Foreign Relations of the United States, 1989–1992, Volume XXXI, START I, 1989–1991

159. Memorandum From Richard Davis of the National Security Council Staff to the President’s Assistant for National Security Affairs (Scowcroft)¹

Washington, September 28, 1990

SUBJECT

• Definitions of “RV” and “Simulated Release” in START

Two of the remaining blanks in our START position are definitions for “RV” and “simulated releases” of RVs during flight testing. Both terms are important for the treaty, but have defied attempts to find accurate definitions that do not leave open obvious circumvention paths or capture potential U.S. capabilities (e.g. penaids) that would not otherwise be subject to START limits. In neither case do we need a definition to make the Treaty work. In both cases the issues are whether specific definitions will help or hurt us politically during ratification and whether they will clarify or exacerbate compliance disputes later.

RV Definition

We and the Soviets have agreed to include a definition of RV in START, but neither side has proposed specific language. The concept of RV is central to making the linkage between flight-testing observations of RVs and treaty accountability of warheads under the 6000/4900/1540 limits.

It will seem odd to the Senate not to have a definition of a central term in START when over 130 other terms, many of them minor, are defined. (In INF at the Senate’s urging we went back to the Soviets during ratification to get a common understanding of the term “weapon.” We have incorporated the substance of that definition into START: any mechanism or device that, when directed against a target, is designed to damage or destroy it.)

On the other hand, it is not clear that we need to have a definition of RV. In SALT there has never been a compliance question about whether or not an object should be treated as a RV; we know a RV when we see one. Moreover, most potential definitions could have unintended negative affects on U.S. options. A definition that included a weight threshold could count certain penaids as RVs or exclude future light-weight RVs if they fell under the threshold. A definition that attempted to exclude non-nuclear defense suppression weapons would be difficult to negotiate.

There are three options:

1. Do not propose a definition. DOE supports; OSD and JCS prefer.

2. Propose the following definition:

The term “reentry vehicle” means an object carried on an ICBM or SLBM that is capable of surviving reentry through the Earth’s atmosphere and is also capable of containing a weapon.

OSD, JCS and State can accept.

3. Modify the language in option 2 to include only those objects “capable of containing a nuclear weapon.” ACDA supports; State prefers.

Although OSD and JCS prefer that RV remain undefined, they believe that a definition will be necessary politically and can accept option 2. They note that option 2 would capture [less than 3 lines not declassified] They see nothing in the definition that the Soviets could use to advantage. State agrees with this outcome as long as the phrase “capable of containing” is part of the definition.

ACDA supports option 3 on the assumption that the U.S. needs to protect the option of not counting a non-nuclear defense suppression weapon (DOD maintains that it does not need to protect this option). [4 lines not declassified]

DOE favors option 1, arguing that any definition would create more problems than it would solve.

NSC staff favor option 2. The desirability of including a definition of RV is primarily political. Our overriding objectives in the specifics of the definition are to avoid language that the Soviets could exploit to circumvent the Treaty [1 line not declassified] Options 2 and 3 satisfy both of these criteria. Option 3 is weaker than option 2 because it would require us to monitor the distinction between objects capable of containing a nuclear weapon from those capable of containing only a non-nuclear weapon. Further, option 3 would be more difficult to negotiate because it would raise again the neuralgic issue of whether START should limit all strategic weapons, or only nuclear weapons.

Finally, we note that OSD and JCS agree that option 2 is acceptable [less than 1 line not declassified]

Simulated Release Definition

The U.S. and Soviet positions on RV counting differ in the details of how to assign a number of RVs to a specific type of missile. Both would count the number of actual RVs released during flight-testing. Both would also attempt to count simulations of RV releases, but would do so in different ways. The U.S. position would simply count the number of “simulated releases” which is not further defined in the treaty. The Soviets have proposed the phrase “procedures for releasing and dispensing,” which they define as the maneuvers of a missile associated with targeting or dispensing RVs.

Technical experts in the USG have studied this issue extensively and have been unable to identify a definition of simulated release that would both: 1) capture all Soviet test practices that would give the Soviets confidence that additional RVs could be carried and released from a particular PBV, and 2) [2 lines not declassified] Under SALT the Soviets have never raised questions about U.S. practices. [2 lines not declassified]

Given the importance in START of accurately attributing RVs to missiles, the Senate will likely examine the definition—or absence of definition—of the term “simulated releases.” Although “simulated releases” is a technical detail, not a central concept in the treaty as is “RV,” we would have to be prepared during ratification to defend the utility of any definition we negotiate. Either a weak definition or an absence of a definition could attract criticism.

There are 2 options:

1. Accept the Soviet language and elaborate on it to ensure that 1) U.S. PBV maneuvers are not captured and 2) any mass at least 80% of the weight of the lightest RV is counted as a RV.

2. Do not propose a definition.

ACDA favors option 1, arguing that clarity and precision in the definition will clarify U.S. undertakings in the treaty and will minimize future compliance questions.

OSD, JCS, DOE, and the DCI’s Rep oppose any definition of simulated release. They argue that a specific definition could stimulate Soviet questions about specific U.S. practices that are unrelated to RV releases, and that no realistic definition could close off plausible Soviet circumvention opportunities.

State believes the term “simulated release” should be defined but has not identified a definition that is acceptable.

NSC staff agree that a substantively useful definition has not been found. In the end, we may have to compromise with the Soviet definition, but no definition should be our preferred outcome. In an inherently gray area, this gives us the necessary latitude in our own programs and maximizes our flexibility to raise questions with the Soviets. We especially want to avoid a situation in which we sanction questionable Soviet practices by excluding them from the definition. [less than 3 lines not declassified] Given the technical nature of the term “simulated release,” these arguments should be helpful in explaining to the Senate during ratification why we chose not to negotiate a definition.

RECOMMENDATION

That you approve the Sittmann memorandum to counterparts at Tab I.² This memorandum reaffirms the current U.S. position not to define “simulated release” and approves option 2 for the definition of RV:

The term “reentry vehicle” means an object carried on an ICBM or SLBM that is capable of surviving reentry through the Earth’s atmosphere and is also capable of containing a weapon.³

Condi Rice and Philip Zelikow concur.

1. Source: George H.W. Bush Library, Bush Presidential Records, Brent Scowcroft Collection, Scowcroft Close Hold Files, NSC Original Files—IF—Brent Scowcroft Folder #2 [9020666–9021348]. Secret. Sent for action. Sent through Kanter. A stamped notation at the top of the memorandum reads: “Signed.”
2. Attached but not printed. In an October 4 memorandum to Vice President’s Assistant for National Security Affairs Carnes Lord, Department of State Executive Secretary J. Stapleton Roy, Department of Defense Executive Secretary John A. Dubia, Department of Energy Director of the Executive Secretariat Ann Lavin, Office of Management and Budget Associate Director for National Security and International Affairs Robert Howard, Central Intelligence Agency Acting Executive Secretary James Pittman, Joint Chiefs of Staff Administrative Assistant to the Chairman Colonel George Sumrall, and Arms Control and Disarmament Agency Executive Secretary Barbara Starr, Sittman wrote: “The following decisions have been made with respect to the U.S. proposal in START 1. The U.S. will continue to propose that the term ‘simulated release,’ as it applies to the testing of ballistic missile reentry vehicles, not be further defined. 2. The U.S. should propose the following definition of RV: ‘The term “reentry vehicle” means an object carried on an ICBM or SLBM that is capable of surviving reentry through the Earth’s atmosphere and is also capable of containing a weapon.’”
3. Scowcroft initialed his approval.