In Defence of a ‘No First Use’ Nuclear Doctrine
thedaleyreview.wordpress.com
thedaleyreview.wordpress.com
So, instead of "no first use", I would just have a global agreement never to have or use nuclear weapons for any reason, not even for defense.
Nuclear pulse propulsion should be allowed for spacecraft, however. This was outlawed years ago by the https://en.wikipedia.org/wiki/Partial_Nuclear_Test_Ban_Treat... but would be much more efficient than other currently used methods and safe if used with clean bombs in space. See: https://en.wikipedia.org/wiki/Nuclear_pulse_propulsion
Weapons by their nature will continue to escalate as long as technology improves, so if nuclear weapons continue to be allowed, time will surely bring a weapon even more destructive to counter them. Whether you believe they've brought peace or not, do you want someone to invent an even greater weapon and test it as the next "weapon of peace"? I don't.
What are you talking about? NASA is facing a plutonium-238 shortage, but that is because we ran out of our stock-pile that was produced as a by-product of bomb manufacturing. Production of plutonium-238 started again back in 2013.
The only international laws I could find regarding nuclear spacecraft seem to be reasonable safety requirement (eg, not crash into Earth while still radioactive).
>So, instead of "no first use", I would just have a global agreement never to have or use nuclear weapons for any reason, even for a defensive strike.
The problem is that this policy removes the main dynamic that has prevented use of nuclear weapons: the threat of reprisal. If a country (verifiably) agrees to never use nuclear weapons, then a bad actor country can safely break the agreement. Under the current system this does not happen, because that would lead to a nuclear counter attack.
But we did actually build prototype engines of a similar principle. There were actual mechanical artifacts test-fired. This is a much more viable looking option then Orion which only proved the pusher-plate design seems viable on a small scale, and which is unlikely to ever clear the problem of "build a lot of nuclear weapons...then deliberately start setting them off".
And how would that be enforced?
At the same time, keeping the policy ambiguous has a cost, namely that the other side will feel more threatened, raising the risk of an accidental nuclear war. (C.f. Able Archer 83).
Ruling out first use is the option that gives the lowest risk of nuclear war. To give the maximal stability, everyone should adopt as restrictive a policy as possible (the ideal example is China, who rule out both first use and use against non-nuclear countries). Such a policy makes the weapons less "useful", in the sense that they only used to deter a nuclear attack and not for "getting away with" weaker conventional forces, but it is safer.
During Operation Desert Storm there was consideration of using nuclear weapons to demolish very-deeply-set bunkers until it was determined that they weren't built under rock sheets. Although technically excellent thanks to Yugoslav and East German expertise, they had been dug-and-covered rather than drilled-out.
A GBU-28 conventional penetrator can punch through about 20 feet of rock, whereas a shallow-detonation 400kT B61-11 will crack structures several hundred feet deeper through shock effects.
The B61-11 remains in the inventory and has now been supplemented by the penetrating B61-12. I wouldn't rule-out first use of either type in the face of suitably challenging targets.
Under MAD an attack on the nuclear capabilities of a nuclear armed state or any attempt at unilateral disarmament is usually defined as a nuclear attack/first strike. This often interpreted both indirect and indirect means which is why the ABM work the US is been doing over the years in Europe causes such a big mess with the Russians.
The US spends trillion on missile defense, trillions that Russia cannot afford, so Russia is forced to spend it on improving their nuclear delivery capabilities. Russia restarting their investment in their nuclear arms made the US go and say "see Europe, we weren't just fear mongering, they are rearming you need our missile defense" which pushes the Russians even further.
Could a conventional attack on a power plant and the fallout from it be considered a nuclear attack? What about a nuclear powered air craft carrier?
Is an attack on a nuclear armed submarine, would they be considered an attack on their nuclear capabilities? The attacker might not even know what type of submarine it is.
Like everything in life you have to take it into proportion, if Russia or the US sink a single SSBN that's not exactly nuclear disarmament, if one party launches a preemptive non-nuclear strike against the SSBN fleet at large as well as the fixed and mobile ground based platforms that would likely invoke a very harsh response which traditionally under MAD has to be nuclear because if you do not use your nuclear arsenal while you still have it effectively you've lost it for good.
And if the other party knows or suspects you would not use your nuclear arsenal unless you've been attacked with nuclear or non-conventional non-nuclear weapons first even if the conventional attack is against your nuclear capability they would attempt to disarm you without thinking twice.
MAD works well when you under no circumstances would use your nuclear arsenal first in an offensive manner, but have very clear and public guidelines for nuclear response.
I don't believe this is true. You lower the risk that an enemy will misinterpret something as a nuclear attack, but you raise the risk of getting involved in a conventional war that goes nuclear.
Probability of enemy nuclear first strike decreased
Probability of enemy conventional first strike increased
Probability of enemy nuclear escalation ?
I am only addressing policy effects on enemies because the policy may be an intentional lie.
This could also be a potential boost for military contractors since because now we need a stronger conventional military so more, tanks, bombers, drones, etc. If anyone asks why they can say "because our doctrine changed, so a bigger military means more peace".
Also, only an insane attacker would test the limits of "no first use" by invading a nuclear state's sovereign territory. If there's any time to renege on such a policy, that's it. Not to mention: if an enemy invades, your nukes are at risk of falling into their hands.
The way I see it, the only rational thing to do once you have been hit hard by nuclear weapons is to forget your original plan of retaliation and to surrender immediately. Retaliating would only kill more people and potentially render the planet uninhabitable, but would give you no gain.
MAD only works if you tie your own hands. If you make the retaliation absolutely automatic and unstoppable. As far as I know, the only such system that was built was the Soviet "Perimetr" or "Dead Hand" system, which was never completely activated. As long as there is a human component, you will be able to, and rationally have to, turn around once you have taken the hit.
I think the naive story about MAD comes from applying game theory but assuming your strategy is static throughout the whole "game".
Likewise, the main reason we (US, NATO) have not used another nuclear bomb is that we have not been in a "real" war against an opponent where it would make sense. I mean for example an all out land war against Russia on Russian or European territory and/or where US territory is under danger. Where it would make sense to use a nuclear bomb to quickly destroy a huge tank unit or weapons factories, or to bomb the enemy into submission (if you have given up on sparing civilian lives). Once we would be stuck in such a horrible war, a "no first use" doctrine like any peacetime strategy would be up for discussion immediately, IMO.
And to the MAD scenario, I too came to the conclusion that the only sane thing to do if I was someone in the chain of command to launch nukes was
a) Always make your unwavering commitment to launch clear - to everyone, including at home.
b) When it happens, refuse to launch, contradictory to anything people know about you.
You can't tell b) to anyone though, because there is a chance the enemy may find out. It's important to only reveal that you won't ever launch nukes when it actually happens, not before.
I always found it mind-boggling that the Soviets kept "Perimeter" secret, but the theory you write about makes a lot of sense.
[1] there's so much non-information about what would actually happen that the uncertainty is the strongest signal in the channel.
The point of MAD is the "logic" ( that actually inspired Heller to write "Catch-22" ) which is:
Nuclear weapons are no good if you use them. They're only any good if you don't use them.
Played properly, not only is there no nuclear war, there is no conventional large scale war. This happened. We still ran one of the warm spots in the Cold War at the German/Russian border. And nothing actually ever happened.
We could only have stupid and in cases, possibly ruinous proxy wars in Vietnam and Afghanistan (1980 version). That's (globally) better than WWII, conventional or nuclear. Those were not psychologically sustainable. So they ended.
I attribute this to MAD. Any technology which allows us to use one simple koan of bullsh8t to keep from killing each other is a very useful technology.
Care to elaborate?
- If attacked by nuclear weapons, retaliate by completely destroying the enemy with nuclear weapons
- Do everything you can to preserve second-strike ability
Then it follows it is rational to not use nuclear weapons.
So far so good and true, but there is one problem with this argument: if your enemy does use nuclear weapons on you for any reason, is it then still rational to stick to your strategy?
Before the war, you do anything you can to cement your retaliation strategy, to preserve deterrence, MAD, and ultimately peace.
But after the war starts, and once you realize you cannot win anymore, it does not make sense anymore to go through with the second strike. You have nothing to gain. You don't punish your enemy so he doesn't attack you again "next time". If your preparations in cementing your strategy haven't been 100% thorough, and you have a way to stop retaliation, then the logical thing would be to do so and surrender unconditionally (assuming you value preserving life over revenge, when there is nothing to gain militarily anymore).
What I meant with "naive" is that the standard story about MAD assumes a steady state - both sides commit to a strategy, there is no time development, from this it follows that you should not use nuclear weapons ("the only winning move is not to play"). But if you allow the strategy to be a function of time, players can change their mind, and this changes the whole situation.
You no longer have to ask, "what is rational before playing the game" (=> not playing), but "what is rational at every time t >= 0". I believe then it is no longer about being able to second-strike, but about being able to abort the second-strike. If your enemy is in any way able to abort the second-strike (and you are already in an escalated global war, and don't value sparing enemy lives over winning), then the most rational step at t > 0 becomes to nuke the enemy as soon as possible, so they can no longer achieve their goals. If they are rational, they should immediately proceed to surrender.
I believe this "alternative story" is somewhat justified, as at least it was believed at some point by certain decision makers in both west and east militaries - I say that because there were many different nuclear strategies floating around during the cold war, and some were very different than the "Dr. Strangelove" scenario. There were ideas to fight a conventional tank war supported by nukes for example. Another rather spectacular one was (declassified one or two years ago I believe), if the soviets were to declare war and to occupy Berlin, to overkill Berlin with nuclear weapons (-> they can no longer achieve their goal, and you'd take out a significant portion of their tanks)... but of course we were never anywhere near a level of escalation where such a crazy, inhuman plan would have been considered.
Yup. If you're the kind of agent that wouldn't stick to your strategy of retaliation, then that changes the best strategies for other players in a way that harms you.
Interestingly, there's a significant amount of "removing the steering wheel" going on in terms of nuclear strategy. I remember reading that British nuclear submarine commanders would have a sealed envelope, to be opened and read only in case nuclear war destroys the British Government. It contains one of four orders - retaliate with nuclear weapons, do not retaliate, try to place yourself under American command, or to use your own judgement.
The very interesting part is that it convinces others that nuking London might automatically mean that they get hit back, but does not obligate Britain to actually making nuclear war worse.
https://www.edge.org/conversation/william_h_press-freeman_dy...
https://en.wikipedia.org/wiki/John_von_Neumann#Mutual_assure...
[edit]
As a note, I can't find a date for Von Neumann coining the MAD term. It seems likely to me that it would predate Nash's work, though as I said earlier, the existence of such equilibria was already well known (It was jut not proven that for large classes of non-zero-sum games there must exist at least one such equilibrium, which was Nash's contribution).
Uhhh, did the author forget about 1945?
Not saying it's correct to try to sweep the WWII bombings under the rug that way, but it is a thing I've seen people try to do.
"Firstly, it is important to recall that in the entire span since the United States acquired nuclear weapons, outside the two instances in Nagasaki and Hiroshima, it has not seen fit to use them."
And others have also set off the bomb: https://en.wikipedia.org/wiki/List_of_nuclear_weapons_tests
Granted, weapons testing is a little different than intentional use in war against an enemy.
https://en.m.wikipedia.org/wiki/Nuclear_testing_at_Bikini_At...
Not that I'd vote for either of them, but it bothers me when people claim that a Trump presidency would be any more dangerous in this regard then if Clinton were elected.
Only the worst kind of human would ever actually use a nuclear weapon in any scenario.