also connects to debates on other potential existential threats such as biological engineering and AI.

Payne’s book connects directly to Allison’s. Allison focuses a lot on the circumstances in which crises could spin out of control and end in US-China war. Payne’s book is the definitive account of nuclear strategy and its intellectual and practical problems. Payne’s book in a nutshell: 1) politicians and most senior officials operate with the belief that there is a dependable ‘rational’ basis for successful deterrence in which ‘rational’ US opponents will respond prudently and cautiously to US nuclear deterrence threats; 2) the re-evaluation of nuclear strategy in expert circles since the Cold War exposes the deep flaws of Cold War thinking in general and the concept of ‘rational’ deterrence in particular (partly because strategy was dangerously influenced by ideas about rationality from economics). Expert debate has not permeated to most of those responsible or the media. Trump’s language over North Korea and the media debate about it are stuck in the language of Cold War deterrence.

I would bet that no UK Defence Secretary has read Payne’s book. (Have the MoD PermSecs? The era of Michael Quinlan has long gone as the Iraq inquiries revealed.) What emerges from UK Ministers suggests they are operating with Cold War illusions. If you think I’m probably too pessimistic, then ponder this comment by Professor Allison who has spent half a century in these circles: ‘Over the past decade, I have yet to meet a senior member of the US national security team who had so much as read the official national security strategies’ (emphasis added). NB. he is referring to reading the official strategies, not the explanations of why they are partly flawed!

This of course relates to the theme of much I have written: the dangers created by the collision of science and markets with dysfunctional individuals and political institutions, and the way the political-media system actively suppresses thinking about, and focus on, what’s important. [Bold added]

I spent a lot of time in 2020 trying to push changes across ‘national security’ issues. I also went around some of the deep state sites to talk to people engaged in operations and try to figure out what important things were being starved of money/focus because of the general MoD horrorshow (leaked by a minister at the time to cause trouble). For example, I found our special forces are deprived of tiny amounts of money for crucial things — literally often so tiny you’re talking THOUSANDS, not even millions, while the MoD blows BILLIONS.

I spent time in the no-phones room under No10 disussing nuclear wargames and the UK nuclear enterprise.

In autumn 2020, I forced the PM to carve out 3 hours to discuss the nuclear enterprise. I had wanted to have a whole weekend at Chequers, including sessions with outside specialists, but he balked at just a few hours. After sort of listening, including to an account of rotten infrastructure and the truly horrific bills amounting to many tens of billions we face in coming years because of 25 years of rot and shockingly bad procurement under both parties, we left the room.

He picked up his phone (left outside for security), turned to me angry, and spat out.

What a waste of my time.

This sums up a lot not just about him but about our political system. Their single most important job is not seen as a priority!

Below the level of ministers and Permanent Secretaries there are many extremely able and dedicated public servants who have spent years trying to protect the UK. (The No10 Private Secretaries working on these things were exceptional and they’ve been repaid for their service by the PM embroiling them in his police investigations.) They pleaded with me to try to get ministers and other senior officials to listen and focus. They have been massively and repeatedly let down by politicians of both parties (and some senior officials).

These are the people now sitting in meetings being briefed on how Putin might interpret Truss saying ‘go and fight’ or Wallace giving Poland a blank cheque to attack Russia, as he did a few days ago (slapped down by Washington).

For many years I’ve said that a Golden Rule of politics is that, given our leaders don’t take nuclear weapons seriously never assume they’re taking X seriously and there is a team deployed on X with the incentives and skills to succeed.

People think this is an overstated metaphor but I always meant it literally.

Having explored the nuclear enterprise with deep state officials 2019-20, I can only stress just how extremely literally I mean this Golden Rule.

It was good that Daniel Finkelstein recently advised people to read Payne. I hope the current crisis means more in SW1 return to taking these subjects seriously and they return to being high status in SW1. Our MPs do not realise how the ‘people, ideas, machines’ of our nuclear enterprise have been allowed to rot by both parties. There is a closed loop of failure and classification of the failures. The same closed loop suppresses learning viz Russian/Chinese espionage and penetration of critical infrastructure, which is also much worse than is known to almost the entire House of Commons.

With this No10, it is totally impossible for this country to take these issues seriously.

Last year I urged MPs not to let No10 rewrite history over covid or we would hit the next crises with the same broken crisis management system that I warned in 2019 would fail, and we all saw fail. The most interesting thing about covid is not even the collapse and disaster. That was likely, not surprising. The interesting thing is the way in which both parties have sort-of-without-openly-coordinating colluded in trying to memory-hole the institutional failure and ensure that business-as-usual continues. (Seen Starmer press on how to change COBR, or even say one interesting thing about the management of No10/Cabinet Office? No.) You see the same phenomenon in D.C.

MPs did not act in 2021.

We got the totally avoidable omicron shambles.

And here we are on nuclear escalation. Yet again we have Boris Johnson sitting around the Cabinet table or in COBR with the same unstructured meetings, jumping from actual serious issue to ‘what does the Telegraph say’, with the few who understand often not in the room or ‘against the wall’ while ministers read out scripts (literally read out scripts). And the officials who understand walk out, clutch their heads and try to minimise the mayhem.

If you care about ‘preserving western values’, I strongly advise that you focus on regime change in London and Washington, not in Moscow. Otherwise we will face these recurring problems with rotten systems and leaders like Boris and Trump.

(I’ve stopped posting on Twitter and instead post small things here including on UKR situation. If you’re a journalist wanting more details on nuclear issues, in no circumstances will I discusss anything about this. You should encourage MPs to reform the Intelligence committee in Parliament. It’s a joke and treated as a joke by Whitehall so there is no forum for classified subjects to be properly discussed. This adds to the rot. This is why I told Boris that putting Grayling in charge of it was an immediate sign of him failing in his responsibilities, one of our first rows in Jan 2020. Similarly the way some ministers disgracefully leak from NSC meetings has meant that people such as ‘C’ and the head of GCHQ are rightly very guarded in what they say. This also adds to the rot.)

‘Rational’ does not mean ‘reasonable’ but western leaders have confused them

American leaders repeatedly define ‘rationality’ in a certain way then expect their opponents to behave predictably which includes responding ‘rationally’ to pressure from Washington.

A definition of ‘rationality’ supposedly brings understanding, predictability, and control. However, Payne argues ‘reason’ and ‘reasonableness’ should be considered very differently.

Rationality:

[It’s] a mode of decision-making that logically links desired goals with decisions about how to realize those goals. For the rational decision-maker, a particular course of action is chosen because, based on available information, that course is calculated to be most suitable for achieving the preferred goal. A rational decision-maker also is expected to prioritize goals—some being more important than others—and recognize that trade-offs among goals may be necessary… Rationality does not imply that the decision-makers’ prioritization of goals and values will be shared or considered “sensible” to any outside observer. The goals and values underlying decision-making do not need to be shared, understood or judged acceptable by any observer for the decision-making to be rational…

The rationality of a leader’s decision-making process is not related to the question of whether the desired goals, or chosen routes to those goals, will be compatible with accepted norms, customs, and values. The goals and/or actions of a rational decision-maker may seem unreasonable to an outside observer, even bizarre, without compromising the rationality of the decision-making process.

Reasonableness:

The judgment that another’s decision-making and behavior is “reasonable” typically implies much more than its “rationality.” Pronouncing another’s decision-making or behavior to be “reasonable” suggests that the observer understands that decision-making and judges it to be sensible based on some shared or understood set of values and standards.

Confusing the two is disastrous.

To assume rationality, however, and on that basis to expect behavior that is reasonable, that is behavior predictably driven by familiar, understandable norms and goals, is to risk lethal surprise.

Leaders often described as irrational such as Hitler and Saddam in fact were rational by this minimal standard just as a shopper is rational if he accepts that buying one thing means less cash to buy something else, but rational people can disagree with how one judges the value of different products.

In short, rational decision-making can underlie behavior that is judged to be “unreasonable,” shocking, and even criminal by an observer because that behavior is so far removed from any shared norms and standards. Historically, relatively few leaders have, in fact, been functionally irrational. Many, however, have been quite unreasonable. Those commonly labeled irrational, such as Adolf Hitler and Saddam Hussein, appear on closer examination to have been capable of rational decision-making. The goals and values underlying their decision-making, however, were far outside familiar norms. As a result, their behavior surprised observers.

Washington tends to define rationality with its own ideas and therefore comes to the comforting conclusion that its opponents can be manipulated by carrots and sticks. A lot of hard, confusing work to figure out what X really thinks and why can be ignored. Once these assumptions are made, then conflicting evidence is routinely ignored or “rationalised” away. If surprising stuff happens, the bureaucracy will try to explain it away, often for so long it wastes precious time and generates disaster.

We will not be surprised by an opponent we define as pragmatic and rational; and, with suitable levers, we can control its behavior. What could be more comforting?…

[D]ecision-making and behavior considered scarcely plausible in Washington can appear entirely reasonable to a foreign leadership, not because foreign leaders are irrational, but because the definition of what constitutes ‘reasonable’ can differ so dramatically. Consequently, surprises are frequently in store for those who believe that a foe’s basic rationality permits confident prediction of its behavior.

A lot of people in Westminster like to talk about ‘the rules-based international order’. Most of them genuinely think it makes everyone richer, safer and freer. But such beliefs can easily cause crucial people to kid themselves about what people like Putin and Xi think.

Rationality/reasonableness and Chamberlain/Hitler

Hitler repeatedly and openly stated his goals including ‘blood cleansing’ of Europe.

Adolf Hitler, for example, held fairly consistently to a set of goals from at least the early 1920s until his suicide in 1945, and he initiated policies logically linked to the realization of those goals. The goals and policies themselves reflected a combination of Hitler’s intense hatreds and hubris, his aggressive rejection of “bourgeois” moral norms, his fanatical belief in racial myths, in his own role as the Führer of Germany, and in his race-centered version of Social Darwinism. Nevertheless, Hitler could be rational in the sense that he was capable of calculated decision-making that logically linked his policy choices to his ghastly goals.

Chamberlain and others decided not to take most of his statements seriously. They persuaded themselves that much of what he said was not ‘serious’ and thought instead that “reason” would draw him away from those goals.

Chamberlain’s behaviour probably forestalled a coup to remove Hitler. It definitely induced contempt.

Our enemies are little worms, I got to know them at Munich, said Hitler.

[If I remember right, after Britain declared war in 1939, Hitler, like Wilhelm II in 1914, was shocked.]

NB. Taking seriously what people like Hitler, Lenin, Stalin, Mao et al say causes immediate severe problems. You have to face things like military weakness, budgets, spending less on other things, making long-term sacrifices, political problems. So there are very powerful incentives pushing minds to ‘rationalise’ away their public statements and observable behaviour. If X does not materialise for 5-10 years, it’s someone else’s problem.

Sound familiar?

(I’m extremely critical of Cameron/Osborne but imagine if they’d said in 2010: ok on top of the economic nightmare we also need to change approach to Putin, those who said he’s mafia and it’s mafia government were right, we need to cut spending on XYZ and make long-term sacrifices to…’ They’d have been shouted down by Labour and most of the media. Similarly: imagine if the Cabinet Secretary had said in 2019: PM, yes we have the biggest constitutional crisis since 1914 but you must focus on pandemic preparations, it’s a disaster… When I started recruiting people to deal with disaster response in Jan 2020 when covid was already here and spreading, it was mostly seen in SW1 as loony.)

US ideas about Soviet ‘rationality’ and nukes were wrong

According to Payne, America planned its nuclear strategy on the basis of a certain concept of “rationality” that we now know was false, that the Soviet leadership would be reasonable by Washington’s standards, and “this expectation was contrary to much of the evidence available at the time.”

Soviet war plans for Europe involved “the early and heavy use of nuclear weapons” and NATO’s planning was “far removed from the realities of Soviet planning”.

Some now believe this good fortune and our survival of the Cold War to be the result of our mastery of nuclear deterrence policy, and therefore advocate extending that policy into the post-Cold War period.

Soviet responses were thought to be understood and predictable:

[B]ecause US nuclear threats were so lethal, any sane opponent would be deterred from extreme provocations. This confident expectation was not based on the close examination of any challenger’s particular beliefs and filters. Rather, it reflected the heroically optimistic assumption that rationality and reason, as understood in Washington, would overcome any particular or unique beliefs and modes of thought that might otherwise lead a challenger away from predictably bowing to severe US nuclear threats.

“Overconfidence in Washington” based on the false belief of wise Cold War policy and mastery of deterrence will lead to the extension of false ideas into new international relations. Future opponents will “astonish” the US and, given WMD, may well bring “an unprecedented catastrophe”.

CH1: Surprise, Surprise

Russia, 1904. Tsar Nicholas thought Japan was too weak to attack and dismissed the possibility.

JAP, 1941. It was confidently predicted that Japan would not attack America. Eg. Acheson told Roosevelt in August 1941 that “no rational Japanese could believe an attack on us could result in anything but disaster for his country.” Japan attacked four months later because it thought itself doomed if it did not.

China/Korea. In November 1950, General MacArthur predicted to Truman that China would not intervene in the Korean War because Mao and Stalin feared American reprisals. The CIA’s “daily summaries” made this prediction up to 17 November. On 25 November, China threw nearly 200,000 troops over the Yalu River. Stalin said to Mao: “If a war is inevitable, then let it be waged now.” Mao thought America planned to encircle China (from Korea, Indochina, and Taiwan) and agreed with Stalin. Both sides misunderstood each other: China wrongly thought America was preparing a planned attack on China, America wrongly thought China would not strike pre-emptively.

Cuba. On 19 September 1962, less than a month before photo evidence proved Soviets had placed missiles in Cuba, the CIA predicted that the Soviets would not put missiles into Cuba. It “would be incompatible with Soviet practice to date and with Soviet policy as we presently estimate it. It would indicate a far greater willingness to increase the level of risk in US-Soviet relations than the USSR has displayed thus far”. Afterwards, Sherman Kent (head of the National Board of Estimates) said, “We missed the Soviet decision to put the missiles into Cuba because we could not believe that Khrushchev could make a mistake.” Bundy described the move as “crazy”. A senior State official said that he initially rejected the reports because “I didn’t want to believe them.” In 1998 McNamara said the world came ‘within a hair’s breadth of nuclear war’ and ‘we lucked out’.

Vietnam. During Vietnam, America’s political and military establishments repeatedly predicted that the North Vietnamese would cave in because of fear. McNmara later said, “The North Vietnamese were prepared to absorb far greater punishment than was ever delivered by the American bombing.”

Egypt, 1973. In 1973, American and Israel ignored evidence that Egypt and Syria would launch a war because, as Kissinger said, “no one believed” the Arabs would attack; “Our definition of rationality did not take seriously the notion of [them] starting an unwinnable war to restore self-respect. There was no defense against our own preconceptions.”

Iran, 1979. In 1978, President Carter described Iran as “an island of stability”; the following year, the Shah was history. Then Director of the CIA said, “We didn’t adequately predict the fall of the Shah… we were unwilling to believe that he would not call out the troops when the crisis came and spill blood on the streets if necessary… [W]e didn’t make the right assumption.”

Kuwait. In 1990, US intelligence did warn politicians of an imminent threat of attack by Iraq on Kuwait. Bush and his senior officials did not believe it. Ambassador Glaspie said of Saddam, ‘We foolishly did not realize that he was stupid.’

1990s US intel on WMD. In 1995, the US intelligence community predicted that “no country, other than the major declared nuclear power, will develop or otherwise acquire a ballistic missile in the next 15 years that could threaten [continental America].” In 1998, North Korea tested a missile with potential ICBM range, prompting a new assessment that it was “most likely” that such a danger would occur. The State Department had a misplaced belief that other countries had an attachment for arms control regimes (such as the Missile Technology Control Regime) that mirrored their own. Russia and China are quite happy to violate agreements they have officially agreed to.

India, 1998. The Indian party’s campaign commitment to nuclear testing was dismissed and it was expected that in government they would move away from nuclear weapons. When instead the new government cracked on with their plans, Washington’s reaction was — to question their rationality! In 1998, the CIA learned about India’s nuclear tests from CNN. Albright suggested to India that “the rest of the planet is racing to leave [nuclear weapons] behind”. US intelligence had defined such Indian behaviour as irrational and therefore predicted it would not occur. Clinton officials briefed the media that their error had been to assume India would behave “rationally”. The subsequent inquiry concluded that intelligence officials had not understood that India might place a high value on something that DC thought was ‘irrational’.

In 1999, America was surprised by Serbia’s resistance to bombing. Milosevic said, “I am ready to walk on corpses and the West is not. That is why I shall win.”

In 1990s Washington, it was ‘irrational’ for countries to pursue nuclear weapons therefore they thought it would not happen. In fact some countries thought it rational and did pursue them.

[Cf. Richard Betts, Surprise Attack.]

CH2: Cold War Deterrence Theory and Practice

America thought that if both sides possessed a manifest and secure nuclear capability for massive destruction, then mutual deterrence would be ‘stable’ because no rational enemy would risk widespread mutual destruction. This amounted to the tautology: rational leaders would be deterred otherwise they would irrational.

Thomas Schelling, an economist who also wrote much about nuclear strategy, explicitly identified ‘mirror imaging’ as a useful method for conceiving nuclear strategy:

You can sit in your armchair and try to predict how people will behave by asking how you would behave if you had your wits about you. You get, free of charge, a lot of vicarious, empirical behavior.

[Sounds like a recipe for disaster!]

Payne writes:

This methodology was taken to the extreme in the United States. With little or no reference to the specific thought, goals, and values of the Soviet leadership, numerous officials and academic commentators assumed that they knew how any “sane” Soviet leader would view nuclear weapons, and how deterrence would therefore operate…

Actual U.S. nuclear targeting policy evolved over time, but for much of the Cold War deterrence analyses simply posited that the Soviet Union would be the rational, pragmatic, and reasonable (and hence predictable) image of the United States, and proceeded to model the outcome of large-scale nuclear wars…

With this methodology, analysts and commentators offered highly confident and precise answers to questions such as, would deployment of accurate ICBM warheads, missile defense, or single-warhead mobile ICBMs “stabilize” or “destabilize” deterrence? The center of attention in this approach to deterrence was on U.S. and Soviet strategic nuclear forces, including ICBMs, long-range bombers, and submarine-launched ballistic missiles (SLBMs).

Bundy (NS Advisor to Kennedy and Johnson) explained the tautologous logic in 1969. ‘In light of the certain prospect of retaliation there has been literally no chance at all that any sane political authority, in either the United States or the Soviet Union, would consciously choose to start a nuclear war. This proposition is true for the past, the present and the foreseeable future.’

Consequently, US strategists tended to conclude that things that perpetuated mutual vulnerability were ‘stabilising’ (e.g. inaccurate SLBMs) and other things ‘destabilising’ (e.g. missile defence (BMD), or an accurate force able to strike pre-emptively).

As Harold Brown, SecDef in 1979 said, ‘In the interests of stability, we avoid the capability of eliminating the other side’s deterrent…’

Hence, in 1972, the ABM Treaty effectively banned BMD and the US pursued for two decades, via SALT and START, the limit of more accurate (therefore more ‘destabilising) strategic weapons.

‘Stable’ deterrence would follow from mutual vulnerability to secure, retaliatory threats of nuclear annihilation.

Nuclear strategy became quite simple and numerical. Strategists had a set of things to count and measure. They did not have to worry about the actual thoughts and feelings of their enemy.

Fred Kaplan wrote about how senior officials under Secretary McNamara thought about basic deterrence:

It all appeared scientific and precise, but in fact it had little to do with any formulation of how much would be to enough to deter the Soviets. It was the output of a computer program designed by Alain Enthoven ‘laying down’ 1-megaton bombs against Soviet cities and calculating, at various points, how much additional damage one additional bomb would do.

American thinking did not reflect the reality of Soviet thinking.

For NATO, nuclear warfighting was unthinkable.

NATO policy sought to maintain a lid on the potential for nuclear escalation via sophisticated “intra-war deterrence” concepts of rational wartime bargaining and “limited nuclear war.” These U.S. deterrence concepts mistakenly assumed a similarly minded, “rational” opponent, and thus were wholly incompatible with Soviet war plans [which] called for very heavy and very early nuclear and chemical strikes throughout Western Europe in the event of war. As a consequence, had large-scale war occurred in Western Europe, it is likely that Western civilian leaders would have looked to their Soviet counterparts for a “rational” mutual deterrence contest of limited war, bargaining and escalation control, while the Soviets planned for large-scale nuclear and chemical strikes throughout Western Europe. The West’s expectations of Soviet “rationality” could hardly have been more ill-fitting with apparent Soviet war plans…

The possibility that the Soviets might steadfastly resist Washington’s logic of mutual nuclear stalemate was, against considerable evidence, dismissed by most of Washington’s cognoscente.

Post-1991 access to archives and interviews has shown western views were false.

Only recently, courtesy of greater access to past Soviet decision-making practices, has it become virtually unarguable that the Soviet leadership never accepted the West’s definition of rationality with regard to nuclear weapons, and that Soviet expectations of U.S. behavior, largely derived from the dogma of Marxist-Leninist ideology, appear to have been a significant factor in Soviet nuclear war planning. Indeed, greater access to Soviet archives demonstrates how consistently Soviet decision-making appears to have been shaped more by ideological vision than by pragmatic, realpolitik.

Odom argues that the Soviet leadership thought a nuclear war could be won and rejected western ideas about ‘stability’.

It was not until Nixon’s Schlesinger Doctrine of 1974 that Washington even claimed to have tailored deterrence to specific ideas about the Soviet leadership (p.24ff).

Academics wrongly accepted Washington’s confidence.

• Joseph Nye and Allison in 1988 wrote that ‘current nuclear postures have substantially solved the problem of deterring deliberate nuclear attack’ (W. Post, 4/9/88).

• Michael Howard gave a similarly confident and wrong verdict (cf. Survival, Vol. 43, 1994).

• In 1995, John Deutsch, Deputy Secretary of Defense, said that ‘Deterrence is ensured by having a survivable [nuclear] capability to hold at risk what potentially hostile leaders value’.

It’s important to realise that the foundations of the official view were very widely shared across GOP/DEM and so-called hawks/doves. Arguments focused on the type and number of weapons that produced ‘stability’.

All such analysis boils the problem down to the capability to destroy and the credibility of threats.

But deterrence is psychological, it is not just about the size and capabilities of respective forces.

CH3: Why the CW deterrence framework is inadequate

Kagan’s On the Origins of War concluded that practical utility and material gain play a small part and ‘some aspect of honor is decisive’ in decisions for war and peace.

One of the great passages in Thucydides tells how Athens made clear, credible and severe threats to destroy Melos. Melos rejected surrender and was destroyed.

When Pilate brought images of Caesar into Jerusalem, in contravention of Jewish opposition to graven images, there was a protest. Pilate issued a severe, well-understood, and credible threat: either stop protesting or be slaughtered. The Jews lay on the ground and laid their necks bare saying they would rather be dead than accept the transgression of their laws. Pilate “was deeply affected with their firm resolution” (Josephus) and withdrew Caesar’s effigy from Jerusalem.

In 1914, the Russians decided to fight because they were worried about the loss of prestige despite having no vital strategic interest in the Balkans and despite everybody thinking that their relative power would increase in coming years so they had an incentive to wait.

General Anami tried to persuade the Japanese leadership not to surrender after Hiroshima and Nagasaki but to fight on. ‘Would it not be wondrous for this whole nation to be destroyed like a beautiful flower?’ he said.

General Galtieri was warned before he invaded the Falklands that Britain would fight but he replied, “Why are you telling me this? The British won’t fight.” After the war when asked about perceptions of the chances of an aggressive British response, he said, ‘Though an English reaction was considered a possibility, we did not see it as a probability. Personally, I judged it scarcely possible and totally improbable… Why should a country situated in the heart of Europe care so much for some islands located far away … which do not serve any national interest? It seems so senseless to me.’

Castro urged Krushchev to launch a nuclear war in the event of an invasion. Krushchev said that Castro was ‘very hot tempered [and] failed to think through the obvious consequences of a proposal that placed the planet on the brink of destruction.’ There is a lot of evidence that Castro and Che Guevara were aiming for a nuclear war with America in which they knew Cuba would be destroyed. Soviet Deputy Premier Mikoyan complained to Krushchev that the Cubans were ‘emotional, nervous, high-strung, quick to explode in anger, and unhealthily apt to concentrate on trivialities’ and ready to ‘die beautifully’. One of the participants said the Cubans believed, ‘Cuba will perish but socialism will win.’

One of Saddam’s generals told after 1990 that he was accompanied by two astrology-based soothsayers (one of whom was a twelve year-old) who kept telling him he would win. Saddam himself warned that he would fight despite overwhelming odds because ‘without pride life has no value’. Many of Iraq’s leaders thought that America was deliberately constricting Iraq via Kuwait and the choice lay between fighting and being ‘suffocated and strangled’ (Aziz) in the future.

General Yakovlev, Commander in Chief of the Russian Strategic Rocket Forces, said astrology is ‘a quite serious science’.

It helps us launch spacecraft, missiles; we use it broadly to forestall suicides among the personnel. Experience shows it is unreasonable to reject it.

Nancy Reagan provoked dismay by repeatedly consulting with a psychic astrologer, Joan Quigley.

According to the Syrian Defense Minister Mustafa Tlas, he warned Lebanese terrorists not to harm Italian soldiers because he was so fond of the beautiful Italian Gina Lollobrigida.

Hitler elevated race and brutality explicitly saying that normal ideas of human morality (‘a mixture of stupidity and cowardice’) must be smashed (‘We want to be barbarians! It is an honourable title’). He had total faith that some sort of non-Christian Providence guaranteed his success. ‘I will follow the path assigned to me by Providence with the instinctive sureness of a sleepwalker… Trust your instincts, your feelings…Never trust your knowledge.’ ‘I have to attain immortality even if the whole German nation perishes in the process.’ At the end, his ‘Nero orders’ were intended to utterly destroy the remaining people and material in Germany because he thought that the East had proven itself stronger and Germany deserved to perish. Speer judged that at the end Hitler wanted to destroy all Germans: “He tried to throw them definitely into the abyss.” Hitler was also taking large, daily dosages of stimulants and sedatives, and from 1944 daily cocaine.

What are their personal beliefs; do they believe in God, astrology? What is their temperament; are they happy to keep doubling down and ‘die beautifully’; do they think of themselves as almost God-like, like Mao who was prepared to sacrifice hundreds of millions of Chinese to achieve his goals?

What cognitive errors distort judgement (Hitler cancelled the V2 rocket program because of a dream and only Speer’s intervention restored it)? What about drugs?

• Hitler: coke, speed, sedatives.

• Mao: barbiturates.

• Kennedy: steroids and amphetamines.

There were notable exceptions in the defence establishment who objected to the fallacies of Cold War deterrence: e.g Ikle, Alexander George and Richard Smoke. In 1964 the American Group for the Advancement of Psychiatry issued a pessimistic statement on the prospects for deterrence because ‘it rests on certain dubious psychological assumptions’. George and Smoke described deterrence theory as ‘seriously incomplete to say the least’. Nathan Leites (RAND) warned in 1951 that America should study the particular ideas of enemy leaders (The Operational Code of the Politburo).

The exceptions were exceptions, and largely ignored.

CH4: CW deterrence in post CW world

After the Cold War, senior US figures repeatedly stated that the strategy of deterrence that had ‘succeeded’ in the CW would continue to work in the future including against emerging ‘rogue nations’.

• An October 1998 report by the Department of Defense’s Defense Science Board, Nuclear Deterrence, attempts to “define concepts of deterrence relevant to the changing world.” It describes the deterrence requirement for “the assurance that no rational adversary could believe they could gain by employing nuclear weapons (or other weapons of mass destruction) against the US or an ally under the US nuclear umbrella.” The report subsequently focuses almost entirely on U.S. capabilities for nuclear threat to provide that “assurance.”

• Then CIA boss Tenet said (2000), ‘To a large degree, we expect our mutual deterrent and diplomacy to help protect us [from Russia and China] as they have done for much of the last century.’

• Admiral Blair, Commander in Chief of the Pacific Command, said (2000) that there was a ‘very strong deterrent capability’ against North Korea because the leadership knew that a conflict would mean ‘the end of their regime’.

• In 1996 Defense Secretary William Perry said that “no rogue nation today has ICBMs; only the established nuclear powers have ICBMs. And if these powers should ever pose a threat, our ability to retaliate with an overwhelming nuclear response will serve as a deterrent. Deterrence has protected us from the established nuclear arsenals for decades, and it will continue to protect us.”

• Influential professor Stephen Walt has said that, ‘If we could deter the “evil empire” for four decades, we can almost certainly deter today’s rogue states.’

• ‘Spurgeon Keeny, executive director of the Arms Control Association, goes so far as to claim repeatedly and against all logic that even “irrational” challengers will be deterred reliably because U.S. threats are so lethal: “Even fanatical, paranoid regimes are deterred by the prospect of catastrophic consequences,” and, “Even the most ‘irrational’ rogue state would be deterred from threatening, much less undertaking, such an attack [against the United States] by the prospect of overwhelming U.S. retaliation.”’

Washington also assured Russia that its deterrent would not be compromised by any plans for missile defence.

The flawed logic is that all rational challengers will:

1. Value most highly that which Washington can threaten.

2. Comprehend and believe U.S. threats to those values.

3. Calculate that conciliation to Washington is preferable to risking U.S. wrath.

4. Thus be deterred reliably and predictably.

This logic is based on the CW ‘mirroring’ analysis which ignored psychology, history and other fields. It is therefore wrong because deterrence is necessarily about psychology and should weigh the chances that other leaders will, like Hitler, Mao or Castro, consider risks very differently than Washington. Those who come to power in bloody coups are likely to be very different sorts of people than those who rise to the top of Washington bureaucracies. ‘[L]eaders can hold to distorted, self-serving interpretations of reality, rely on dubious sources of information, be motivated by extreme emotions and goals, and esteem some values more highly than their own lives and positions.’ Hitler issues Nero orders. There can be no ‘assurance’.

CW logic applied to the post-Cold War world has another major problem. NATO’s Cold War deterrence theory was:

[A]lthough the Soviet Union possessed significant conventional force advantages in Europe, Soviet fear of NATO’s threat of nuclear escalation (nuclear “first use”) would deter it from any sizable Soviet conventional or nuclear attack. It was believed that NATO’s nuclear escalation threat presented an unavoidable risk of costs to the Soviet leadership that would far outweigh the benefits of any purposeful offensive aimed at Europe. The Soviet leadership would be deterred because it could never be confident that it could exploit its conventional force superiority without triggering NATO nuclear retaliation.

Leave aside that this logic was unhistorical, and Soviet leaders actually calculated differently. Assume the logic is correct. Washington also argues that it can deploy conventional superiority against regional powers with WMD because the regional power will be deterred from escalating to WMD. But this contradicts DC’s logic viz the Soviets!

Given NATO threatened nuclear war in the event of conventional attack and thinks this worked to deter such a conventional Soviet attack, why does it think that other regional powers now will not think similarly? Further, in such a regional conflict it is likely that the local power will be much more highly motivated than will America.

Sound familiar?! How much does Putin care about UKR versus the median American voter?

CH5: A new direction

Nothing can ‘ensure’ deterrence:

… no particular nuclear balance, no weapon system, no declaratory policy, no technological advance, no presidential statement, no intelligence breakthrough, and no organizational gimmick can predictably “ensure” deterrence. Doing so would require omniscience and omnipotence, qualities even Washington lacks.

We ultimately cannot prevent leaders from making decisions based on ignorance, folly, the self-serving or self-induced distortion of reality, or a conscious willingness to court disaster. At some point in America’s future deterrence assuredly will again unexpectedly fail as a result of one or a combination of these factors.

As Sun Tzu said, defeating the other’s strategy is the highest form of war but it requires that you ‘know yourself and know the enemy’.

US force structure and threats should depend on answers to questions.

• What are the enemy’s goals, motives, morale, how do they make decisions, how do they view risk, how do they view you?

• What are the stakes?

• What is the regional political/security context?

• What sources of power are available to each?

• How do certain strategies and operations match up against each other, and therefore iteratively how might they be improved?

Payne provides a detailed framework for how to assess threats and deterrence. He stresses that all such analysis will have holes, often serious, and we must drop talk of ‘assurance’ and ‘ensure’ deterrence. There can be no assurance.

The official 1990s policy was that START III levels of nuclear weapons would by definition be effective against an open set of WMD challengers with unknown motives and stakes.

CH6: Testing the deterrence framework - China/Taiwan

Payne applies his frmework from CH5 to a Taiwan scenario.

In 1996, CIA Director John Deutsch admitted that they knew ‘very little about Beijing’s future leadership and intentions’. Many China specialists argue that Chinese leaders emphasise the value of exploiting opportunities and are less risk averse than Washington. They also look for the brilliant coup for which they have intriguing names (such as ‘kill with a borrowed knife’, ‘the empty city’) in order to avoid ‘exhaustion in protracted fighting’ (qiongbing duwu). They have grown up learning Sun Tzu’s maxims about the possibilities of bluff and deceit.

Some say also that Chinese expect formal courtesy in private meetings and therefore discount what is said while focusing on public comments, which often confuses Americans who tend to regard public statements as much more dishonest and often said for domestic political reasons.

Taiwan is seen as an existential issue of the survival of China and the survival of the regime. The humiliation of Japan ruling it 1895-1945 then losing it in the civil war must be erased. Deng said, ‘Whoever lost Taiwan must step down and stand condemned through the ages.’ Many other senior Chinese have repeated similar ideas. They fear it would lead to a domino effect in non-Han regions, such as Tibet, Xinjiang, and Inner Mongolia. They fear that ‘danger from without’ will bring ‘trouble from within’. This connection also leads them to extreme suspicion of foreigners’ motives in discussing China and Taiwan.

Military preparations are aimed at deterring American involvement and China has repeatedly raised the stakes. Mao originally embarked on a nuclear program to escape American intimidation during Korea and the Taiwan crises of the 1950’s. Mao was responsible for the deaths of tens of millions and remarked on the ability of Chinese to ‘eat bitterness’.

PLA Lt. Gen. Xiong Guangkai, deputy chief for intelligence and foreign affairs, claimed to Charles Freeman in 1995 that China would sacrifice ‘millions’ and ‘entire cities’ to prevent Taiwan’s independence: ‘In the 1950’s, you three times threatened nuclear strikes on China [Korea, 1954, 1958] and you could do that because we couldn’t hit back. Now we can. So you are not going to threaten us again because, in the end, you care a lot more about Los Angeles than Taipei… [W]e’ve watched you in Somalia, Haiti, and Bosnia and you don’t have the will.’

Zhu explicitly warned in 2000 that it would a huge error for America to think that the balance of forces was such that it was obvious that a war over Taiwan would be lost therefore China would duck it: ‘… those people … do not understand and do not know about the Chinese history. The Chinese people are ready to shed blood and sacrifice their lives to defend the sovereignty and territorial integrity of the motherland.’

Generally, the PRC has not been a country that bluffs about force. When it threatens, it is prepared to act.

China has invested enormously in a range of weapons to counter America’s ability to interfere. I won’t go into this here, I discussed it a bit in this piece on Allison’s book and will return to it. Obviously since Payne’s book was written China’s capabilities have grown enormously.

America has not been clear in the 1990‘s. ‘We don’t know what we would do and you don’t,’ said Nye to Chinese officials in 1995. Various senior US officials have warned Taiwan publicly and semi-publicly that America would not fight for them if they declare independence. Clinton also said in 1998 that America does not support Taiwan independence.

Payne summarises:

• PRC leaders are calculating;

• They are willing to take significant risks;

• Their willingness to absorb costs is high;

• The fate of Taiwan is a survival issue for them;

• There is a political consensus in China for reunification;

• Taiwanese independence is unacceptable to Beijing;

• PRC leaders are ready to use force to protect Chinese territorial integrity in general, and in particular to deny Taiwan independence;

• In the absence of U.S. intervention, China ultimately can subdue Taiwan;

• PRC leaders believe the “stakes” over Taiwan are far less significant for the United States than they are for China, and view the U.S. commitment to Taiwan in this regard as uncertain;

• PRC leaders consider Washington unwilling to absorb significant costs for the purpose of preventing China from subduing Taiwan;

• PRC leaders believe that Washington will be vulnerable to Chinese deterrence threats in a crisis over Taiwan;

• For Beijing, keeping the U.S. out is the key to success.

He concludes:

Consequently, to conclude that Chinese leaders … simply may not be deterrable in practice is not to question Chinese rationality. It is a conclusion logically derived from an attempt to understand the basic features of the Chinese (and U.S.) cost-benefit and risk calculus in a specific context.

Payne summarises what he thinks the actual PRC situation is viz a notional deterrable opponent.

My summary:

1. It is clear that there is a very widespread view in China that losing Taiwan risks not just the collapse of the PRC’s grip on power but the crackup of the country, as has happened before, therefore the risks of a major war with America are worth paying, even including nuclear strikes. This is not an idiosyncratic view held by one or two people. The entire leadership has similar views, there is no meaningful hawk/dove division.

2. They also think: Taiwan is an island on the other side of the Pacific that most Americans could not identify on a map, it is self-evidently not an existential issue for America.

3. Further, if America starts threatening over Taiwan, it may well be seen as part of an attempt by the American government to collapse the regime/country.

4. China built nuclear weapons partly because Mao experienced nuclear threats from America in 1954 and 1958.

5. China suffered tens of millions of deaths under Mao. People like Xi who have grown up in this culture think very, very differently about death to western politicians.

6. China could easily conclude that it has been entirely clear that it would risk nuclear war over Taiwan because it is an existential issue for the nation and regime, while simultaneously America could wrongly think that a nuclear war would be so destructive for China that it must surely back down.

7. China may also wrongly conclude that America will not dare intervene because, to the Chinese, it would be stupid to trigger war for a non-vital issue and Americans are too weak to risk large casualties. American leaders including Clinton have stressed they do not support independence or Taiwan being streated as a separate state.

8. Given America’s ‘strategic ambiguity’, it would not be a surprise if China were confused by a strong American response in a crisis, just as Galtieri later complained that it was stupid for Britain to intervene over irrelevant islands thousands of miles away. British ambiguity over Belgium was catastrophic in 1914. It could easily be catastrophic for America.

9. Further if America were to fight, then that would surely be taken as evidence in China that America really does want to destroy China and therefore they were right to resist! Mao thought (wrongly) in 1950 that America was planning to invade hence his attack.

10. The American public will not support nuclear threats over Taiwan. European publics will not support US nuclear threats over Taiwan.

11. It is likely that there is no US deterrence policy that would be viewed by China as both sufficiently damaging and credible such that it would deter China.

I don’t see any sense in America threatening war/escalation over Taiwan. (I said this in 2017, it isn’t a response to UKR.) It should stress there should be a peaceful resolution. It should resign itself to Taiwan rejoining China. It should think about actually credible red lines for China’s expansion. And it should prepare for Japan to go nuclear and for Australia to ask for a NATO-style pledge from America and Britain to threaten China with nulcear war in the event of invasion. When China takes Taiwan — and it’s when, not if — it will inevitably provoke counter-moves from other regional players. Shaping this is much more important for America than trying to interfere in what it has itself conceded is an internal Chinese issue on the other side of the Pacific.

I strongly recommend reading Lee Kuan Yew’s views on this. He was a realist with a long-term view. He tried to limit US aggression on this subject and stressed that America should allow and encourage a peaceful unification in time, and avoid confrontation. He also feared a disastrous outcome of DC neocons controlling policy.

Lee Kuan Yew understood the world much, much better than those neocons in Washington who talked so much nonsense about Afghanisatan and Iraq and now want America to threaten nuclear war over Taiwan ‘for freedom’.

CH7: The new deterrence framework

Joseph Nye, as Assistant Secretary of Defense for International Security Affairs, confidently claimed:

I know how to deter [Chinese missiles]. If deterrence prevented 10,000 Soviet missiles from reaching the United States, it baffles me as to why it wouldn’t prevent 20 Chinese missiles from reaching Alaska.

a. How do you know deterrence prevented Soviet missiles flying?

b. Even if it’s true it doesn’t mean it’s true viz China/Taiwan.

Many US officials and think tanks etc have spoken similarly, as if all that matters is a/ assume China is ‘rational’, b/ we have more/better nukes than them.

If you want to make deterrence work against China, America must acquire greater capabilities including BMD and make political changes such that this change is more credible.

But even extreme moves in this direction do not guarantee success and may well be seen as non-credible in China.

NB. This pessimistic conclusion seems justified by developments since Payne wrote.

CH8: Lessons of this case study

First, how to ‘deter the deterrent’ is a new challenge unsolved in the CW (though people wrongly think it was!).

Nothing can ‘ensure’ the functioning of deterrence… With sufficient work and available evidence, we should be able to reduce our ignorance about how a specific opponent thinks, and thus improve our chance of anticipating how it will behave in response to U.S. deterrence policies. But the hubris regarding our mastery of nuclear deterrence ‘stabilty’ that was common during the Cold War, and extended by many into the post-Cold War period, is built on the demonstrably false assumption that Washington’s interpretation of what is rational and sensible behavior also will be the basis of our opponents’ behavior.

In truth, no matter how lethal the U.S. deterrent threat, no matter how ‘rational’ we regard an opponent’s choice to bow to U.S. demands, there remains an irreducible level of uncertainty concerning how an opponent will behave in response to US threats. And in no prospective case will we even be able to define that level of uncertainty with precision. In the future, the United States will face crises in which there can be no confident expectation that its deterrence threats will control the opponent’s behavior. The prospects for deterrence success may be know only after the fact… [From start but I’ve put here]

There is no existing basis for believing that U.S. leaders will be any more willing to persist in the context of nuclear threats to U.S. cities than we believed Soviet leaders would be willing to provoke NATO’s nuclear threat during the Cold War. Indeed, U.S. efforts to deter and coerce will likely be at a great disadvantage in terms of the costs, benefits, and the interests involved in prospective regional crises when American cities are vulnerable to WMD escalation.

Second, to deal with this problem and use force regionally against powers with WMD, the US will have to develop new capabilities, including regional anti-access/area denial, missile defence. BMD is not an ‘isolationist’ idea, it will be necessary to allow US forces to operate in regional conflicts. The absence of a BMD is ‘a particularly egregious vulnerability’.

Third, America must make far greater efforts to understand its adversaries in detail and not rely on the simplicities of the CW framework.

If we cannot fully understand the acts of other people, until we know what they think they know, then in order to do justice we have to appraise not only the information which has been at their disposal, but the minds through which they have filtered it. Walter Lippmann

Simply cutting nuclear forces on the assumption that it only needs a few to deter a ‘rational’ enemy is folly. Washington’s dynamics make it almost impossible to review arms control treaties once they’re agreed. They lock everyone’s thinking in place in dangerous ways. GOP and DEM governments seek to simultaneously uphold the treaties and evade them, leading to technical and legal contortions. E.g the ABM Treaty:

The United States has been locked into the treaty by a domestic norm that views strategic arms control agreements as sacrosanct… The past deterrence framework, and the approach to arms control derived from it, are wholly ill-suited to the fluidity of the post-Cold War era.

Fourth, US must prepare simultaneously for new ideas on deterrence and the possibility of its failure or irrelevance.

[D]esperate leaders driven by an internal or external imperative may distort reality in a self-serving fashion; they may be inattentive, drugged, foolish, or simply so cost/risk tolerant in pursuit of an absolute goal that U.S. deterrence policy is impracticable. The notion that U.S. nuclear weapons, or any particular type of threat, can “ensure” deterrence and that nuclear deterrence is “existential” is a dangerous myth.

Deterrence is inherently unreliable because challengers, including rational decision-makers, are not wholly predictable or controllable under any circumstances, and under conditions of great stress may often be beyond predictable reason and practicable control. This problem is certain to be exacerbated by Washington’s lack of familiarity with, and the unpredictability of, regional rogue powers.

WMD proliferation means we should prepare defences. They should not be ignored because they undermine “stability”. Defences may help deterrence and help if it fails. E.g Missile defence should be a major priority.

Defense against ballistic missiles, and other forms of defense that have been rejected in the past on the basis of overriding confidence in deterrence, should now be seen as potentially contributing to deterrence in some cases, and, perhaps more importantly, as wise insurance against the near-certainty that at some future point deterrence will unexpectedly fail. Continuing assertions that the United States does not need defenses because deterrence will work reliably is an extension of the Cold War deterrence tautology; it is an obscurant and high-risk proposition in the post-Cold War era.

AJP Taylor said that deterrence may work 99 times out of a hundred but the hundredth ‘produces catastrophe’. The real odds are worse than 99:1 and the dangers are worse than ever.

Few thoughts

I won’t go into much here, will wait until after covering Payne’s next book.

First, you can see in many comments over the past month how relevant this work is to our situation. A subject that has been very niche for a quarter of a century now needs huge focus.

Second, the current situation with Putin is a perfect case study for what Payne was talking about. There’s been a lot of confused commentary on Putin’s ‘rationality’.

Like with Hitler, Stalin and others Putin has been clear about many goals. His character has been clear. His behaviour has been clear. For years western leaders and the officials and media with power and influence acted as people have in the past.

They averted their eyes from the evidence, they chose to believe ideas that justified 1) avoiding otherwise hard decisions and 2) punting the consequences to the future.

Now we are dealing with consequences of system error. But very few of those now influencing debates and taking decisions are aware of this important history.

Further, Putin himself has made the same error about his enemy! It’s clear from pro-Putin Russian nationalists that they greatly misunderstood internal dynamics in UKR. They thought that all sorts of entities would fold. They thought that subversive operations would quickly triumph. They did not realise their own agents had been bribed, spoofed, captured, killed etc. Putin has had a nasty suprise.

Third, Payne’s point about making much greater efforts to understand the adversary is crucial.

In the Cold War our intelligence services made huge errors and were thoroughly and repeatedly compromised by high level moles.

The record of the CIA’s predictions and analysis was often woeful, as the ONA in the Pentagon has catalogued.

I know from my own exploration of the deep state 2019-20 that despite spending billions, we enormously neglected:

1. Our thinking about the likes of Putin and Xi.

2. Our defences against their intelligence services, IP theft etc.

3. Investments in people, ideas, machines — in that order! — regarding enemy intentions and WMD.

---

PS. AI definitions of rationality / expected utility

There is ‘rationality’ as used by historians, economists, and nuclear strategists.

There is ‘rationality’ as used by AI researchers, neuroscientists and cognitive scientists.

Intelligence is the ‘ability to achieve goals in a wide range of environments… It seems hard if not impossible to define rationality without the notion of a goal. If rationality is reasoning towards a goal, then there is no intelligence without goals’. This can be rigorously formalised mathematically — general, non-anthropocentric, objective, complete (Hutter).

The scientific way of measuring intelligence involves measuring problem solving capacity. (Schmidhuber)

Coherence among perceptions, actions, and goals is the essence of rationality. If we represent goals in terms of preference over outcomes, and conceive perception and action within the framework of decision-making under uncertainty, then the AI agent’s situation aligns squarely with the standard economic paradigm of rational choice. Thus, the AI designer’s task is to build agents that best approximate rationality given the limits of their computational resources. (Parkes, Science 2015)

Theoretical rationality. Von Neumann in his famous 1945 book on game theory (written while he was also helping the Manhattan Project and other critical wartime projects) presented an axiomatic formalisation of preferences and derived the ‘principle of maximum expected utility’ (MEU).

‘Accepting a compact and compelling set of desiderata about preference orderings implies that ideal decisions are those actions that maximize an agent’s expected utility, which is computed for each action as the average utility of the action when considering the probability of states of the world’ (Gershman & Tenenbaum).

AI, neuroscience & cognitive science have converged on a view of intelligence defined as

… computational rationality: computing with representations, algorithms, and architectures designed to approximate decisions with the highest expected utility, while taking into account the costs of computation in real-world problems in which most relevant calculations can only be approximated… Deliberation about the best action to take hinges on an an ability to make predictions about how different actions will influence likelihoods of outcomes and a consideration of the value or utilities of the outcomes.

Theoretically perfect rationality is impossible in practice. The complexity and nonlinearity of the world means even simple problems like chess are computationally intractable — you cannot calculate all possibilities and find the best possible solution.

‘Rationality’ for a human/artificial agent is therefore a search for approximations and heuristics taking into account the costs of computation/search: ‘The optimal rational action is very expensive to compute [NP-hard] so rational action must be approximated’ (Omohundro) & this requires meta-reasoning.

Interestingly, when solving problems with approximations and meta-reasoning AIs evolve ‘cognitive biases’: i.e ‘biases’ are byproducts of rational approximating heuristics under uncertainty when sampling the environment is costly.

NB. Want to kill all the Jews? Want to invade Iraq? Want to invade Ukraine? Want to turn all the atoms of the universe into paperclips?

Be careful about definitions of ‘rational’, it is not the same as reasonable or humane or civilised.

This paper — Computational rationality: A converging paradigm for intelligence in brains, minds, and machines, Science, 2015 — is a good summary of modern ideas in economics, AI, neuroscience etc about intelligence and rationality.

---

If you found this useful please share and consider a gift.