How LBJ Manipulated Events to Bring on War


Mr. Ellsberg was prosecuted by the Nixon administration for releasing the Pentagon Papers to the public. His latest book is: Secrets: A Memoir of Vietnam and the Pentagon Papers.

August 4, 1964 was Daniel Ellsberg's first day on his new job at the Pentagon as an analyst. A courier came running in with a message. It was the news of the second attack in two days on the USS Madox, a destroyer operating in the Gulf of Tonkin. President Lyndon Johnson told the American people that the North Vietnamese -- the Democratic Republic of North Vietnam (DRV) -- had fired upon U.S. ships in an unprovoked and unequivocal attack in international waters. Congress quickly granted the president the authority to respond to the attack. In his new book, Secrets: A Memoir of Vietnam and the Pentagon Papers, Ellsberg recounts his shock at the course of events he witnessed first hand from his perch at the Pentagon.

In chapter one he notes that nearly every statement the president made in a televised address to the country about the attacks was inaccurate. The attack was not unprovoked; the U.S. had recently shelled several DRV islands in an operation run by the United States, codenamed 34A. Nor was the attack unequivocal; there was no second attack--the ship's radar had picked up false readings of torpedoes that had never been fired. Finally, the Maddox and a sister vessel, the Turner Joy, were operating in an area long claimed by North Vietnam. The ships were on a secret mission, codenamed DeSoto, designed to elicit intelligence about DRV activities.

In chapter four, excerpted below, Ellsberg tells how the president used the Tonkin Gulf resolution to step up military action that turned into full-scale war.

From early September 1964, US"retaliatory" capability against North Vietnam was a cocked pistol. Officials just below the President were waiting for something to retaliate to and increasingly ready to provoke an excuse for attack if necessary. Six days after Assistant Secretary of Defense John McNaughton's September 3rd plan"to provoke a DRV response and to be in a good position to seize on that response. . . to commence a crescendo of GVN-US military actions against the DRV,"(1) the highest officials forwarded the proposal to the President for his decision. After recommending the immediate resumption of DeSoto patrols off the coast of North Vietnam and the resumption of 34A actions, both suspended since August 5, they added:"The main further question is the extent to which we should add elements to the above actions that would tend deliberately to provoke a DRV reaction, and consequent retaliation by us. Examples of actions to be considered would be running US naval patrols increasingly close to the North Vietnamese coast and/or associating them with 34A operations."(2)

I recall that these proposals excited a flurry of concrete suggestions by the Joint Staff as to how best to provoke an attack on US forces by the North Vietnamese if it proved hard to get a rise out of them. Along with running a US destroyer increasingly close to beaching on their coast, U-2 reconnaissance planes over North Vietnam could be supplemented by low-level reconnaissance jets flying progressively lower over populated areas, culminating, if necessary, in a supersonic flight that would break every window in Hanoi with a sonic boom.

But nothing so spectacular proved to be necessary. On the night of October 31 there was an attack on U.S. forces, killing five Americans, wounding thirty, and destroying or badly damaging eighteen of the B-57 jet bombers that had been deployed to Bien Hoa airbase in South Vietnam as part of a buildup rationalized by the Tonkin Gulf incidents.(3) The VC guerrillas didn't rely on advanced weaponry from the Soviet Bloc to accomplish this destruction. Having moved through heavily populated areas up to and within the American air base near Saigon without giving warning, they used 81 mm. mortars and satchel charges. Again Ambassador Taylor and the JCS strongly demanded retaliation, this time urging plausibly that to fail to respond would show weakness. The JCS proposed initial attacks in Laos and North Vietnam, to be followed by a night attack by B-52s on Phuc Yen airfield near Hanoi and a dawn strike by tactical fighters on other airfields and oil storage in the area of Hanoi and Haiphong. But the VC attack was three days before the election, and once again the pistol stayed cocked by decision of the candidate in the White House.

The military and Ambassador Taylor were extremely unhappy with this degree of restraint, predictably. They were assured, by Rusk among others, that after November 3rd things would be different. The organization of the NSC Working Group under Bundy on November 2 was part of this assurance. That group eventually reported consensus on the strategy of graduated pressures. The President endorsed this in principle on December 1, without committing himself to a definite date of beginning it. The consensus did not really include the JCS. They continued to urge a"hard knock" rather than a gradual approach, beginning soon, with or without provocation, with the attack they had urged on November 1. But they accepted the gradual approach as a first step toward their own strategy.

With the election over, it was taken for granted that the President would order a Tonkin Gulf-like retaliation, at least-with the likelihood it would launch a systematic campaign-- if there were another attack on US troops. Yet another exception turned up. On Christmas Eve, 1964, I was called in my apartment, within walking distance from the Pentagon, by the International Security Affairs (ISA) duty officer with the news that the Brinks Hotel BOQ (Bachelor Officers' Quarters) in Saigon had been blown up"by persons unknown." A car bomb had killed two Americans and wounded fifty-eight, plus thirteen Vietnamese. Ambassador Taylor, CINCPAC and the JCS called for an immediate reprisal bombing by 40 strike aircraft against the Vit Thu Lu Army barracks in North Vietnam.

Various reasons were given for the President's rejection, once again, of this recommendation. But John McNaughton told me at the office that the reason was simple. Despite earlier incidents and recommendations, there had been no attacks on the North since August 4th, and the President had also rejected advice to announce the reprisal policy openly. The public remained entirely unaware of the secret discussions, internal advocacy and preparing for attacks that had been going on for nearly a year within the Administration. Therefore the first bombing since Tonkin Gulf would come as a surprise to the American public, especially after the impression given during the election campaign. The President did not wish to present them with that particular surprise on Christmas morning; and later seemed too late for a clear signal to Hanoi.

Knowing the argument that had been going on for months between advocates of the gradual or all-out bombing campaigns, and the indications that the President might be reluctant to undertake either strategy and was not even eager to undertake the one-shot reprisals he had secretly promised, McNaughton and I found the VC direct attacks on Americans extremely reckless, even incomprehensible, if they didn't want to see North Vietnam blown apart. We were glad that Johnson had held back so far, for whatever reason, but we knew that couldn't happen again. Indeed, at the end of the year the President assured Ambassador Taylor that the next attack on Americans would bring a reprisal.

A month later on January 27, though I didn't know it at the time, McNamara and McGeorge Bundy argued forcefully to the President that the time had come"to use our military power in the Far East and to force a change of Communist policy."(4) He was no longer inclined to wait passively for an excuse for a"retaliatory" strike on the North. On January 28, DeSoto patrols were ordered back into the Tonkin Gulf for the first time in five months, with the mission of provoking an attack.(5) Naval retaliatory forces to be in position before the patrol commenced on February 3. If the Communist attackers didn't come to our troops on land, as they had at Bien Hoa and the Brinks, we would go to them by sea, as close as necessary to get them to attack. The American public, in the dark about the Administration's objectives and sense of commitment in Vietnam, still needed to be given a plausible reason for dropping bombs on North Vietnam. But it shouldn't take long now for one to come around. McGeorge Bundy recalled later that it was like waiting for a streetcar.(6)

As it had been once before, in late July 1964, the American pistol aimed at North Vietnam was not merely cocked but on a hair trigger. This time, with no election campaign pending, it was loaded for more than a single shot.


1.The Pentagon Papers, Gravel ed., vol. 3, pp. 193, 559.

2. ibid.

3. ibid., 208.

4. Robert S. McNamara, In Retrospect: The Tragedy and Lessons of Vietnam (New York: Times Books, 1995), p. 168.

5. H. R. McMaster, Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff, and the Lies that Lead to Vietnam (New York: HarperCollins, 1997), pp. 213-214.

6. Stanley Karnow, Vietnam: A History (New York: Viking, 1983), p. 411.