A line in a briefing book
The National Security Archive at George Washington University posted an Electronic Briefing Book titled “Nuclear Weapons and the Law of War” on June 6, 2023. It collects U.S. government records from the negotiation of Additional Protocol I to the 1949 Geneva Conventions, a treaty meant to sharpen wartime protections for civilians. The posting reads as documentary history, not disclosure of wrongdoing: one compilation among the Archive’s holdings, drawn from memoranda and reviews spanning a multi-year diplomatic process. Among the papers reproduced is a memorandum from the Pentagon’s Joint Staff, dated September 1975. The available material gives no fuller title, no named author, and no classification marking, the stamp, such as Secret or Confidential, that would show how tightly the Pentagon held the document at the time.
Geneva, 1974 to 1977
The treaty took shape at the Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law Applicable in Armed Conflicts, which opened in Geneva, Switzerland, on February 20, 1974, and ran across four sessions into 1977. A photograph of the opening session, credited to the International Committee of the Red Cross, appears among the Archive’s illustrations. One strand of the talks concerned civilians directly: limits on reprisals, a proportionality rule weighing expected civilian harm against military advantage, and a specific ban, under what became Article 56, on attacking dams, dikes, and nuclear power stations where doing so would release dangerous forces onto nearby civilians. According to the posting, U.S. and NATO officials worried that language drafted with conventional war in mind might be read, intentionally or not, as reaching nuclear weapons too. A West German aide-memoire preserved in the same record raised a sharper version of that worry: such a reading, Bonn noted, could call into question the first use of nuclear weapons, an idea built into NATO’s own deterrence doctrine.
What the September 1975 memo says
Into that atmosphere, the Joint Staff put a position on paper. The memo concluded that protecting civilian populations would not be “feasible” in the event of “general nuclear war.” It registered particular concern over proposed restrictions on reprisals against civilians, which defense officials viewed, according to the record, as an essential deterrent against violations of the law of war by an adversary, and which they wanted to preserve for conventional warfare as well as nuclear. None of this is presented as a public U.S. position or a formal submission to the conference. It reads as an internal assessment, one entry in a negotiating record running across years, grappling with a scope problem: rules drafted for the battlefield meeting a nuclear doctrine built on separate assumptions. The same posting includes later reviews by NATO’s Military Committee and a Defense Department working group, both of which simply assumed the protocol’s rules did not reach nuclear weapons at all. U.S. officials held that view from the start of the negotiations and had, per the record, a tacit understanding with the Soviet Union to the same effect. George H. Aldrich, the State Department lawyer who led the U.S. delegation, later proposed clarifying statements to the same end.
What this record is not
This record should not be over-read. It is a compilation, not an investigation into an accident or a death, and there is no incident here for anyone to resolve. What surfaced on June 6, 2023, is the existence of one internal Pentagon position from September 1975, sitting inside a much longer paper trail of treaty-making. The substance is the memo and the worry it records, not a mystery waiting on a verdict.
What the record does not settle
Three gaps are worth keeping separate. First, the posting does not trace a line from the Joint Staff’s “not feasible” finding to the treaty’s final wording, or to the United States’ later decision on signature and ratification. The conference adopted Protocol I on June 8, 1977; the United States eventually signed it but never ratified it, citing general objections that some provisions exceeded customary international law, a different and broader argument than the nuclear concern recorded in 1975. Whether that 1975 finding fed into the ratification decision at all is not addressed in the material reviewed here. Second, the memo is identified only by its month and year: no full title, no named author, no distribution list, and no classification level appears in the available excerpt. Third, how negotiators ultimately handled the nuclear question inside the finished text, by silence, by deliberate ambiguity, or by a shared assumption nobody wrote down, is not described in this posting.
No aftermath supplied
The posting includes no later statement from a participant, no retrospective account, and no advocacy-group assessment of what the memo meant. That absence is a fact about this particular record, not proof that no such statements exist elsewhere; other archival material on the later years of this same policy question is not part of what is reviewed here. What remains, roughly half a century after it was written, is the documented fact of the finding itself: in September 1975, the Pentagon’s Joint Staff judged, on paper, that civilians could not be protected in a general nuclear war, in the middle of a negotiation built to protect them. What that judgment did to the outcome is not in the file.