Excavation · Gulf of Tonkin Incidents (August 1964)

Did a second attack on US destroyers happen in the Gulf of Tonkin on 4 August 1964?

Area: Politics and Government · History · Question: Did it happen?, What did people know, and when? · Starts from a widely shared claim

The short answer

No (high confidence)

Where it stands

Current assessment · as of 2026-10-02

No. The second attack did not happen (high confidence). What senior officials knew about that is unsettled.

Central claim: On 4 August 1964 North Vietnamese torpedo boats attacked USS Maddox and USS Turner Joy a second time.
Evidence ●●●●○   Public belief ●●○○○   Source checked: primary

What would settle it. New intercepts, ship logs or North Vietnamese records showing boats present and firing on 4 August, or the missing original text showing it concerned 4 August.

The full reasoning

NSA's own historian re-read the intercepts and concluded that no attack took place that night. The intercept reports describe aircraft and salvage of boats damaged on 2 August, not an attack on the ships, and doubts reached the White House within hours. The 2 August clash with USS Maddox did happen. Whether senior officials knew the 4 August attack had not occurred when they sought the Resolution is a separate question, and the documents read support only that they knew of doubts. Two limits: the refutation rests largely on Hanyok's re-analysis, which we read but did not repeat, and the original Vietnamese text of the key report cannot be located.

Based on: tonkin-aug4-attack-occurred tonkin-aug2-maddox-engaged tonkin-senior-doubts-documented tonkin-senior-knowledge

We give no probability figure. Confidence words describe how sure we are of the basis, and no source supplies a number that would not be invented.

Challenge this assessment A challenge needs a specific claim and evidence we can check. How challenges work.

The headline finding

refuted: the claim below is false confidence high · our conclusion

On 4 August 1964 North Vietnamese torpedo boats attacked USS Maddox and USS Turner Joy a second time.

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. The intercept reports cited as evidence of 4 August read as follows [read]: Report 7 (instructions to ready two Swatow boats for "military operations" that night; boat T-333 to join if oil can be found); Report 12 (04/1630Z, a Swatow boat reports an enemy aircraft "falling into the sea; enemy vessel perhaps wounded"); Report 13 (04/1644Z, "we shot at two enemy airplanes and at least one was damaged. We sacrificed two comrades"). Report 13 mentions aircraft and not ships. Hanyok, who had access to 122 relevant SIGINT products, concludes: "no attack happened that night", and that "Hanoi's navy was engaged in nothing that night but the salvage of two of the boats damaged on 2 August" [read]. He also reports: no intercept showed an attack being planned, much less carried out [read]. The Navy's on-scene commander, Captain Herrick, sent a flash message (1:27 p.m. Washington time, 4 August) that many contacts "appear doubtful", that "freak weather effects" and "overeager sonarmen" may account for many reports, and that there were "no visual sightings" by the Maddox [quoted by Prados from a 28 Aug 1964 chronology; secondary]. A pilot overhead, Stockdale, later wrote there was "absolutely no gunfire except our own" (secondary). Hanoi has consistently denied any 4 August action; Bundy's own staff meeting recorded that "Hanoi has denied making the second attack" [read].

Would change if. new intercepts, ship logs or North Vietnamese records showing boats present and firing on 4 August, or recovery of the missing original Vietnamese text showing it concerned 4 August

Evidence nodes (what each does to the statement above)

Assumptions this rests on.

  • The declassified intercept reports, read through Hanyok's study, are complete enough that no unreleased 4 August material shows an attack.
    If wrong: New intercepts, ship logs or North Vietnamese records showing boats present and firing on 4 August would reopen the claim.
  • Hanyok's re-analysis of the intercepts is sound. We read it and the reports but did not repeat his analysis.
    If wrong: A reanalysis showing that Reports 7, 12 or 13 do describe an attack on the ships would weaken the refutation.

Alternatives considered.

  • An attack took place but left no trace in the signals intelligence.
    Why not preferred: Hanyok, with 122 relevant SIGINT products, concludes 'no attack happened that night' (read), and no source read shows one.
  • The key after-action report did concern 4 August.
    Why not preferred: Its original decrypted Vietnamese text and translation cannot be located (claim tonkin-original-text-missing), so it cannot be tested either way.

Why this confidence. Starts at high (The anchor was read directly (primary)).

  • risk of bias: no change Hanyok writes from NSA's own records against his own institution's earlier account, and the intercept reports themselves were read. The weight rests largely on his re-analysis, which we did not repeat.
  • unreported negative results: no change The original text of the key after-action report cannot be located. Considered and not applied because the other reports read do not describe an attack on the ships. Watch this one.
  • independent corroboration: no change Giap's account (via Prados, secondary) agrees; counted as corroboration only, not as a basis.

These reasons were drafted by the project and have not yet been reviewed.

Challenge or add evidence

Two separate events are routinely fused: the 2 August 1964 clash between USS Maddox and North Vietnamese torpedo boats, and the reported 4 August attack that led to the Tonkin Gulf Resolution. Stratah keeps them apart. The first happened. The second, on the declassified record, did not. Three further questions are kept separate again: how the false 4 August picture was assembled inside NSA, what senior officials were told and when, and what they knew. The first two have documented answers. The third is a genuine gap, and it is where conspiracy claims and the evidence part ways. This dig is also a calibration case: a real, documented failure of government reporting, filed strictly on its own evidence so that it neither inflates nor excuses claims in other digs.

Where evidence and belief differ

Evidence and adoption diverge on three points. On 4 August, historians agree no attack occurred, while the public memory of 1964 is of an attack. On method, the popular framing ('false flag', 'manufactured evidence') is stronger than the record: NSA's own historian found that only the SIGINT supporting the attack claim was passed up, calls the actions deliberate, but writes that this was not conspiratorial and involved no manufactured evidence or collusion at all levels. On knowledge, the documents show doubts reaching the White House within hours and a senior adviser welcoming the events as a justification, while the same documents show that adviser still believed an attack had happened. That gap is the product.

Timeline

Every point on the timeline links to its source. Open the timeline.

Challenges addressed

Objections raised against this excavation, in the form people raise them, with the test we ran and where it stands. Open ones are shown as plainly as answered ones. To add a challenge or evidence, use the button on the claim: it must name the claim and give a source we can check, because disagreement, popularity and repetition carry no weight. How challenges work.

The Gulf of Tonkin attack never happened.

partly answered raised by: Popular framing (see thread-tonkin-popular-framing); no single source

Correct for 4 August (refuted, high confidence, NSA historian's study read). Wrong for 2 August, when the Maddox was attacked and returned fire.

Test and claims

Separate the two dates and check each against the declassified record.

tonkin-aug2-maddox-engaged tonkin-aug4-attack-occurred

It was a false flag: the US manufactured the evidence.

partly answered raised by: Popular framing; no single source

Documented: only the intercepts supporting an attack reached senior officials. Whether individuals at NSA did this deliberately is proposed, not established. The historian calls it not conspiratorial, with no manufactured evidence and collusion at all levels, and writes that staff believed the attack had happened.

Test and claims

Read what NSA's own historian says about how the 4 August intelligence was assembled.

tonkin-sigint-selectively-presented tonkin-midlevel-deliberate-skew

The administration knew there was no attack and lied to get its resolution.

open: not yet testable raised by: Popular framing; no single source

Documented: doubts reached the White House within hours, and a senior adviser welcomed the events as a reason for a resolution already wanted. Not found in what was read: a statement that no attack had occurred. The same adviser's memo says there was an attack. Rated contested, low confidence. Unread sources are listed in the dig.

Test and claims

Look for any document in which an official says no attack occurred, as opposed to doubting the evidence.

tonkin-senior-knowledge tonkin-senior-doubts-documented tonkin-bundy-welcomed-resolution

The Maddox was sent in as bait to provoke the 2 August clash.

open: not yet testable raised by: Popular framing; no single source

No such document found in what was read (searched gap, low confidence). Covert raids on the North Vietnamese coast did precede the engagement, which is a separate and documented fact.

Test and claims

Look for an order or document directing the ship to provoke an attack.

tonkin-bait-order tonkin-34a-raids-preceded

Congress was kept in the dark about the covert raids.

partly answered raised by: Popular framing; no single source

The record read shows McNamara's testimony to the Senate about the raids differed from what he told the President privately. In 1968 he said he had told the committee about the raids, and Fulbright said it was unfair to ask the committee to vote on the resolution with the evidence uncertain. Intent is not decided here.

Test and claims

Compare what McNamara told the President privately, what he told the Senate, and what he and Fulbright said in 1968.

tonkin-mcnamara-34a-testimony tonkin-mcnamara-1968-account tonkin-fulbright-1968-evidence-uncertain

Tonkin shows official accounts of any event are fabricated.

answered: finding stands raised by: Common argument by analogy

No. A documented case lends no evidential weight to a different event; each is tested on its own evidence.

Test and claims

Ask whether a documented failure in one case adds evidence about a different event.

Publication review

Status: published before the review gate existed; partial review, no independent reviewer · reviewed 2026-10-02. Published by the owner before the gate existed. Read directly during the build: Hanyok's study, the Bundy staff-meeting memo, the declassified intercept reports, the Congressional Record pages for 6-7 Aug 1964 and 24 Jun 1970 (checked on page images), and the 1968 hearing text. Read only through secondary sources: the Prados essay and the LBJ tape transcriptions' commentary, and Herrick's message. Not read: FRUS, SNIE 50-2-64, the full floor debates (searched, not read in full). Quotes from the Prados essay and tape transcriptions have no stored text here and were not re-checked in this review. Headline status is draft. The review record.

All claims

refuted: the claim below is false confidence high · our conclusion

On 4 August 1964 North Vietnamese torpedo boats attacked USS Maddox and USS Turner Joy a second time.

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. The intercept reports cited as evidence of 4 August read as follows [read]: Report 7 (instructions to ready two Swatow boats for "military operations" that night; boat T-333 to join if oil can be found); Report 12 (04/1630Z, a Swatow boat reports an enemy aircraft "falling into the sea; enemy vessel perhaps wounded"); Report 13 (04/1644Z, "we shot at two enemy airplanes and at least one was damaged. We sacrificed two comrades"). Report 13 mentions aircraft and not ships. Hanyok, who had access to 122 relevant SIGINT products, concludes: "no attack happened that night", and that "Hanoi's navy was engaged in nothing that night but the salvage of two of the boats damaged on 2 August" [read]. He also reports: no intercept showed an attack being planned, much less carried out [read]. The Navy's on-scene commander, Captain Herrick, sent a flash message (1:27 p.m. Washington time, 4 August) that many contacts "appear doubtful", that "freak weather effects" and "overeager sonarmen" may account for many reports, and that there were "no visual sightings" by the Maddox [quoted by Prados from a 28 Aug 1964 chronology; secondary]. A pilot overhead, Stockdale, later wrote there was "absolutely no gunfire except our own" (secondary). Hanoi has consistently denied any 4 August action; Bundy's own staff meeting recorded that "Hanoi has denied making the second attack" [read].

Would change if. new intercepts, ship logs or North Vietnamese records showing boats present and firing on 4 August, or recovery of the missing original Vietnamese text showing it concerned 4 August

Evidence nodes (what each does to the statement above)

Assumptions this rests on.

  • The declassified intercept reports, read through Hanyok's study, are complete enough that no unreleased 4 August material shows an attack.
    If wrong: New intercepts, ship logs or North Vietnamese records showing boats present and firing on 4 August would reopen the claim.
  • Hanyok's re-analysis of the intercepts is sound. We read it and the reports but did not repeat his analysis.
    If wrong: A reanalysis showing that Reports 7, 12 or 13 do describe an attack on the ships would weaken the refutation.

Alternatives considered.

  • An attack took place but left no trace in the signals intelligence.
    Why not preferred: Hanyok, with 122 relevant SIGINT products, concludes 'no attack happened that night' (read), and no source read shows one.
  • The key after-action report did concern 4 August.
    Why not preferred: Its original decrypted Vietnamese text and translation cannot be located (claim tonkin-original-text-missing), so it cannot be tested either way.

Why this confidence. Starts at high (The anchor was read directly (primary)).

  • risk of bias: no change Hanyok writes from NSA's own records against his own institution's earlier account, and the intercept reports themselves were read. The weight rests largely on his re-analysis, which we did not repeat.
  • unreported negative results: no change The original text of the key after-action report cannot be located. Considered and not applied because the other reports read do not describe an attack on the ships. Watch this one.
  • independent corroboration: no change Giap's account (via Prados, secondary) agrees; counted as corroboration only, not as a basis.

These reasons were drafted by the project and have not yet been reviewed.

Challenge or add evidence

established confidence high · our conclusion

On 2 August 1964 North Vietnamese torpedo boats attacked USS Maddox in the Gulf of Tonkin; the Maddox and supporting aircraft returned fire.

Evidence●●●●●Public belief●●●●●Source checkedprimary
Evidence and limits

Anchor. Hanyok writes that on 2 August "the SIGINT system performed admirably when it provided sufficient warning to the Maddox to allow it to defend itself against the attack by the three DRV torpedo boats" [read]. Prados reports photographs of the boats and a dud shell kept by Admiral Moorer (secondary). Giap confirmed the 2 August action; at a 1997 conference he said it was ordered by a local naval command and not by Hanoi (secondary).

Would change if. records from either side showing the 2 August clash did not occur as described

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · what a source says

On 2 August the Maddox fired three warning rounds at the approaching North Vietnamese boats before they attacked. The Johnson administration did not report this and said the Vietnamese boats fired first.

Evidence●●●○○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. Hanyok: "At about 1505G, the Maddox fired three rounds to warn off the communist boats. This initial action was never reported by the Johnson administration, which insisted that the Vietnamese boats fired first." [read]

Would change if. Navy action reports or the Maddox log showing the warning rounds were not fired, or were reported at the time

Next step. Read the Maddox action report and the Navy's official history of the engagement; log whether the warning rounds appear.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · our conclusion

Covert OPLAN 34-A raids on North Vietnamese coastal targets, including the islands Hon Me and Hon Ngu, preceded the Maddox engagement (about half past midnight on 31 July) and were followed by another raid on the night of 3/4 August. The Maddox's mission was to collect North Vietnamese radio and radar emissions, expected to rise after a raid. Hanoi associated the destroyer with the raids.

Evidence●●●○○Public belief●●●●○Source checkedsecondary
Evidence and limits

Anchor. On 3 August McNamara told LBJ the link to OPLAN 34-A should be explained to Congress: "There's no question but what that had bearing on. And on Friday night, as you probably know, we had four [PT] boats from Vietnam manned by Vietnamese or other nationals, attack two islands." [read, transcription]. Prados reports that 34-A was directed and paid for by the United States, that the targets were chosen by the CIA, and that SNIE 50-2-64 shows the US directed 34-A to pressure Hanoi (secondary); the raids are described elsewhere as carried out by South Vietnamese forces.

Would change if. records showing the raids were not 34-A, or that Hanoi did not connect the destroyer to them

Next step. Read SNIE 50-2-64 (May 1964) and the declassified OPLAN 34-A documents; confirm the raid dates and who controlled the boats.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · our conclusion

The signals intelligence about 4 August reaching senior decision-makers was assembled and presented in a way that left out almost all the material showing no attack, and did not give them the complete and objective picture.

Evidence●●●●○Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Hanyok [read]: "SIGINT information was presented in such a manner as to preclude responsible decisionmakers in the Johnson administration from having the complete and objective narrative of events of 4 August 1964. Instead, only SIGINT that supported the claim that the communists had attacked the two destroyers was given to administration officials." The withheld information "constituted nearly 90 percent of all available SIGINT". The handful of reports supporting an attack "contained severe analytic errors, unexplained translation changes, and the conjunction of two unrelated messages". The six products used in the NSA chronology are "a bit more than 10 percent of the total available"; the other 90 percent was not offered to congressional committees.

Would change if. an independent audit of the 122 products shows the omitted material did not contradict the attack story

Next step. Obtain the NSA's declassified set of 140 documents and spot-check Hanyok's account of which products were and were not used.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

By 8 a.m. on 5 August 1964, the White House national security adviser told staff that the evidence for the second attack was "less than we had yesterday", because of "eliminating double counting and mistaken signals", that details such as the number of boats and torpedoes were "still somewhat uncertain", and that Hanoi had denied making the attack. He also said "This much seemed certain: There was an attack."

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. Memorandum for the Record, White House Staff Meeting, 5 August 1964 [read]: "On the first attack, the evidence would be pretty good. On the second one the amount of evidence we have today is less than we had yesterday. This resulted primarily from correlating bits and pieces of information eliminating double counting and mistaken signals. This much seemed certain: There was an attack. How many PT boats were involved, how many torpedoes were fired, etc. -- all this was still somewhat uncertain. This matter may be of some importance since Hanoi has denied making the second attack."

Would change if. evidence the memorandum is not authentic, or a different record of the same meeting

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

At the same meeting, when a new staffer questioned the need for a congressional resolution, the national security adviser replied jokingly that the matter should perhaps not be thought through too far; as for himself, he "welcomed the recent events as justification for a resolution the Administration had wanted for some time."

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. Same memorandum [read]: "Bundy, in reply, jokingly told him perhaps the matter should not be thought through too far. For his own part, he welcomed the recent events as justification for a resolution the Administration had wanted for some time."

Would change if. a second record of the meeting showing the remark was different or not made

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · what a source says

At 9:43 a.m. Washington time on 4 August, before any attack on the destroyers was reported, McNamara told President Johnson the ship was "allegedly" to be attacked that night, and they discussed what to authorize if it was attacked.

Evidence●●●○○Public belief●●○○○Source checkedsecondary
Evidence and limits

Anchor. McNamara [read, Prados transcription]: "...the tracks are laid down very clearly; they go through the three command channels to get out there; this ship is allegedly uh to be attacked tonight - we don't like to see a change in operation plan of this kind at this time." He then reports a recommendation that the commander "be authorized to pursue the attacking vessels in the event he is attacked and destroy their bases."

Would change if. the LBJ Library audio or its own transcript differing materially from this transcription

Next step. Listen to, or read the LBJ Library's own transcript of, the 4 August 9:43 a.m. call.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

On 3 August 1964 McNamara privately told President Johnson that OPLAN 34-A "had bearing on" the Maddox engagement and should be explained to Congress. In formal testimony to the joint Senate committees on 6 August 1964 he said: "Our Navy played absolutely no part in, was not associated with, was not aware of, any South Vietnamese action, if there were any."

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. 6 August testimony as read by Chairman Fulbright into the hearing of 20 Feb 1968 (S. Hrg. 90-187, p. 30, "page 23 of the transcript of hearings") [read, OCR text]: "Our Navy played absolutely no part in, was not associated with, was not aware of, any South Vietnamese action, if there were any. I want to make that very clear to you. The Maddox was operating in international waters, was carrying out a routine patrol..." McNamara, asked by Senator Cooper, confirmed it: "That is correct" and that it was his belief then and now. 3 August call [Prados transcription]: "...explain this Op Plan 34-A, these covert operations. There's no question but what that had bearing on."

Would change if. the 1964 transcript itself (Executive Sessions vol. XVI, or the 1966 edited hearing) showing different wording

Next step. Read the 6 August 1964 transcript itself (the edited joint hearing, and Executive Sessions vol. XVI, pp. 291-299) and listen to the 3 August LBJ call; confirm the page numbers of the 1968 quotation on the page images.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

In his prepared statement of 20 February 1968 McNamara said he had told the Senate Foreign Relations Committee about the South Vietnamese raids "in an informal meeting on August 3, 1964, after the attack on the Maddox", that the subject "was again raised in lesser detail" in his testimony on 6 August, that Senator Morse asked that day about a connection between the patrol and the raids and he answered "that there was no connection", and that he "learned subsequent to my testimony of August 6, 1964, that another South Vietnamese bombardment took place on the night of August 3-4". On the second attack he said "some uncertainty existed, and to this day exists, about some of the precise details of the attack. But there should be no uncertainty about the fact that an attack took place."

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. Hearing before the Senate Committee on Foreign Relations, 20 Feb 1968 (S. Hrg. 90-187), prepared statement pp. 14-16 and the questioning pp. 30-31 [read, OCR text].

Would change if. the 3 and 6 August 1964 committee records showing that what he told the senators then differs from this account

Next step. Read the 3 August informal meeting record, if any survives, and the 6 August transcript (Executive Sessions vol. XVI) to test his 1968 account against 1964.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

At the hearing of 20 February 1968 Fulbright, as chairman, said: "I think this committee ... it was very unfair to ask us to vote upon a resolution when the state of the evidence was as uncertain as I think it now is ... none of those intercepts were mentioned to us, I don't believe, in the testimony on August 6. Your statement and General Wheeler's was without any doubt, any equivocation that there was an all-out attack."

Evidence●●●●○Public belief●●○○○Source checkedprimary
Evidence and limits

Anchor. S. Hrg. 90-187, 20 Feb 1968, p. 80 [read, OCR text; the page number is from the text layer]. Fulbright qualified the point with 'I don't believe'.

Would change if. the 6 August transcript showing the intercepts were mentioned to the committees

Next step. Read the 6 August 1964 transcript to test 'none of those intercepts were mentioned'.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

Congress passed the Tonkin Gulf Resolution (7 August 1964) in response to the reported attacks, with the 4 August report central to the administration's case.

Evidence●●●●○Public belief●●●●●Source checkedprimary
Evidence and limits

Anchor. Congressional Record, 7 Aug 1964 [read]: p. 18470 gives the Senate roll call "yeas 88, nays 2" (nays Gruening and Morse; 10 not voting) and says "So the joint resolution (H.J. Res. 1145) was passed". The debate prints, as "Exhibit 1", the President's address of 4 August: "The initial attack on the destroyer Maddox on August 2 was repeated today by a number of hostile vessels attacking two U.S. destroyers with torpedoes." The enacted preamble says Communist naval units "have deliberately and repeatedly attacked United States naval vessels". House: the Record prints 416-0 (Roll No. 210, p. 18555; see tonkin-house-debate-aug7). Signed 10 August per Senate.gov (secondary).

Would change if. the full floor debate showing the attacks were not central to the case made to Congress

Next step. Read the whole Senate debate of 6-7 August 1964 and the House debate of 7 August; log which speakers cite the 4 August attack and which do not.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

Opening the Senate debate on 6 August 1964, Fulbright told the Senate: "On August 4 the Maddox and another destroyer, the C. Turner Joy, were again attacked by North Vietnamese torpedo boats in international waters. The attack, which lasted for over 2 hours, was without any doubt a calculated act of military aggression." He also said the committees had approved the resolution with only one dissenting vote.

Evidence●●●●●Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Congressional Record, Senate, 6 Aug 1964, pp. 18399-18400 [read on the page image, p. 18399].

Would change if. a different printing of the speech in the permanent Record

Next step. Read what the two committees were told before reporting (Foreign Relations and Armed Services joint session, 6 Aug; Senate Historical Series vol. XVI).

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · what a source says

The House passed the Tonkin Gulf Resolution on 7 August 1964 under suspension of the rules: the Record prints "yeas 416, nays 0, answered 'present' 1, not voting 14" (Roll No. 210, p. 18555; the one "present" is Powell), on Morgan's motion to pass H.J. Res. 1145 "with an amendment". In the debate (pp. 18538-18555, read in full by a delegated reader) only Zablocki cited the 4 August attack by date ("on August 4--U.S. destroyers, the Maddox and the C. Turner Joy, were again attacked... The attack lasted 2 hours"); Albert, Broomfield and Roush referred to two attacks without dates; no Member doubted that the attack occurred or asked about the evidence for it; no Member named OPLAN 34-A or South Vietnamese raids. Several (Derwinski, Chamberlain, Alger) complained about the information given to Congress, and Halleck, Morgan and Barry referred to what they were told at White House or committee briefings.

Evidence●●●●○Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Congressional Record, House, 7 Aug 1964, pp. 18538-18555 (roll call p. 18555), with a Senate message at p. 18612. Delegated reading; quoted passages image-checked by that reader.

Would change if. a re-reading of the pages showing a Member who questioned the attack or the evidence, or the later pages (18556-18625) containing relevant debate

Next step. Re-check the Zablocki, Halleck, Barry and Chamberlain quotes on the images myself; read the committee amendment text (the 'Constitution of the United States' wording) and reconcile 'with an amendment' in the House with the Senate message that it passed H.J. Res. 1145 'without amendment' (p. 18612); read what the Foreign Affairs Committee was told (its report).

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence high · what a source says

On 24 June 1970 the Senate voted 81 to 10 to agree to the Dole amendment repealing the Gulf of Tonkin Resolution effective at the end of the 91st Congress. The ten nays were Allen, Bellmon, Eastland, Ervin, Fulbright, Hollings, Long, McClellan, McGee and Stennis.

Evidence●●●●○Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Congressional Record, Senate, 24 June 1970, p. 21132 [read]: "The result was announced-yeas 81, nays 10" [No. 167 Leg.], "So Mr. DOLE's amendment (No. 715) was agreed to." Voteview numbers the same roll call 398 in its index.

Would change if. the Record's roll call being corrected

Next step. Read the amendment text (Amendment No. 715) and the later vote on S. Con. Res. 64 (10 July 1970, 57-5 per Voteview) in the Record.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

established confidence moderate · what a source says

In the 24 June 1970 Senate debate on the Dole amendment, the ten senators who voted Nay gave different reasons, or none in the pages read. Fulbright said he had intended to vote for repeal and changed his vote after "the assurance of the majority leader" that the Mathias concurrent resolution would be brought up. Allen and Hollings spoke against repeal itself. Stennis mixed opposition to repeal with a procedural complaint. McGee called the repeal a "political charade". No stated reason was found for Bellmon, Eastland, Ervin, Long and McClellan. Several senators who voted Yea (Javits, Moss) said a concurrent resolution would be better but still backed the amendment. A 1970 vote on this amendment is therefore not a clean measure of a senator's view of the 1964 Resolution.

Evidence●●●○○Public belief●○○○○Source checkedprimary
Evidence and limits

Anchor. Congressional Record, Senate, 24 June 1970. Fulbright, p. 21131 [read on the page image]: "I said earlier that I would vote for the repeal because, of course, I am committed to repeal and have been for a long time. However, upon further consideration, in view of what has been said on the floor of the Senate, and also in reliance upon the assurance of the majority leader that he would bring up the Mathias resolution--which is a concurrent resolution..."; p. 21132: "I shall vote against the repeal today and preserve the integrity of the procedure which I believe is very important to this body." Earlier that day, p. 21119: "I am going to vote for its repeal. I am committed to it, of course." Nay list p. 21132. Reasons of Allen (p. 21123), Hollings (p. 21122), Stennis (p. 21121), McGee (pp. 21130-31) per the delegated reading.

Would change if. a reading of the pages showing a different reason for any of the ten, or of the pages 1-14 and 34-45 not yet read in full

Next step. Re-check each non-Fulbright quote on its page image; read the 22 June debate (the Record file read did not include it) and the text of Amendment No. 715.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Evidence nodes (what each does to the statement above)

Challenge or add evidence

contested confidence low · our conclusion

Senior administration officials knew the 4 August attack had not occurred when they ordered reprisal strikes and sought the Resolution, as opposed to acting on a genuine, if flawed, belief that it happened.

Evidence●●○○○Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Doubts reached the White House within hours, and a senior official welcomed the events as justification. The same record shows that official still thought an attack had happened, and the study found the SIGINT reaching senior officials had been filtered. "Knew of doubts" and "knew nothing happened" are different claims; the documents read support the first and do not settle the second.

Would change if. a dated senior communication stating, or clearly showing knowledge, that no attack had occurred, or a document showing senior officials were told the SIGINT was being filtered

Next step. Read FRUS 1964-68 vol. I documents dated 4-7 August 1964 and the LBJ Library tapes for those days; log who said what about whether an attack occurred.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Disputed by.

  • historians, not named in the sources read: divided on what senior officials knew
    this claim's adoption note (source not read here)

Challenge or add evidence

proposed confidence moderate · our conclusion

Individuals at NSA deliberately presented only the SIGINT that supported the attack claim, in an active effort to make it fit the Navy's account, without manufacturing evidence and without a conspiracy at all levels. Hanyok judges that they themselves believed the attack had happened and rationalized the contradictory evidence away. The exact how and why are not known.

Evidence●●●○○Public belief●●●○○Source checkedprimary
Evidence and limits

Anchor. Hanyok [read]: "From this evidence, one can easily deduce the deliberate nature of these actions... This was an active effort to make SIGINT fit the claim of what happened during the evening of 4 August." And: "This mishandling of the SIGINT was not done in a manner that can be construed as conspiratorial, that is, with manufactured evidence and collusion at all levels. Rather, the objective of these individuals was to support the Navy's claim that the Desoto patrol had been deliberately attacked." On motive: "The question why the NSA personnel handled the product the way they did will probably never be answered. The notion that they were under 'pressure' to deliver the story that the administration wanted simply cannot be supported." Also: the NSA personnel "believed that [the attack] had occurred"; "The overwhelming body of reports, if used, would have told the story that no attack had happened. So a conscious effort ensued to demonstrate that the attack occurred"; "what was issued in the Gulf of Tonkin summaries beginning late on 4 August was deliberately skewed to support the notion that there had been an attack... That the NSA personnel believed that the attack happened and rationalized the contradictory evidence away is probably all that is necessary to know in order to understand what was done." He finds no "smoking gun" memoranda or notes in the files outlining any plan or stating a justification.

Would change if. a document showing the omissions were mistaken rather than deliberate, or showing direction from outside the NSA personnel, or evidence that intercepts were fabricated

Next step. Read the oral-history interviews in the NSA's 2005 release and check what participants said about how reports were chosen.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Challenge or add evidence

searched gap confidence moderate · result of a search

The original decrypted Vietnamese text of the key "after-action" intercept used to support the 4 August attack, and an important translation from the San Miguel listening station, cannot be located.

Evidence●●●○○Public belief●○○○○Source checkedprimary
Evidence and limits

Anchor. Hanyok: "the original decrypted Vietnamese text, and an important translation from San Miguel cannot be located." [read] He calls it "unusual" that the original text could not be found (and was not used as early as October 1964), and that these are the only missing papers in the San Miguel reports "allows for suspicion to shade any further discourses". A 1972 note cited by Hanyok records NSA's deputy director, Dr. Louis Tordella, conceding that the 4 August intercept could refer to the 2 August attack, and being unable to produce the original. Report 13 above is the NSA summary that was relied on; the original it summarized is the missing item.

Would change if. the original text or the San Miguel translation is found

Next step. Search the National Archives, NSA, CIA and LBJ Library finding aids for the original Vietnamese text and the San Miguel translation; log every hit and every miss.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Challenge or add evidence

searched gap confidence low · result of a search

No passage has been found in the Senate debates of 6 and 7 August 1964 in which a senator questioned whether the 4 August attack on the destroyers occurred. Senators whose words were found citing it treated it as fact: Fulbright ("without any doubt"), Smathers ("The facts are indisputable"), Keating ("no doubt whatsoever that the second attack"), Scott and Russell (who mention the "second attack" and where the ship was). Others in the same debate (Morse, McGovern) raised the earlier raids and how close the destroyers patrolled to the coast.

Evidence●●○○○Public belief●●○○○Source checkedsecondary
Evidence and limits

Anchor. Machine search of the Congressional Record text (OCR, column order partly scrambled). The search terms were tied to 4 August; a doubt expressed in other words would be missed. Church (6 Aug) said the destroyers, "from all that we have been told, were not engaged in any aggressive action directed against the shores of North Vietnam" [OCR wording, not verified on the image]: senators were acting on what they were told, which is the P8 question the dig has not yet answered.

Would change if. a senator's words in the debate casting doubt on whether the 4 August attack occurred

Next step. Read both days in full, page by page (6 Aug pp. 18399-18436; 7 Aug pp. 18441-18471), verify each quote on its image; also the House debate of 7 Aug (pp. 18538 ff.).

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Challenge or add evidence

searched gap confidence low · result of a search

No document read or reported so far shows an order to provoke the 2 August clash by sending the Maddox in as bait.

Evidence●●○○○Public belief●●○○○Source checkedsecondary
Evidence and limits

Anchor. Text searches of the Hanyok study found no mention of bait; its only uses of "provocation" concern the Aug 4 incident as the pretext for the Resolution and September 1964 patrols. Prados's essay and the tape commentary say the Maddox's mission was to record North Vietnamese radio and radar emissions expected to rise after a 34-A raid (so the patrol was coordinated with the raids), and that the mission commander "had been briefed in Taiwan previously that there would be" no attack and was unaware of the raids. None describes an order to provoke.

Would change if. a document ordering or planning the Maddox patrol as a provocation is found

Next step. Read SNIE 50-2-64 and the NSA's remaining declassified Tonkin documents (about 140 released in 2005); search each for provocation or bait language; log every hit and every miss.

Start this step → (signs you in with GitHub; trusted contributors can send it to the dig agent)

Challenge or add evidence

Sources

The data

These pages are generated from plain YAML: claims.yaml · sources manifest. If you can show a finding is wrong, challenge it.