• Email
  • Single Page
  • Print

Tonkin Bay: Was There a Conspiracy?

Normally an incident of this sort involving so many uncertainties would be followed by a naval board of inquiry. Such a review followed the so-called “third” Tonkin Gulf incident of September 18, 1964. The board found that although two other US destroyers had held numerous radar “contacts,” had reported attack, had seen “tracer bullets” and “light flashes,” and in the end had fired some 300 rounds of ammunition, there had in fact been no North Vietnamese attack. The lack of any such inquiry into the August 4 incident is itself one further ground for suspicion.

Washington was not unaware of all this confusion when it ordered the retaliatory air strike. On the contrary Captain Herrick’s first expression of doubt had been relayed to Washington by the Naval Communications Station, Philippines, as early as 1:27 P.M. EDT, August 4 (or 1:27 A.M., August 5, in the Tonkin Gulf):

Review of action makes many reported contacts and torpedoes fired appear doubtful…. Freak weather effects and over-eager sonarman may have accounted for many reports. No actual visual sightings by Maddox. Suggest complete evaluation before any further action.

But it was about this time that President Johnson had agreed in principle to an American reprisal, making it clear that he wanted more positive information before the reprisal attack was launched. McNamara’s testimony showed that for the next four and one-half hours there was confusion and debate over whether to proceed: “I personally called Admiral Sharp and…said we obviously do not want to carry out retaliatory action unless we are ‘damned sure what happened.’ “6 Certainty was restored, according to his account, by two corroborating sources.

The first was Admiral Sharp, who communicated many times with subordinate commanders, then phoned in at 5:23 “stating that he was convinced the attack had occurred and that all were satisfied it had.” If Sharp really said this, it was a gross misrepresentation. Herrick, according to the cables seen by the Committee staff, had merely confirmed the “apparent attempted ambush,” not the alleged attack two hours later. Some time after the strike order had been released, after 6:00 P.M. EDT, Herrick sent another cable which itemized his continuing doubts: for some reason this cable was not received in Washington until 10:59 P.M., when the retaliating airplanes were already airborne. At 11:00 the ship Turner Joy was again asked in an “urgent” cable for witnesses and evidence. Their reply indicated still further grounds for doubt.

McNamara’s second source of corroboration, the allegedly “unimpeachable” intercepts of North Vietnamese communications, deserves far more scrutiny than Goulden gives it. There were four groups of intercepts concerning the events of August 4, all of them supplied by the intelligence personnel aboard the Maddox to Herrick, to CINCPAC, and to Washington. The first two are credible, but do not prove an attack. The latter two, which if true would clearly confirm an attack, are both highly dubious, and not just because they both contain false information. McNamara chose to summarize the intercepts for the record:7

No. 1. Located the position of the Maddox and the Turner Joy.

No. 2. (Received Washington about 9:20 A.M.) Directed three vessels “to make ready for military operations.”

(It is unfortunate that Goulden twice—pp. 147, 207—echoes McNamara’s original characterization of this intercept as “North Vietnamese orders…to initiate the attack.” Senator Gore, who had just acquired the text, successfully challenged McNamara on precisely this point, and forced him to accept the more moderate phrasing.)8

No. 3. (Received Washington about 11:00 A.M.) Reported an American plane falling and ship wounded.

No. 4. (Received Washington “immediately after the attack ended,” i.e., just after 1:30 P.M.) Reported “that they had shot down two planes and sacrificed two ships,” and added further details.

The third and fourth groups, in other words, came in after the first highly excited flurry of cables from the American destroyers reporting attack. What made them so credible at the time is that in part they also echoed the cables: the Maddox had reported the disappearance of an unidentified aircraft from its screen, and the Turner Joy had reported sinking two enemy boats. Intercept group No. 4, which arrived in Washington in the crucial minutes right after Herrick’s expression of doubts, must have seemed like a clincher. As Herrick himself told Goulden in an interview, “We heard…their damage report confirm our assessment that two of the boats had been sunk.” It must have had an equal impact on McNamara, who has stated that his own decision at 6:07 P.M. to go ahead with the strike order was based particularly on “the communications intelligence.”

Nevertheless, as Goulden points out, Herrick at the time was not completely convinced by the last intercept, and for four and one-half hours McNamara was not either. What is more important: by evening (Washington time) the Turner Joy was no longer certain it had sunk two vessels; and in retrospect the grounds for believing so seem suspect. The Turner Joy had seen a “target” disappear from its radars, supposedly accounting for one ship, and some personnel thought they had seen a column of black smoke, supposedly accounting for another. But the radars were disturbed by atmospheric conditions that night and it was “dark as hell.” Goulden himself doubts that these bits of evidence prove anything and he quotes without challenging it the conclusion of a high-level Pentagon informant that “the so-called second attack of August 4 never took place.”

If this informant is correct, the similarity between the destroyer’s cables and the last intercepts is no longer corroborative, but highly suspicious. It is made all the more suspicious by the absence thus far of any credible evidence that US aircraft were damaged or missing.9 How, then, are we to explain the strange circumstance that a North Vietnamese “intercept” reported information which echoed the cables sent by the Maddox and the Turner Joy but which later turned out to have no convincing basis in fact? One possible explanation is that the North Vietnamese did, in fact, radio the news that two of their ships were lost, but somehow did so in error. It is also possible that the shots fired by the Turner Joy somehow struck two North Vietnamese ships. Both possibilities seem remote ones, however, and in any case neither accounts for the reports about downed aircraft. A third possibility is that American technicians were subconsciously influenced by the destroyers’ cable traffic in their hearing or interpretation of the North Vietnamese messages; but it is difficult to imagine how such errors could have been consistently made in the case of both groups of intercepts. A fourth possibility should therefore be considered: that the intercept may deliberately have been fed in—or distorted in the process of translation or summary—by American intelligence personnel in order to end the fateful and unexpected indecision in Washington.10

This possibility is increased by another undisputed anomaly: the failure of either one of the destroyers to detect any electronic activity from North Vietnamese ships—whether radar or radio communication—after about 2:30 P.M. Tonkin time, or some six hours before the alleged incident. Under these circumstances it is not only hard to imagine how the North Vietnamese could have conducted an attack out at sea in the darkness, it is also hard to imagine the origins of the information in the third and fourth “intercepts.” Herrick confirmed to Goulden what the Fulbright Committee had already learned, that “We had no radio contact, or heard no communications going on between the PT boats” (p. 153). As Goulden quite properly asks:

The communication van’s ability to intercept North Vietnamese messages had been amply demonstrated during the preceding four days; why, then, no intercepts from the PT boats during the August 4 incident? Messages from director ships, or a headquarters on Hon Me, which were audible to the North Vietnamese would also have been audible to the Maddox’s monitors—yet Herrick avows none were heard during the engagement. What, then, was the origin of the damage report?11

If such grave suspicions about the performance of our intelligence network are unfounded, there is much that can be done to put them at rest. The intercepts should be made public, both in their original form and as characterized at the time in intelligence reports. Investigation should be made to determine whether anything happened to US aircraft on that day which would have led the North Vietnamese to think they had shot down two planes. Even the disclosure of an honest error would serve, at this point, to clear the air rather than to poison it.12

In January 1968 Ambassador Goldberg, in presenting the American version of the Pueblo incident to the UN Security Council, did not hesitate to quote directly from intercepts of North Korean PT-boat communications which were only three days old. Yet one month later McNamara would not even discuss the North Vietnamese intercepts with the Fulbright Committee until its staff adviser (who had received the appropriate clearance when in the Navy) had been cleared from the room. Such furtiveness, until it has been explained, only deepens the credibility gap.

McNamara’s dilemma of 1964 must be grasped. To doubt the existence of an attack on August 4 (as Goulden clearly does, though he is not explicit on this point) is to doubt the credibility of the intelligence network which “proved” there was one.13 And if one does not choose with McNamara to believe the “proof,” then there is much more to question in the lower echelons of our national security bureaucracy than Admiral Sharp’s evident eagerness to bomb North Vietnam. Sharp might well have been restrained by McNamara, had it not been for the performance of the intelligence community’s technicians who handled the intercepts.

Sooner or later, most discussions of the Tonkin Gulf incidents (including McNamara’s) return, if only to dismiss it, to the possibility of conspiracy. In fact two kinds of conspiracy have been hinted at, a conspiracy by the Administration, and a conspiracy against it. (Failure to distinguish between these has led to confused accounts in which McNamara appears simultaneously as villain and victim.) On the first point, there have been charges of deliberate provocation of the North Vietnamese, as when I. F. Stone in these pages neatly characterized the Tonkin Gulf incidents as a “question not just of decision-making in a crisis but of crisis-making to support a secretly pre-arranged decision.”14

We have already seen South Vietnam’s first bombardment raids against North Vietnam (at Hon Me Island) were followed by the Maddox‘s persistent feints toward Hon Me Island two days later. In the preceding weeks Hanoi had been subjected to other new pressures. On July 19 General Khanh had made a major public appeal for a bac thien or march to the north, and on July 22 Marshal Ky revealed that CIA-trained commando operations against North Vietnam had been stepped up 40 percent since July 10.15 Goulden reveals that every one of these escalations (including Khanh’s calls for invasion) had been suggested and finally approved as parts of a “measured pressure” plan prepared by an inter-agency Vietnam Working Group appointed by Johnson and headed by William Sullivan, a forty-one-year-old Foreign Service Officer. (One of the Sullivan group’s working papers specifically mentioned the patrol boat base and radar station on Hon Me Island.)

  1. 6

    McNamara explained he could report his exact words because “I have a transcript of that telephone conversation” (Hearing, p. 56). Two hours later Fulbright asked for the transcript, since the Defense Department had agreed to supply all relevant communications. McNamara first replied that he would be happy to make it available, but soon qualified this promise by saying he did “not know how much of this will be recorded.” Forgetting what he had said earlier, he now stated that “the source of my statement is my memory of what I myself said and did” (pp. 60-61). This important discrepancy is unfortunately overlooked by Goulden (pp. 152-53). The transcript has apparently never been supplied.

  2. 7

    Hearing, p. 92.

  3. 8

    The orders were directed to two of North Vietnam’s “Swatow”-type motor gunboats (armed with two 37 mm. deck guns but without torpedoes) and to one PT boat “if the PT could be made ready in time.” It is difficult for those with naval experience to believe that lightly armed motor gunboats could be ordered to attack two US destroyers.

  4. 9

    All of the published official reports about the second incident confirm the original report in the Times that the United States suffered no hits, damage, or casualties (NYT, Aug. 5, 1964, p. 1). Thus the report of two downed US aircraft would indeed seem to be, as Goulden concludes, a “total untruth” (p. 153).

  5. 10

    The hypothesis that American intelligence was responsible for the fourth “intercept” itself raises problems, e.g., why would the intercept say that two planes had been shot down when Washington would have known this was not the case?

  6. 11

    If the third and fourth intercepts are valid, how are we to account for the conflicting testimony of the North Vietnamese prisoner-of-war? US naval intelligence officers who interviewed him for more than 100 hours in 1966 reported that he was “co-operative and reliable…. Yet he specifically and strongly denies that [on Aug. 4] any attack took place.” [Hearing, p. 75: Goulden, p. 213] McNamara anticipated the Fulbright Committee’s questions about the prisoner with a new piece of evidence.

    His [the prisoner’s] disclaimer of PT participation is contradicted by information received from a later captive. A North Vietnamese naval officer captured in July 1967 provided the name of the commander of the PT squadron. In intelligence reports received immediately after the August 4 attack, this commander and his squadron were identified by name and number as participants. [Hearing, p. 18, emphasis added; cf. p. 74.]

    But the Committee staff subsequently established that the second testimony, insofar as it was relevant, corroborated rather than contradicted the first. The Pentagon, it developed, had attached importance to the second prisoner’s testimony, not because he confirmed or denied the second incident (he was not involved and had no information on this point) but solely because he corroborated the name mentioned in the “intelligence reports.” Goulden identifies these reports as “radio intercepts”; and McNamara stated that intercept group No. 4 came “immediately after the attack ended.” (The Pentagon’s interest in the second prisoner itself suggests official concern as to the intercepts’ reliability.) Unfortunately for McNamara’s case, the second in command to the PT squadron commander named in the intercepts was none other than the first prisoner, who had stated “definitely and emphatically that no PTs could have been involved.” [US Congress, Senate, Committee on Foreign Relations, The Gulf of Tonkin, the 1964 Incidents: Part II, Supplementary Documents, p. 13; Goulden, p. 214.] In other words, McNamara’s new evidence about the “intelligence reports” only makes the reports look all the more dubious.

  7. 12

    The North Vietnamese PT squadron commander should also be interrogated, unless (as has been rumored) he has since been returned to North Vietnam in an exchange of prisoners.

  8. 13

    McNamara, in his carefully worded written statement, said it would be “monstrous” to insinuate that “the Government of the United States induced the incident on August 4.” Misquoting him, Goulden says “McNamara has called it ‘monstrous’ to insinuate that the August 4 incident was a manufactured lie, product of a plot of the White House and the military; I agree.” I do not know if Goulden, in this apparently casual language, intended thus to leave open the question of a plot by intelligence personnel. In any case he does not pursue it.

  9. 14

    I. F. Stone, “McNamara and Tonkin Bay: The Unanswered Questions,” New York Review of Books, March 28, 1968, p. 11.

  10. 15

    Saigon Post, July 23, 1964; I. F. Stone’s Weekly, Sept. 12, 1966, p. 3.

  • Email
  • Single Page
  • Print