3 ms·

Somewhat related, I've never believed the conventional doctrine of mutual assured destruction. The way I see it, the only rational thing to do once you have be

Somewhat related, I've never believed the conventional doctrine of mutual assured destruction. 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.

deepfriedbits 10y ago
Nukes were strongly considered during the Korean War due to the massive number of Chinese fighting on the side of the aggressors in that war. MacArthur wanted badly to nuke China or more specifically, Chinese divisions and the threat of nukes helped lure China to the negotiating table when the U.S. wanted to wind down the conflict.
Noseshine 10y ago
Incidentally, I read that "perimeter" - the Soviet automatic MAD system - was actually not meant against the Americans, but against the "hawks" in the Soviet government. Which is why it wasn't publicized. It assured them that the USSR had a way to retaliate in any case - so there was no need to launch nukes right away at the slightest suspicion, but waiting if there actually was an impact somewhere was possible without losing the ability to retaliate. With the "hawks" assured the calmer people now had more time to think, talk, and check data in an emergency and faced less pressure to act immediately. 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.
- captainmuon 10y ago
  Yeah, the most important thing about deterrence is to make a certain impression to the enemy. I always found it mind-boggling that the Soviets kept "Perimeter" secret, but the theory you write about makes a lot of sense.
ArkyBeagle 10y ago
I think you'd have to follow through in case it didn't turn out that the world was destroyed in the first strike[1] and other nuclear capable actors remained. [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.
netheril96 10y ago
> I think the naive story about MAD comes from applying game theory but assuming your strategy is static throughout the whole "game". Care to elaborate?
- captainmuon 10y ago
  Here I'm using "game theory" in the pop-sci sense, no deeper mathematics. Assume you and your opponent commit to the following strategy: - 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.

deepfriedbits 10y ago

Nukes were strongly considered during the Korean War due to the massive number of Chinese fighting on the side of the aggressors in that war. MacArthur wanted badly to nuke China or more specifically, Chinese divisions and the threat of nukes helped lure China to the negotiating table when the U.S. wanted to wind down the conflict.

Noseshine 10y ago

Incidentally, I read that "perimeter" - the Soviet automatic MAD system - was actually not meant against the Americans, but against the "hawks" in the Soviet government. Which is why it wasn't publicized. It assured them that the USSR had a way to retaliate in any case - so there was no need to launch nukes right away at the slightest suspicion, but waiting if there actually was an impact somewhere was possible without losing the ability to retaliate. With the "hawks" assured the calmer people now had more time to think, talk, and check data in an emergency and faced less pressure to act immediately. 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.

captainmuon 10y ago
Yeah, the most important thing about deterrence is to make a certain impression to the enemy. I always found it mind-boggling that the Soviets kept "Perimeter" secret, but the theory you write about makes a lot of sense.

ArkyBeagle 10y ago

I think you'd have to follow through in case it didn't turn out that the world was destroyed in the first strike[1] and other nuclear capable actors remained. [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.

netheril96 10y ago

> I think the naive story about MAD comes from applying game theory but assuming your strategy is static throughout the whole "game". Care to elaborate?

captainmuon 10y ago
Here I'm using "game theory" in the pop-sci sense, no deeper mathematics. Assume you and your opponent commit to the following strategy: - 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.