Congress Concluded There Were Two Gunmen. It Has Never Withdrawn It.

DALLAS — The United States House of Representatives concluded, in a report it published in 1979, that President Kennedy was probably assassinated as the result of a conspiracy.

That is not a claim. That is a citation.

No committee has reconvened to say otherwise. No successor body has superseded it. It sits in the record where it was put, and the thing it rests on is a scratched belt of plastic from a police radio room.

What the recording is

On the afternoon of 22 November 1963 a transmit button on one Dallas police motorcycle stuck open.

For about five and a half minutes, the channel it was on could carry nothing else. Everything the microphone heard — engine, wind, road, whatever else — was cut into a dictabelt at headquarters, which is what police departments used before tape.

Nobody thought about it for fifteen years.

What the engineers did with it

In 1978 the House committee hired acoustic engineers, who found six impulse patterns on the belt and went to Dallas to test them.

They did it the direct way. They fired rifles in the plaza from candidate positions, recorded the echoes off the buildings, and compared those echo signatures to the patterns on the belt. Two physicists then modelled the echoes rather than shooting them, refined the match, and reported that one impulse pattern had about a 95 per cent probability of being a shot from the fence on the knoll.

The committee published in January 1979. The acoustics were the load-bearing part of the finding, and the committee said so.

The drummer

Later that year the recording was pressed onto a floppy plastic disc and bound into a men's magazine as a giveaway.

An amateur drummer bought it and played it over and over, which is what drummers do to recordings, and heard something under the impulses that nobody in the investigation had reported: a voice.

The voice is a sheriff giving an instruction about securing the scene. That instruction was given roughly a minute after the shooting.

If it is under the impulses, the impulses are in the wrong minute.

The panel

A national scientific panel reviewed the whole thing in 1982 and agreed with the drummer.

Its report is not a document about the assassination. It is a document about timing, and it concludes that the impulses occurred about a minute after the shots, and that whatever they are, they are not the shots.

That is where most accounts stop, and it is where most accounts of the other kind stop too — one side finishes at 1979 and the other finishes at 1982, and the two sides have been finishing in different years for four decades.

It came back

In 2001 a paper in a peer-reviewed forensic journal argued that the crosstalk had been mistimed and that the impulses stand. In 2005 a rebuttal argued the reverse.

It has continued in that register ever since: careful people, in print, with working shown, disagreeing.

This newspaper reports that as the state of the question, because it is the state of the question.

What we did

We did not want to argue about audio we had only heard fourth-hand. So we bought a first-generation transfer, and we gave it to an acoustic engineer who had never worked on this case, with a commissioning letter that said nothing about what it was.

We asked him one question. How far apart are the impulses and the sheriff's voice?

His answer is closer than the 1982 panel said and further apart than the 2001 paper said. It supports nobody. His full statement runs at the foot of this story and contains the most useful paragraph we have printed this year, about what a probability is a probability of.

The thing we could not close

The acoustic case needs a specific microphone in a specific place.

For it to work, that motorcycle has to be in the plaza, roughly where the model puts it. Several officers have been proposed over the years. None has ever been established. One of them said, on the record, that he was not there.

Our correspondent asked the archivist whether the dispatch logs could settle it. They cannot. The logs record who was assigned, not where a man's front wheel was at half past twelve.

Where that leaves the record

An official finding of probable conspiracy, standing, unwithdrawn, and load-bearing on evidence that the government's own review panel said was in the wrong minute.

Both halves of that sentence are true and they have been true together since 1982.

This newspaper's view is that the interesting fact here is administrative rather than acoustic. A conclusion entered the permanent record of a legislature, the basis for it fell apart three years later in a different institution, and there is no mechanism by which the second event edits the first.

Nobody buried anything. The filing system simply does not have a rubber stamp that says never mind.