The Grassy Knoll: What the Echo in Dealey Plaza Really Was

At 12:30 in the afternoon on 22 November 1963, a presidential motorcade turned off Houston Street and onto Elm, into a shallow bowl of concrete and grass in downtown Dallas called Dealey Plaza. Within about eight seconds it was over. John F. Kennedy was slumped against his wife, a bullet had passed through his neck and another had opened his skull, and the plaza had become the most photographed, filmed, measured and re-measured patch of ground in American history. And almost at once, dozens of the people standing in it — police officers, railway workers, a soldier, a press photographer — turned and ran the same way. Not toward the tall brick building behind the motorcade where the shots were supposed to have come from, but up a low grassy slope to the front and right, toward a white pergola and a wooden picket fence. They were certain the sound had come from there. Their certainty is the seed of everything that followed.

The slope with a name

The grassy knoll is a real place, and a small one. It is the landscaped rise on the north side of Elm Street, crowned by a curved concrete pergola built in the 1930s and backed by a stockade fence at the edge of a car park and the railway yards beyond. From the sixth-floor window of the Texas School Book Depository, where Lee Harvey Oswald’s rifle and three spent cartridge cases were found, the knoll sits diagonally opposite, across the plaza. A shooter there would have been firing at the President from the front and to his right, rather than from behind.

The Warren Commission, reporting in September 1964, concluded that there had been three shots, all fired from that sixth-floor window, all by Oswald, acting alone. It is worth being fair to the Commission: the physical evidence for Oswald firing from the Depository is genuinely strong. His rifle, his palm print, his cartridge cases, his prior possession of the weapon by mail order, his flight and his shooting of a Dallas police officer forty-five minutes later all point one direction. The lone-gunman case is not a fantasy imposed on the public. It is, on the forensic record, the likeliest reading of what happened.

But the Commission was writing over a hole it could not fill, and the hole was the witnesses. The Dallas County Sheriff’s Department took statements from people who had been in the plaza, and a striking share of them located the shots, or at least some of the shots, on the knoll. The photographer on the follow-up car saw a puff of what looked like smoke drift out from the trees on the rise. A railway supervisor named Sam Holland, standing on the overpass above the plaza, was adamant he had seen smoke under the trees near the fence and heard a report from there. A police motorcyclist, Bobby Hargis, riding just to the President’s left rear, was hit by blood and matter with such force that he later said he thought for a moment he himself had been shot. To him the spray came from the wrong direction for a bullet fired from behind.

That is the kernel. The grassy knoll is not a thing invented decades later by men with corkboards and string. It is where a large number of ordinary, sober eyewitnesses, in the first shocked minutes, believed a gun had been fired. Any honest account has to start by conceding that the impression was real, widespread, and immediate.

Why a plaza lies about where a sound comes from

Here is the first fork, and it is a quiet one, because it does not involve a second shooter at all. It involves architecture.

Dealey Plaza is an almost perfect acoustic trap. It is a shallow concrete bowl, open at one end onto a triple underpass, flanked by hard-faced buildings and a colonnade of pergolas, with a railway overpass closing off the far side. A rifle shot in such a space does not travel to the ear as a single clean crack from a single clean direction. It arrives as the muzzle report followed by echoes — off the underpass, off the Depository, off the pergola walls, off the retaining wall on the knoll itself. The human ear localises a sound by tiny differences in when it reaches each side of the head, and in a hard, reverberant bowl those cues can be captured by the loudest echo rather than the original report.

Witnesses on the overpass and near the fence were standing beside large flat reflecting surfaces. It would have been stranger if some of them had not heard the shots seeming to come from close by. As for Sam Holland’s smoke: a steam locomotive stood in the railway yard behind the fence, motorcycle exhaust drifted across the plaza, and it was a cool day. Smoke and haze in that corner is not proof of a rifle; it is what a busy rail yard on a cool afternoon looks like.

None of this proves there was no second gunman. Absence of a good acoustic reason would not summon one, and the presence of a good acoustic reason does not banish one. But it changes the weight of the eyewitness testimony from “dozens of people independently detected a shooter” to “dozens of people had the experience a reverberant plaza would produce.” The impression is honest. The inference from it is where the caution begins.

The tape that convinced Congress

For most of the 1960s and 70s the knoll lived on rumour and eyewitness memory. Then, in the late 1970s, it acquired something the Warren Commission never had — what looked like hard physical data, and an official body willing to stake its name on it.

The House Select Committee on Assassinations was set up in 1976 to re-examine the killings of Kennedy and Martin Luther King. Late in its work, investigators turned to an unglamorous piece of evidence: a Dictabelt recording from the Dallas police radio system. During the motorcade, the transmitter on one police motorcycle appears to have jammed open, broadcasting continuously and recording roughly a minute of open-microphone sound onto the belt. If that motorcycle had been in Dealey Plaza at the moment of the shooting, the theory went, the belt might contain the actual gunfire.

The Committee hired the acoustics firm Bolt Beranek and Newman, who had earlier analysed the Watergate tapes, and then two scientists, Mark Weiss and Ernest Aschkenasy, to refine the work. They compared impulse patterns on the belt against test shots fired in Dealey Plaza in 1978, matching the echo signatures a real shot would leave. Their conclusion, delivered in the Committee’s final days, was startling: the belt appeared to contain four shots, and one of them — with, they estimated, a 95 per cent probability — had been fired from the grassy knoll.

On the strength of that single analysis, the House Select Committee on Assassinations concluded in 1979 that President Kennedy “was probably assassinated as a result of a conspiracy,” and that there had probably been a second gunman on the knoll who missed. For the first time and the only time, an official organ of the United States government had put its name to the grassy knoll. This is the high-water mark of the theory, and it is why the knoll is not a fringe belief but a mainstream one. Millions of people did not talk themselves into it. Congress handed it to them.

The drummer and the minute that did not fit

The fork that matters most came from an amateur, which is part of why it is worth telling honestly. It is a reminder that the correction of an official error can come from anywhere.

Steve Barber was a drummer in Ohio who bought a flexi-disc copy of the Dictabelt audio bound into a magazine and played it, over and over, the way a musician listens to anything. Buried under the crackle he caught a voice. It was faint, bleeding through from another radio channel — “crosstalk,” in the jargon — and it was saying something like “hold everything secure.” Barber recognised it. It was the voice of Sheriff Bill Decker, and the phrase was one Decker had issued after the shooting, when he ordered men into the rail yard and the buildings.

If Decker’s post-shooting order is audible on the same stretch of belt as the supposed gunshots, then the “gunshots” were not recorded at 12:30 as the President was hit. They were recorded roughly a minute later. And impulse noise on a stuck police microphone a minute after the assassination is not gunfire; it is the ordinary static of a busy radio channel and a moving motorcycle.

In 1982 the National Research Council convened a panel of specialists, chaired by the Harvard physicist Norman Ramsey, to review the acoustic evidence. Building on Barber’s crosstalk timing and their own analysis, they found that the Dictabelt impulses did not correspond to the shooting at all, and that the statistical match to a knoll shot was without foundation. The motorcycle whose microphone jammed open was probably not even in Dealey Plaza. Later work reinforced the point: the timing of the belt is inconsistent with the Zapruder film, the frame-by-frame home movie that remains the truest clock of those eight seconds. The one piece of hard physical evidence for a second gunman had dissolved. What Congress endorsed in 1979, science withdrew by 1982 — though the retraction never travelled a fraction as far as the claim.

What the knoll is really about

Strip the acoustics away and a plainer question remains, and it is the one that actually keeps the grassy knoll alive. It is not really a question about echoes. It is a question about proportion.

A President of the United States, at the height of the Cold War, is killed in the middle of the day in front of hundreds of people. And the man arrested for it is a twenty-four-year-old ex-Marine, a defector who came back, a drifter between jobs and cities, who bought his rifle by post for less than the price of a good coat. The cause and the effect do not balance. The scale of the loss demands a cause of matching weight — a plot, an agency, a conspiracy large enough to be worthy of the hole it tore in the century. A lone misfit with a mail-order rifle feels like an insult to the dead. This is the deepest engine of assassination theories everywhere: the mind’s refusal to accept that enormous consequences can have small, shabby, accidental causes.

The grassy knoll answers that refusal. It offers a second gunman, and a second gunman implies coordination, and coordination implies that the murder meant something proportionate to its weight. And the suspicion was not irrational, because the era genuinely earned it. This was the government that really did run MKUltra, dosing unwitting citizens with LSD; the government that a year later would manufacture the Gulf of Tonkin incident to widen a war; the government whose plans, as in Operation Northwoods, showed it was willing to contemplate killing its own people to justify a policy. A public that had been lied to that thoroughly had every reason to disbelieve an official finding, even a correct one. The knoll grows in the soil those real deceptions left behind — the same soil, in the same decade, that produced the durable American conviction that the desert at Roswell hid something the authorities would never admit.

So the honest close is not a verdict stamped on a slope of grass. It is this. The witnesses were not fools; a reverberant plaza told them the truth of their own ears. The theory was not fringe; a committee of Congress once carried it. The evidence was not fabricated; it was analysed by serious people and then corrected by serious people, one of whom was a drummer with a good ear and a cheap record. What survives, after the echo has been traced to the wall it bounced off, is not a killer behind the fence. It is the shape of a country’s grief — the entirely human need for the death of a President to have been the work of something as large as the wound it left.