Jeffrey LewisDoctrine for Joint Nuclear Operations (Joint Publication 3-12)

*[DTIC subsequently removed these documents. Here are the four documents in a single, zipped file.]*

The Joint Staff hates freedom.

What else should I conclude when its staff knowingly place classified material they believe could be harmful to national security on a public webserver?

Through that noted hacking program “Google,” I discovered that the Joint Staff keeps draft documents and comments on a public webserver located at

I decided to download (and compare) multiple copies of the draft Doctrine for Joint Nuclear Operations (Joint Publication 3-12) from September 2003 and March 2005, as well as first and second rounds of comments from the services and a contractor.

The US Army helpfully warned that “security classification of this paragraph content [regarding High Altitude Nuclear Detonations] as written appears to expose a critical vulnerability of US systems and is typically only found in classified documents” and Joint Staff complained that a description of US “inactive stockpile” was “inappropriate in an unclassified context.”

Now, I am just taking the piss out of the Joint Staff. There aren’t any real secrets being revealed. The term “inactive stockpile” has been used by DOD offficials in an unclassified context, and DOD concern about High Altittude Nuclear Detonations is well documented.

Some Stuff I Didn’t Know (Or Was Surprised To See In Print)

Still, the documents contain some helpful some revelations:

First, the new Joint Publication will contain a list of nuclear weapons related National Security Presidential Directives, including NSPD’s 34 and 35 which have not previously been disclosed:

  • NSPD 14: Nuclear Weapons Planning Guidance (June 2002).
  • NSPD 17: National Strategy to Combat Weapons of Mass Destruction (December 2002). Although the commentator notes that “the subject of NSPD-17 is classified and will therefore not be reflected in this document,” the classified Appendix is described in detail in stories in the Washington Post and Washington Times.
  • NSPD 28: United States Nuclear Weapons Command and Control, Safety, and Security (June 2003), which “provides direction on various nuclear issues, to include security.”
  • NSPD 34: Fiscal Year 2004-2012 Nuclear Weapons Stockpile Plan (May 2004), which “directs a force structure through 2012.”
  • NSPD 35: Nuclear Weapons Deployment Authorization (May 2004).

Second, the defense community clearly holds the 2002 Treaty Between the United States of America and the Russian Federation On Strategic Offensive Reductions (Moscow Treaty) in contempt. An early draft of the joint pub described the Moscow Treaty as “a new framework for accounting and compliance,” which drew this withering assessment from the Navy:

The [Moscow Treaty] has no requirement for accounting and compliance. It is not a binding agreement like START. Further, the Moscow Treaty does not have any compliance-related language in it. In fact, the 1700-2200 limit only applies for one day 31 December 2012.

The Moscow Treaty’s use of the term “operationally deployed” is treated as a polite fiction. When the term was used in reference to Russian forces, a Joint Staff commentator noted that “The term ‘operational deployed’ is not relevant to the Russians and is not part of the Treaty. The term operationally deployed refers to the method the US will use to reach their Treaty limit.”

The idea that “operational” forces are simply an accounting device to describe the readiness of forces is reinforced by the way the document introduces the Nuclear Posture Review’s concept of an “augmentation capability”:

augmentation capability (nuclear). The inventory of US strategic nuclear warheads that are not operationally deployed and that could serve to augment the deployed forces should the US strategic nuclear force requirements rise above the level of the Moscow Treaty. In a developing crisis, the augmentation capability may be required to increase the number of operationally deployed strategic nuclear warheads above the limits of the Moscow Treaty. Such a change to the US operational nuclear force level could only be considered following a US withdrawal from the Moscow Treaty and appropriate action by the President and the Congress.

That sentence referring to “a developing crisis” originally called for “remaining nonalert nuclear assets quickly integrate to favorably alter the strategic situation.”

Yeah, that’s stabilizing.

Finally, the Air Force recommended deleting langauge relating to STRATCOM’s resposibility to “prepare Whiskey messages” and “draft Romeo messages” in favor of the less evocative “prepare Emergency Action Messages.”

Use and Abuse of Classification

In addition to the substantive insights provided by the documents, the drafting process reveals the uses and abuses of classification. In one blunt paragraph, the Joint Staff insists on using a phrase provided by STRATCOM—technical course of action—“as an obfuscation for technical IO [information operations].”

The Navy suggested that, due to classification, the publication contained “very little of substance.” The Joint Staff fired back that

I’m sure we all have attempted to add value to substance of the pub – however many things remain under development in classified fora, like the integration discussion.

The “integration discussion” refers to the integration of nuclear and conventional forces—an important public policy question that has been shielded from public scrutiny by classification. Similarly, the Army’s suggestion that the publication include “some discussion of the policy concept of preemption as well as an expanded discussion of retaliation” consistent with “current national security strategy” was rejected by the Joint Staff as “not appropriate for an unclassified pub[lication].”

This is really obscene—the Administration asserts that it has reduced reliance on nuclear weapons, but the relevant discussions related to the integration of nuclear and conventional forces or the role of the pre-emption remain classified.


  1. Alex Montgomery (History)

    I sincerely hope that you downloaded everything on that site. These goofs do occasionally get discovered.

    Also interesting are some of the revisions made, e.g. changing “rogue” states to “regional” states (whatever the latter might mean). As well as the deletion of the phrase mentioning that changes to the operational nuclear force level could only be considered after abrogation… see 3_12fc2.pdf, p.II-13.

  2. marko (History)

    Brilliant Detective Work. I did a masters in US nuke policy during the Clinton era and this collection and analysis here is the best I have seen in regard to Bush policy…didn’t know about NSPD 14, will be interesting to learn the differences with PDD 60.

  3. Earl Kirkman (History)

    You vile hacker, yu nd your kind are whuts rong wif this kuntry today…

    But seriously dear, how can you expect anything else?? The nukes are the vballs of the contemporary mil establishment. We just can’t have any restriction of that phallic uber-statehood exist, it would make us speak in high soprano. I despair of the mind-set that keeps the earth-burn capability above peace.

  4. J. (History)

    I used to work on the Joint Staff. That comment matrix brought a smile to my face (and a grimace – 153 pages!!!). It is an art to be able to take a joint document and arbitrate all the comments from OSD, the Joint Staff directorates, four Services, nine combatant commands, and other defense agencies. You have to expect a little fire and shots back and forth in the process of arbitration.

    This will be an interesting document when it gets published. You’ve shown some of the debate behind the scenes which is always interesting (to people like me). While I agre that some policy on nuke use would have been of value, I can understand why it was rejected – might reveal too much on CONPLAN development. If the NSPD is classified, then any talk about policy is classified. Maybe they could have dumbed it down, but (given the political climate and hardball attitude of the Bush administration), maybe that wasn’t a good way to go.