The Gulf of Tonkin Incident

//The Mainstream Narrative

The conventional account treats August 1964 as a turning point produced by genuine confusion in the fog of war. North Vietnamese P-4 torpedo boats did attack the destroyer USS Maddox on August 2 — that engagement is documented and uncontested. Two nights later, in poor weather and heavy seas, the Maddox and USS Turner Joy reported a second attack and fired hundreds of rounds into the dark. Within hours, President Johnson informed the nation of an "unprovoked second attack" and ordered retaliatory air strikes. Three days later, Congress passed the Gulf of Tonkin Resolution by 416–0 in the House and 88–2 in the Senate (Senators Wayne Morse and Ernest Gruening dissenting).

In this framing, intelligence officers and ship crews acted in good faith on bad data, and the Johnson administration responded to what it sincerely believed had happened. The escalation that followed was, in this telling, a tragic consequence of honest error in a tense Cold War theater.

//Under-Reported Dimensions

The historical record is harder than that account suggests. In 2005 the NSA released a 200-page internal study by agency historian Robert J. Hanyok titled "Skunks, Bogies, Silent Hounds, and the Flying Fish" — completed in 2001 but kept classified for four years, partly out of concern about the parallels to Iraq WMD intelligence. Hanyok reviewed the original SIGINT cables, dispatcher logs, and intercept summaries from August 1964 and concluded the second attack never occurred. Worse, he documented that NSA analysts at the time had cherry-picked fragments of intercepts to construct an attack narrative, and that contradictory raw traffic — including intercepts that strongly suggested the August 4 cables described the August 2 engagement — was suppressed in the briefings sent up the chain.

This was not a mistake at the radar-scope level alone. The Pentagon Papers, leaked by Daniel Ellsberg in 1971 and now held in full by the National Archives, show that the Johnson administration's own internal planning treated the Tonkin Resolution as an instrument prepared in advance, not a reaction. A draft of what became the Resolution had circulated inside the executive branch months before any incident, awaiting a triggering event.

//Credible Dissenting Voices

Some historians — notably Edwin Moïse in Tonkin Gulf and the Escalation of the Vietnam War (UNC Press, 1996, revised 2019) — argue the case for deliberate executive deception is overstated. In Moïse's reading, Johnson and McNamara were genuinely uncertain on the night of August 4, leaned toward "attack probably happened" within hours, and were not handed clearly contradictory SIGINT in time to reverse course. The deception, in this view, came later: in how the administration described the certainty of the attack to Congress and the public, not in fabricating an event whole-cloth.

Other scholars push harder. Daniel Ellsberg, who worked inside the Pentagon at the time, has argued the Tonkin episode was a contingency long planned for and that the SIGINT manipulation went higher than the NSA. McNamara himself, in his 1995 memoir In Retrospect, acknowledged the August 4 attack "probably didn't happen" but stopped short of accepting that the administration had known so at the time.

//Follow the Money

The Resolution mattered because of what it enabled. Without a declaration of war, it became the legal scaffolding for nine more years of escalation — including the deployment of more than half a million U.S. troops, sustained bombing of North Vietnam, and tens of thousands of American and millions of Vietnamese deaths. It was repealed in 1971, after the Pentagon Papers exposed the pre-planned posture of the administration, but the precedent — that an Executive can use a contested incident in international waters to extract sweeping authorization from Congress — has shaped every subsequent use-of-force debate, from the 2001 AUMF onward.

//Open Questions

Three questions remain genuinely unresolved, and dossiers that flatten them get the history wrong:

First, the question of executive knowledge in real time. Hanyok's NSA study proves the August 4 attack didn't happen. It does not prove Johnson and McNamara knew that on the night they ordered retaliation. The available record is consistent with both deliberate deception and with motivated reasoning under pressure.

Second, the role of mid-level intelligence officers. The SIGINT cherry-picking Hanyok documented happened well below the cabinet level. Whether it was bureaucratic self-protection, conformity bias, or directed manipulation is still contested.

Third, how complete the declassified record actually is. NSA declassified Hanyok's study; the underlying raw intercepts have been released in heavily redacted form. Independent verification of his interpretation against the full intercept traffic remains incomplete.