Jeffrey LewisChina and No First Use

I am sure you have, by now, noticed that Kyodo News has published a story titled, “China military eyes preemptive nuclear attack in event of crisis,” which was picked up by the Associated Press and AFP.  The article claims that an “internal document” from the PLA indicates that under certain conditions, China would abandon its no first use pledge.

The internal document in question appears to be the Science of Second Artillery Campaigns (第二炮兵战役学), a 2004 a textbook printed by the Chinese National Defense University and used for training military officers.   This textbook is one of a series of textbooks circulating in the United States including the better known Science of Military Campaigns. (I picked up copies of several of these books in Beijing, but was stopped by a conscientious clerk who noted the ones marked “nei bu” — not for foreigners. Oh well!)

Basically this is an well-known textbook that is widely circulated and cited, which has been taken out of context and then further abused by the translation into English. The Chinese military is not eyeing a preemptive nuclear attack in a crisis, but the slightly longer explanation is interesting.

Like most news stories, the headline does not accurately describe what is written, which is more careful.  (Apparently, the writing is even more careful in the original Japanese.) Rather than a general sense that China might use nuclear weapons preemptively in a crisis, the story indicates that China might abandon its “no first use” policy (a warning, followed by a strike) in response to a conventional attack on Chinese civilian targets like nuclear power plants, hydroelectric facilities and urban centers.

(The Chinese Foreign Ministry denied the report, but in that utterly unconvincing manner of apparatchiks and functionaries the world over.)

As I noted, this is one of a series of textbooks circulating in the United States. The Science of Second Artillery Campaigns has been cited in several scholarly publications, including articles by Andrew Erickson, as well as Taylor Fravel and Evan Medeiros.  The 2004 document should be interpreted in the context of the entire body of textbooks, including the two volumes of the Science of Campaigns (战役学, 2000, 2006), which suggest China’s no first use policy remains an operational constraint.  China trains, equips, and postures its forces on the expectation that Chinese leaders will ride out a nuclear attack.

The narrow case of what China would do in the event of a conventional attack on civilian infrastructure and population centers deserves perhaps some additional consideration.  A well-established problem with “no first use” and other categorical statements about nuclear weapons employment, as I have argued before,  is that it is always easy to imagine implausible but not impossible scenarios, akin to the tendentious and artificial “ticking time bomb” hypotheticals used to justify torture.  This is one reason I have suggested we talk about why we possess or maintain nuclear weapons, but not when we would use them.

(As an aside: In our Track II discussion, US experts sometimes try to illustrate this problem, asking what would happen if the US used conventional weapons in some particularly provocative manner, in order to demonstrate the American aversion to “no first use” and other categorical declaratory policies. We should stop doing this.  The Chinese, as you will soon see why, think this is intended as a threat.)

In the early 2000s, some Taiwanese (and American) strategists began discussing the use of conventional weapons against high-value civilian targets.  The 2004 edition of Chinese Military Power, for example, noted that Taiwanese “proponents of [conventional] strikes against the mainland apparently hope that merely presenting credible threats to China’s urban population or high-value targets, such as the Three Gorges Dam, will deter Chinese military coercion.”

As you might imagine, the Chinese were enraged at this suggestion of striking the Three Gorges Dam (which would inundate many Chinese cities and drown a lot of Chinese citizens).  In response to the report, the PLA trotted out Lt. Gen. Liu Yuan, son of Liu Shaoqi, to write a very tough response in the China Youth Daily calling leaders who attacked civilians in this way “whores,” arguing that this would be worse than anything Osama  Bin Laden has ever done and generally leaving the impression that a strike on the Three Gorges Dam would be a very bad idea.  He did, however, avoid threatening to use nuclear weapons.

It appears that the discussion of early 2000s discussion of conventional strikes against Chinese civilian targets prompted a debate within the PLA that amounted to, well, what would you do in that case? This was the context for the (in)famous remark by Gen. Zhu Chenghu that “If the Americans draw their missiles and position-guided ammunition on to the target zone on China’s territory, I think we will have to respond with nuclear weapons.”

The 2004 textbook was written in this same era, when Chinese leaders were struggling with the question of how to reconcile “no first use” with the possibility of mass casualty conventional attacks.  References to striking civilians targets clearly initiated a discussion about conventional strike and the “no first use” policy.  There is some question about how to characterize this conversation — is it a debate about whether to discard or condition the pledge? Or ought we understand the dialogue as part of the continuous and often difficult process of developing plausible operational concepts within the strictures of the seemingly immutable policy on no first use.  (I happen to the think the second explanation is much more intriguing.)

Now, six years later, Kyodo has a copy of a textbook from this period.  From a handful of discussions, my sense is that Kyodo’s coverage of the textbook (which I still need to acquire) is similar to the public reaction to a 2005 comment by Chu Shulong, a Chinese scholar, that was shorn of its context and distorted by being transplanted from a domestic debate to a foreign one.

In 2005, Chu gave an interview to a Chinese newspaper, titled in Chinese “PRC Expert: China’s Policy on Nuclear Weapons Remains Unchanged.” FBIS, however, picked up the interview and titled the document “PRC Expert Warns PRC May Renounce ‘No First Use’ of Nuclear Weapons in War Time.”  Quite a difference!

How did that happen? As I blogged at the time, despite Chu’s direct argument that China would continue to maintain “no first use” as an operational assumption (and subsequent statements to that effect), Chu also admitted the obvious: No one could be certain what Chinese leaders might do under extreme scenarios. This, too, ended up in Chinese Military Power.  The 2006 edition highlighted Chu’s remark as further evidence that China was moving away from “no first use,” despite Chu’s obvious intention to make the point that it would remain the policy for the foreseeable future.

This is a textbook that has circulated, if not widely, at least among scholars for some time now and was little remarked upon by experts because in context it is broadly consistent with what we already knew: China’s no first use policy probably remains a real operational constraint on how China trains, equips and postures its nuclear forces.  New conventional capabilities may be a new source of pressure on the policy.  And, if the US were to do something like destroy the Three Gorges Dam and drown 20 million Chinese, no one could be confident that an enraged Chinese leader in a hot-blooded moment would scrupulously observe “no first use”.  Which is how the Chinese like it.


  1. Coyote (History)


    Fascinating stuff. I think it is important for states to garner prestige among the international community during peacetime by announcing a “no first use” policy. However, it is even more important for states to abandon such a policy if their survival is at stake. It’s a safe assumption that they will. I think our Chinese partners are prudent for wanting it both ways.

    The best policy of all is to never back a nuclear-armed country into a corner. The trick, of course, is knowing where the red lines are. Declaratory policies can help, but this is not something in which our Chinese partners have demonstrated proficiency. To be fair, we all need to do better at this.

    I wonder how many countries are rushing to proliferate so they can have a “no first use” policy?



  2. Captain Ned (History)

    I got lost in the glorious sound of Merlin engines.

  3. 3.1415 (History)

    The fondness of China for NFU and the aversion of the West for it reflect our cultural differences. The fundamental assumption for most Chinese is that people are born kind (人之初,性本善。). This has been indoctrinated in the Chinese meme for thousands of years. This assumption worked generally fine when there was no foreign invasion. Because Chinese people really like to restore the validity of this assumption, it becomes important to minimize the probability of foreign invasions. Nuke is one of the means and arguably not that effective. Her Majesty has her own nukes and many more in Uncle Sam’s hands. That did not deter Argentina from going to war over a little island with the United Kingdom. Would Buenos Aires be nuked if Her Majesty’s aircraft carrier was sunk? As for China, the more certain deterrence is the symbiosis with the United States. What would happen to the US if there were no China at the next treasury auction? That would trigger the mother of all bank runs. The nuke business is based on an assumption that has been shown to be flawed, if not totally wrong, by people like Steven Pinker. We as a species are getting less and less violent on a log scale plotted against a historical timeline. It has been shown ad nauseam that when someone is more useful to you when alive then dead, you have no incentive to kill that person. NFU is just a declaration; the guarantee is the law discovered by Adam Smith that we can have a non-zero sum game called trade.

  4. Andrew Tubbiolo (History)

    Let me be the fly in the soup say this. If you own a nuclear arsenal isn’t it a smart thing to have a first use policy? Don’t you want that policy out there in the open for all to see, and debate it? At some point any nuclear power will use first if put into a corner. I’d think that any nuclear power would at least want the other nuclear powers just how far not to push things. A public no first use policy makes sense when you are at the levels of China and you have no hope of executing a strategic first strike against the larger nuclear arsenals and there’s no real link of such an action to the ongoing politics. Or if you were the USSR and you think you have overwhelming conventional capability in Europe.

    There’s going to be a conventional standoff against China. There will be a nuclear dimension to this. As China ascends and we decline, we’re going to have to figure out how we will posture. I think living with the fact that China will have a integrated nuclear policy with their conventional stance, included with that will be a first use policy either known to us or unknown to us will be a matter of a lot of future politics.

    If the arms control community is able to prevent this future stand off. Hat’s off and more power to you. I want to see you all succeed, but I’m not going to hold my breath.

    • John Schilling (History)

      I’m not sure you do want the policy to be out there in the open for all to see. If your actual policy is, “we will go nuclear if A happens, and we might if B or C happens”, you probably want potential adversaries to believe your policy is, “we will for sure go nuclear if A, B, or C happens and we will probably go nuclear if D, E, or F happens”, where D-F are things you’d rather not have happen but wouldn’t really start a nuclear war over.

      You might want allies to believe the same thing, or you might want them to believe something different. And you don’t want to be caught it a lie, if someone calls your bluff on point D. So I can easily see someone legitimately wanting their actual first-use policy to be obscure.

      And note that explicit but non-credible statements like “absolutely no first use ever”, have the effect of making your actual first-use policy quite obscure. Probably deliberately so.

    • Scott Monje (History)


      Aside from the question of calculated ambiguity, does each country really know in advance what it will do if any of A-F occurs? I think some of these plans are an exercise in self-deception. Many war plans of a less devastating nature go out the window when the manure hits the fan.

  5. FSB (History)

    The debate is academic.

    In times of high tension any adversary will likely assume that any nice prose in any declaratory policy of NFU should not be trusted.

    Of course, the US used first even when its survival was not at stake.

    To Coyote: yes, many nations are rushing to proliferate so that they may have a deterrent, just like the US.

    What’s good for the goose is good for the gander.

    • Coyote (History)


      It does make for a nice insurance policy, doesn’t it? Like an ace in the hole.



  6. rwendland (History)

    “some Taiwanese (and American) strategists began discussing the use of conventional weapons against high-value civilian targets. … such as the Three Gorges Dam”

    which would, of course, be contrary Article 56 of the Protocol I 1977 amendment to the Geneva Conventions. (As would an attack against a “nuclear electrical generating stations”.)

    Though I see that the U.S. is one of the tiny handful of states that is not in Ratification/Accession of this amendment, so there is no point in asking if anyone can name and shame these otherwise putative war criminals!

  7. anon (History)

    A few points about this book. First, it’s classification is way beyond “neibu”. Second, Fravel and Medeiros do not mention the 2004 version in their recent IS article. They mention an earlier version from the mid 1980. Third, perhaps the fullest public description of the ideas in the Kyodo report can be found is Chase, Erickson and Yeaw Chinese Theater and Strategic Missile Force Modernization and its Implications for the United States The Journal of Strategic Studies Vol. 32, No. 1, 67–114, February 2009. Fourth, PLA 2nd Arty writing on “lowering the nuclear threshold” and abandoning NFU is probably about developing a set of options that they can offer the political leadership should the latter decide that NFU is under stress from conventional strikes on existential targets. Fifth, more-or-less agree with Jeffery’s last paragraph, except that it is not clear whether some of this 2nd Arty thinking might influence crisis management behavior. Prior to abandoning NFU, the book, and other writing, suggests there are steps for signaling the credibility of nuclear use that involve fairly escalatory actions short of nuclear demonstration shots and the like.

  • Jeffrey (History)

    That’s extraordinarily helpful. Thank you.

  • Mark Lincoln (History)

    No First Use is great propaganda. No one is damned for saying they will not use nukes first.

    For decades in the 1950s until now American media assured Americans that ‘we would not use nukes first’ while in fact our intent was to use them first and had been since 1945.

    Is this article anything more than another construct to build a new cold war to justify not cutting US defense expenditures?

    Perhaps I am too old, but I do remember when there was serious talk about nuking China before it had nukes. . .

    • Charles (History)

      A preemptive nuke to avoid more nukes? Where have I heard this recently?

  • jeannick (History)

    George W Bush said the same thing more or less when he stated that the U.S. could go nuclear if attacked with biological weapons
    there is an element of bluster in all of this
    but it make perfect sense to threaten a range of options if faced with an existential threat , if only to dissuade some silly assumptions

  • Anon (History)

    “In response to the report, the PLA trotted out Lt. Gen. Liu Yuan, son of Liu Shaoqi, to write a very tough response in the China Youth Daily calling leaders who attacked civilians in this way “whores,” arguing that this would be worse than anything Osama Bin Laden has ever done and generally leaving the impression that a strike on the Three Gorges Dam would be a very bad idea. ”

    I suggest the NFW policy: No first whores.