The Ghost of the Grassy Knoll Gunman

and the Futile Search for Signal in Noise

 

Josiah Thompson and his latest book. (Graphic: © 2021 Dale K. Myers. All Rights Reserved) 

 

By NICHOLAS R. NALLI, Ph.D.

 

Last Second in Dallas

Josiah Thompson

University Press of Kansas. 504 pp., $29.95

Last Second in Dallas (ku.edu)

 

[EDITOR'S NOTE: Click HERE to download a (1.08 MB) ZIP file containing a PDF version of this article, plus the animated GIFs.]

 

Introduction

 

It has been more than 57 years since the shocking and tragic assassination of President John F. Kennedy in Dallas, Texas, on 22 November 1963. Like the 7 December 1941 Pearl Harbor surprise attack before it, and the 11 September 2001 terrorist attack following, the course of history was changed in an instant. The Kennedy assassination occurred at the height of the Cold War, just over a year after the Cuban Missile Crisis, with tensions in Vietnam escalating, and America embroiled in the Civil Rights Movement.

 

It was within this geopolitical context that President Kennedy made a mundane visit to Texas to maintain electoral support in battleground southern states in advance of the 1964 election. The reception he received during his motorcade in Dallas, riding in an open-top convertible, was warm and enthusiastic… until the last minutes of the downtown route, when gunshots rang out in Dealey Plaza. The news media and American public would soon learn that the suspect arrested in connection with the shooting, a resident in right-leaning Dallas, Texas, was in fact a communist sympathizer, a former U.S. Marine who tried to defect to the Soviet Union. This set of seeming contradictions would be the source of cognitive dissonance for people seeking meaning in the tragedy. While the Warren Commission (WC) would establish that this lone individual murdered JFK, they never established any definitive underlying motive. This, among other things, set the stage for the early first and second waves of “WC critics,” “assassination researchers,” “buffs,” and yes, “conspiracy theorists.” Speculation abounded as to “who killed JFK,” much of it the outcome of the geopolitical context—it was the Mafia, the right-wing segregationists, the Cubans, the Russians, the US military-industrial complex and deep state, ad nauseam... anybody but that “silly little communist”. [1]

 

Now, while broad public interest in the assassination has gradually waned since the 1990s, there is no shortage of new authors on the topic. Interest in the event continues to wax annually around the anniversary, and the Dallas Sixth Floor Museum at Dealey Plaza remains a popular attraction.

 

It has thus been with some anticipation that a new book by Josiah “Tink” Thompson, a respected first-wave author, was welcomed by the assassination research community. The book’s title, Last Second in Dallas, suggests a follow-up of his 1967 classic, Six Seconds in Dallas, and indeed, it is part sequel, part autobiography. In Last Second, Thompson takes us on his transition from being a young, idealistic, left-wing college professor of philosophy at Haverford College in 1964 to a private investigator, told through the lens of his pursuit of the Kennedy assassination case. Thompson’s fascination with the Kennedy assassination was driven by his serendipitous association with LIFE Magazine and other early researchers who questioned the Warren Commission’s conclusions. However, he was ahead of the curve on the subject, and justly deserves recognition for his early contributions (beyond the mandate-limited Warren Commission Report). Thompson is unquestionably among the more reasonable of the “WC critics,” but he also has the rare talent of being a persuasive public speaker. He contends that he remains focused on investigation of the crime scene (Dealey Plaza), and not any “conspiracy theory.”

 

As this paper will show, despite what might be high expectations, Last Second in Dallas (LSID) is ultimately a disappointment. Of course, this is not speaking for entrenched “conspiracy theorists” (or “CTs” for short). Although CTs have wide-ranging scenarios, often in stark disagreement with Thompson, many will nevertheless welcome this encore performance from one of the genre's icons, if for no other reason than keeping the JFK conspiracy debate alive. I write this knowing full well that any criticism will be a priori rejected by some CTs, and I also know firsthand of the ad hominem attacks that can be incurred by critics of the “WC critics.” While Thompson himself is generally somewhat more civil and diplomatic than some of the more militant conspiracy-minded ideologues, throughout the book he airs a lingering personal grudge against the late Nobel Prize winning physicist, Prof. Luis Alvarez, a prominent American scientist who earnestly sought to debunk the CTs of his time.

 

In the Foreword we are advised by Richard Rhodes that “you will find no conspiracy theories here,” presumably because there is no specific hypothesis on who the perps were. But this denial notwithstanding, the book falls solidly within the CT genre because a conspiracy is implied, with three unknown shooters conjectured in Dealey Plaza (i.e., adhering to “triangulation of crossfire” and “Grassy Knoll gunman” scenarios). However, outside of the autobiographical content, Thompson’s implicit goal is not so much as to build a tenable counter-scenario (explaining the who, why and how), but rather deconstruct the well-established “official” scenario. But it is Thompson’s collateral attempt to deconstruct his own legitimate groundbreaking contributions to the case in his original book (which were heretofore generally respected) that I find particularly disappointing about this new book.

 

LSID does not attempt to provide significant new research of his own not included in Six Seconds in Dallas, but instead attempts to invalidate one of the key conclusions (discussed in detail below). This is presumably because this earlier finding, under closer examination, has since inadvertently pointed to an opposite conclusion. This is something not too uncommon in science, and it can often lead to big breakthroughs, but only if the investigator is objective and detached enough to accept the unexpected conclusion at face value. Otherwise, it will eventually fall on others to do so... because facts are stubborn things.

 

The book is divided into four chronological parts spanning the author’s career investigating the case: Part I, “Beginnings—The 1960s”; Part II, “Aftermath—The 1970s”; Part III, “Breaking the Impasse—The 2000s”; and Part IV, “The Signal in the Noise—2013-2017.” Parts I and II are more autobiographical and retrospective in nature, whereas Parts III and IV focus on the author’s renewed interest in the early 2000s along with a subsequent new interpretation of his previous work.

 

Part I: Beginnings—The 1960s

 

The first part consists of nine chapters where Thompson discusses his early sleuthing on the case during the turbulent 1960s, walking the reader through his journey up through the publication of his first book. These chapters are more biographical in nature and thus make for easy reading, basically telling us the story behind his original bestseller that would gain national attention. Thompson’s narrative in pursuit of conspiratorial mysteries will connect with passionate students of the Kennedy assassination. Readers interested in the case will feel the excitement he must have felt, while working for LIFE magazine.

 

Working with various connections he made at LIFE, Thompson was able to gain access to interview key witnesses, many of whom he admits did not lead to any “breakthroughs.” Thompson focused on Dealey Plaza witnesses who were in a position to see the area of the Grassy Knoll during the assassination. Some of these had been previously interviewed by the Warren Commission or given depositions to law enforcement, and Thompson sought to reconcile testimony that appeared to conflict with the WC version of events. Among these were S. M. Holland, a railroad supervisor who witnessed activity behind the stockade fence overlooking the Knoll around the time of the assassination. Holland was detailed about what he saw and heard, especially what appeared to be two gunshots, along with a “puff of smoke” around the stockade fence. He ran to the location with two other rail yard workers and found numerous footprints and cigarette butts around the fence and a parked car, agreeing with what Lee Bowers saw from the switching tower. Certainly, someone was there just before the assassination. Thompson concluded that it would have been possible for a gunman to stow the murder weapon in the trunk of the car, then drive away later without detection. While this is a reasonable hypothesis for how any would-be gunman could have gotten away, it still implies a plan with a high risk of being caught.

 

While Thompson does acknowledge the limitations of witness testimony, he still uses them to support his conclusion on the location of a Grassy Knoll gunman. Thompson also recounts other anomalies that he encountered, including conflicting testimonies and questions about physical evidence, for example, the discovery of the evidence bullet (CE399) on a stretcher at Parkland Hospital, and the three expended hulls found in the sniper’s nest. Although these may have been legitimate questions at the time leading to his initial speculative “reconstruction” (“four shots from three locations in just under six seconds”), most have since been resolved. Nevertheless, their repetition still goes a long way toward sowing doubt in the minds of the “CT inclined.”

 

It was through his LIFE connections that Thompson became among the first to see and study high-resolution frames of the Zapruder Film, the now infamous 8 mm silent home movie of the assassination sequence. For several serendipitous reasons, this film of the assassination is nothing short of an evidentiary miracle. While it might appear somewhat low resolution and grainy by today’s standards (and it’s silent), it was state-of-the-art for its time, using color film optimized for outdoor sunlight conditions, possessing a large optical zoom lens, and filmed from a vantage point that could not have been better chosen.

 

Within Thompson’s autobiography, it was these early years that would lead to what is considered by many to be the best pro-conspiracy book to have been published, Six Seconds in Dallas. Thompson would achieve national fame, appearing on television and radio as an expert WC critic on the case. Unlike many of the first and second wave assassination buffs, who would spend time speculating on grander and grander conspiracy schemes, Thompson was focused on the Dealey Plaza crime scene, primarily areas where the WC Report findings seemed contrary to intuition. This naturally included the “single bullet theory” (i.e., a single bullet hitting both Kennedy and Gov. John Connally), but it was the fatal shot (striking Kennedy in the head, near-instantly killing him) that captured most of his attention. No one can deny that this horrific event captured at Zapruder Frame #313 (Z313) is the “climax” (as Thompson aptly puts it) of the 26-second film.

 

Indeed, I remember when I first saw it in 1975, not yet 10 years old, lying on the floor in front of the TV and hearing the gasps of my parents and older brothers behind me, not knowing fully what I was seeing. (Believe it or not, I had thought that Mrs. Kennedy’s climbing onto the trunk of the limo was President Kennedy being thrown back). It was seeing the sequence on TV as an impressionable kid that initially sparked my own interest. In recent years, I have given a handful of lectures on this topic before academic audiences, and I have asked for a show of hands from those who have not seen the film before showing it. Amazingly, a significant percentage of hands have gone up. When I proceeded to show the film (after due warning), I was nearly brought to tears by the renewed gasps in the room, feeling guilty for having exposed them to such a sad, tragic event, one of the darkest moments in American history.

 

Pertaining to the fatal shot, Thompson, true to the early buffs, continues to be fixated on the “back and the left” motion of the stricken President. Seeing the event at actual speed from Zapruder’s viewpoint, the untrained eye might get the initial impression (especially if suggested beforehand) that President Kennedy is being shot from the front. This seemed to confirm the early CBS reports that afternoon from witnesses who thought the shots came from a “grassy hill” overlooking the motorcade, which would eventually be coined the “Grassy Knoll.” These early reports (which also said that the Secret Service found a “huddled man and woman” on the top of “a grassy hill” who may have been the “would-be assassins”) were based on a handful of eyewitnesses who were closest to the limo (and who thus had the impression that the fatal shot must have come from nearby). But first impressions can often be wrong, and if there is one thing I learned from my career and education, it is that the physical world can be counterintuitive.

 

Consider sound waves, for example. Because sound is a wave phenomenon, identifying the source of a sound (e.g., a gunshot) is not always straightforward, as most of us know from experience. Sound waves echo off hard solid objects, are muffled by others, and, perhaps most importantly, they also bend around objects. Although we are aware of these properties of sound, they nevertheless can sometimes confound the ability to pinpoint accurately the source. This especially applies to a gunshot (more precisely, the muzzle blast) in a complex urban environment, not to mention the additional confusion from the shockwave (i.e., the “ballistic crack”) caused by a supersonic bullet.[2] Interestingly, Thompson seems to understand this and correctly points out that the earwitness testimony (suggesting the shot came from the Knoll) is not necessarily reliable.

 

Thompson understands the primacy of evidence in the case, and that one cannot willy-nilly throw out evidence one does not like. Thompson is quite capable himself of thoroughly debunking the more fringe “conspiracy theories,” such as the preposterous “Zapruder Film alteration” that has metastasized over the conspiracy blogosphere and social media. Zapruder Film alterationists contend that the film was “altered,” an example of taking layers and layers of ad hoc assumptions to their extreme, self-contradictory conclusions.

 

Here Thompson is correct, of course: The extant Zapruder Film is authentic photographic evidence of the crime.[3] It has been confirmed authentic by photographic experts, for example, noted Kodak film expert Roland Zavada., who has unambiguously asserted that “there is no detectable evidence of manipulation or image alteration on the Zapruder in-camera-original [film] and all supporting evidence precludes any forgery thereto”.[4]

 

The technology simply did not exist in 1963 to make undetectable alterations to 8 mm reversal film, and even today, it would not be trivial. We are not talking about using Photoshop to alter an electronic JPEG image—we are talking about making physical alterations to the extant physical film itself. From a common-sense perspective, when one considers, for example, that the image of JFK’s head on the film is about the size of the point of a pencil, and that there would not be a second chance to “Undo (Ctrl+Z)” a mistake, one might imagine just how far-fetched this scheme would be. Sorry, but no. If would-be conspirators (operating in the real world) wanted to conceal damning evidence in the Zapruder Film, they would have simply sought to destroy it, had there been an opportunity. But ultimately, there never was an opportunity for them to get their hands on it, as David Wrone, Thompson and others have documented in the chain-of-custody. It should go without saying that alteration would also require “back-engineering” the entire assassination sequence so that the alterations would conform to a coherent, predetermined assassination cover story.

 

But even with these independent lines of proof, from a purely philosophical point-of-view, objection to “alterationist” nonsense has an even more fundamental underpinning: If the Zapruder Film was “altered,” then how do we even know that JFK was assassinated? What if the whole thing was an orchestrated hoax, like the “Moon Landing Hoax”? And before somebody objects “but the media reported it,” “the witnesses and ER doctors saw it,” “they performed an autopsy,” “they buried him in the ground in Arlington Cemetery during a state funeral,” etc. — sorry, but that train has left the station. Once we start denying key pieces of evidence (e.g., the Zapruder Film, autopsy materials, physical evidence, etc.) without ironclad proof (i.e., other more fundamental evidence), we can then deny all the evidence on the same grounds.

 

And once we have descended to that point, we no longer have any basis for investigation or argument, nor any basis for “believing it”—we are simply wasting our time debating how many angels can dance on the head of a pin.

 

Alas, the Zapruder Film is indeed authentic (as is the other primary evidence) and it shows precisely (and gruesomely) how the 35th President was brutally shot to death in Dealey Plaza. On these points, Thompson and I are in full agreement. Nonetheless, as will be seen, this discussion of the Zapruder Film is germane to what we find later in the book.

 

On Shooting Melons... or Coconuts… or Reasonable Facsimiles 

 

In the concluding chapter of Part I Thompson reveals the origins of what would be a lifelong (and apparently bitter) feud with physicist Luis Alvarez that will manifest itself throughout the remainder of the book. Fellow scientists, engineers and mathematicians out there will understand what a profound achievement it is to receive the Nobel Prize in physics, as Alvarez did in 1968.

 

Like many of us, Alvarez would take keen interest in the case. For him it was apparently kindled by the November 1966 LIFE magazine article, titled “A Matter of Reasonable Doubt,” which Thompson had contributed to and which featured large blown-up full-color frames from the Zapruder film. Alvarez was initially interested in the effect of camera panning error in creating bright streaks from shiny reflective surfaces such as the chrome window frame, hypothesizing that these frames could be indicators of startle responses Zapruder may have shown as the result of gunshots. Based on his quantitative “jiggle” analysis, Alvarez concluded that there were only three shots fired in Dealey Plaza (and he would go on to assert that there was no conspiracy).

 

Given his status, Alvarez was given airtime on CBS News (hosted by the most trusted man in America, Walter Cronkite) to make his case in defense of the Warren Commission. Thompson subsequently came to his own conclusion that Alvarez was not forthcoming about his “jiggle analysis” and had overreached on the conclusion of three shots fired, given that there were more than three “jiggles” evident in the film, which, in all fairness, Thompson had a point on. (Alvarez handwaved that one of them was caused by Zapruder’s startle response to the Secret Service siren.) He wrote Alvarez a letter that by his own account was “sloppy” and “decidedly lacking in tact.” If it were anyone else, Alvarez probably would have dismissed the letter as coming from a nut and tossed it in the proverbial recycle bin (which is what Thompson apparently expected). But Alvarez felt compelled to respond in kind, albeit as if he were swatting an annoying fly.

 

But Thompson would find out from Paul Hoch, then a graduate student under Prof. Alvarez (and a well-respected first-generation assassination researcher), that he apparently had gotten under Alvarez’s skin more than the physicist let on. One gets the sense that Thompson felt a certain sense of satisfaction, given his description of the entire “scuffle” with the Nobel Laureate as being “comical,” and Hoch ultimately acknowledged that Thompson was probably right about the Secret Service siren. Undoubtedly feeling emboldened in “the brash confidence of youth” and Hoch’s siding with him, Thompson then decided to “turn the knife” by sending Alvarez another letter, this time carefully typed and gloating over the fact that Alvarez’s associate Paul Hoch had agreed with a “cheeky assistant professor of philosophy” over the internationally recognized physicist.

 

But it wasn’t Hoch’s opinion that was the “knife turner,” but rather Thompson’s not-so-subtle insinuation that Alvarez had “cherry-picked” his data, a decidedly unethical and unprofessional practice in science. I was dumbfounded when I read this, and I can only empathize with how Alvarez might not have taken too kindly to the gall in the accusation. But rather than toss it into the recycle bin with the first letter, breathe, count to ten, and then let it go, he felt the need to respond, basically explaining (correctly) that one of the main things a scientist does is “to decide which of the evidence he sees must be wrong and must be ignored.” Or, in other words, to distinguish meaningful “signal” from irrelevant “noise”. [5]

 

While he tries to brush off his encounters with Alvarez as somehow “funny,” Thompson obviously did not like Alvarez’s dismissive tone in his replies. But I would have informed him that such a response, if any, would not be atypical from a physics professor, and that it would be best not to take it personally. It may have been one thing for Thompson to challenge Alvarez on his own turf as a philosophy professor, but there is a certain level of presumptuousness in advising a Nobel Laureate in physics on the topic of science.

 

However, dirty laundry aside, we are told that the private communications with Alvarez establish “a context for understanding Alvarez’s thinking and methodology,” and unfortunately, we are not talking about science here, but rather his establishment worldview. At best, this tactic is ad hominem, and at worst, it’s gossipy. But Thompson’s fixation on politics evident throughout this book is telling in its own right. Yes, scientists and investigators may have private political opinions, but the establishment of facts in a crime has nothing to do with “politics.” That is the entire basis behind the scientific method and our post-Enlightenment justice system. Yet the real beef comes with Alvarez’s “second, and far more influential, theory,” namely the “jet effect,” the explanation that Alvarez proposed to explain the “back and to the left” motion of President Kennedy, and thus a direct challenge to Thompson’s book Six Seconds in Dallas.

 

My own backstory on this topic is relevant since many buffs treat the “jet effect” as an explanation conjured up by Alvarez. I had never heard of Luis Alvarez (or the term “jet effect”), until about ten years ago, when, as an anonymous commenter on various blogs, I was offering up my own (at that point qualitative) “theory” that JFK’s “back and to the left” motion was due to a recoil effect (aka, “jet effect”). At some point, another anonymous blogger brought to my attention a physicist named Alvarez and his 1976 paper on the “jet effect.” Up until that point I had read the WC Report (inspired by seeing Oliver Stone’s movie JFK), but I had read little else on the case due to the pressing demands of graduate school and early career, so at the time I was unaware of post-WC developments.

 

Some 15 years earlier I was in graduate school at the University of Wisconsin-Madison when it first independently dawned on me that the explosive wound seen in the Zapruder Film would cause a recoil of JFK’s head—and the more I thought about it, the surer I became of it. I remember an incident in a classical mechanics class when we were studying inelastic collisions. Someone asked a question about gunshots and the TA said that a body struck by a bullet would move very slightly in the direction of the bullet as an inelastic collision. Although I was normally a quiet student, I spoke up saying that no, the bullet would actually pass through and create a much larger exit wound on the other side, and this would cause the body to move toward the shooter… The TA, surprised by this response, pondered it for a moment, and then thoughtfully said, “hmm, that’s true...” And the professor, uncomfortable with the gory direction the discussion had taken, cleared his voice, interjected something to the effect of “okay, that’s enough of that subject...” and then moved us quickly on to the rest of his lecture.

 

Of course, I have since learned that gunshot wounds are a bit more complicated than that, and that “jet effects” from such wounds are probably exceptions to the rule. But the point I wanted to empathize here is that the very same idea came to me without ever hearing of Alvarez. That is because the “idea” is not really about him or me—it’s about how the physical world behaves. Upon learning that he published this idea a good 20 years before I “discovered” it, I was disappointed, for I honestly thought I had stumbled upon a unique and original idea. But I was also comforted in the realizations that (1) I was not out in left field with my physical intuition, (2) Alvarez’s approach was pedagogical in nature and thus there was room for a more detailed independent analysis, and (3) in the interim, “CTs” had gotten comfortable in rejecting Alvarez as a some sort of one-off “government shill,” and thus there would be ample rationale in revisiting it anew (which I eventually got around to publishing in the journal Heliyon in 2018 [6]).

 

Alvarez arrived at his “jet effect” theory after obtaining a copy of Six Seconds from Paul Hoch. He studied the figures showing the change in position of JFK’s head and came to the realization, as I did, that there must have been a real force [7] at play in driving JFK’s body backward in the limousine. But he was not convinced of Thompson’s “near simultaneous gunshot” hypothesis. Given it was a journal directed to physics educators, Alvarez relates how he worked out the essence of his theory on “the back of an envelope” while on conference travel. This is not atypical for a physicist to keep things as simple as possible, and using simple conservation of energy and momentum arguments, he was able to deduce that a recoil of a sufficient magnitude would be theoretically possible given certain assumptions.

 

As a professional colleague, Hoch recommended that Alvarez perform some sort of experimental test that “could demonstrate the retrograde recoil on a rifle range, using a reasonable facsimile of a human head.” They proceeded to carry out such experiments, trying different targets, but ultimately deciding upon a taped melon as the best “facsimile.” Hoch noted that the melons consistently exhibited a “retrograde motion” toward the shooter, and Alvarez thus was able to demonstrate that a recoil effect is indeed possible.

 

Through his correspondence with Paul Hoch, who shared a write-up of the melon experiments with other researchers, Thompson would find a beachhead to launch another “cherry picking” accusation. Hoch simply noted, legitimately once again, the limitations of the melon experiments. Among other things, the shell of a melon is much softer than a skull, and it will thus present much less resistance to the projectile passage. The upshot of this is that the bullet would transfer practically none of its forward momentum to the target, and thus there would be no forward impulse (discussed more below) for a “jet effect” to overcome. While this is a legitimate point, it misses the point of the experiments. In Hoch’s own words: “It should be emphasized that we did not attempt to simulate the conditions of the assassination, but only to show that a recoil towards the gun is physically possible…” Q.E.D.

 

During the course of the “melon experiments,” Hoch also observed the following: “Although taped melons consistently showed the ‘rocket’ effect, different kinds of targets reacted differently.” I would also venture to add that identical targets would react differently depending on where precisely they were hit, as well with different firearms and ammunition. Thus, the point of the “melon experiments” is not to demonstrate that recoil effects are a universally observed behavior of “head shots,” but rather that a bullet can produce such an effect in a “reasonable facsimile.” Of course, Thompson disputes this, claiming that “Hoch’s point is obvious. A taped melon is in no way ‘a reasonable facsimile of a human head’.” Thompson is free to have his opinion about what a “reasonable facsimile” is, but Hoch’s point was anything but. According to him: “We do feel a taped melon is not an unreasonable simulation of a person's head…” [emphasis mine] 

 

Thompson uses the melon experiments along with a perceived “political bias” (which is always something defined relative to one’s own bias) as a cudgel to attack Alvarez and accuse him of “cherry picking” data because not all targets behave as the melons did. But that has never been the point. The point was to show that it was quite possible, thereby providing a physically plausible counter-explanation.

 

But regardless of Alvarez’s melon experiments, or subsequent experiments conducted independently by Dr. John Lattimer in the 1970s and more recently by ballistics expert Luke Haag, a recoil effect of some magnitude had to happen (following the initial forward impulse) given the directional mass expulsion observed in Z313, even if it were the only known case of the phenomenon, because it is an inescapable consequence of momentum conservation. The only question is, could it have been sufficient enough to cause “retrograde motion” by overcoming the initial 5 cm (2”) forward impulse brought on by the transfer of momentum by the bullet to the head? And the quantitative answer to that question is arguably “yes”. [8]

 

One gets the impression that Thompson never got over the fact that Luis Alvarez, undoubtedly one of the greatest scientific minds of his time, would publish a paper in a reputable journal that would offer a plausible scientific counter-explanation for JFK’s “rearward lurch.” But notably, while Alvarez makes extensive reference to Thompson’s book in his paper, at no point does he impugn Thompson’s integrity or work. Indeed, when I eventually read Alvarez’s paper, I never got the slightest inkling that there was any bad blood between these two men. Instead, we have Alvarez saying things like “[Thompson’s book Six Seconds in Dallas] is beautifully printed, with excellent photographs and carefully prepared graphs.” I had expected that Thompson would have since changed his mind on his original conclusions given what we have since learned.

 

But alas, as Dale Carnegie once said, “a man convinced against his will, is of the same opinion still.”

 

Part II: Aftermath—The 1970s

 

In Part II, Thompson focuses on developments during the 1970s following the publication of his book, including his meeting Robert Groden and his role in the unauthorized showing of the Zapruder Film on ABC’s Good Night America program (hosted by Geraldo Rivera and aired on 6 March 1975). The Zapruder Film’s airing on national television was a game-changer event, bringing the JFK “conspiracy theory” to many Americans for the first time, influencing public opinion about the crime, and leading to the House Select Committee on Assassinations (HSCA) government reinvestigation of the case in 1976.

 

The HSCA “Acoustics Evidence”

 

It was around this time that Thompson had decided that being a college professor of philosophy wasn’t really for him. As he was becoming disillusioned with his academic career path, Thompson coincidentally found himself invited by the HSCA to join a group of authors (primarily WC critics) gathered in Washington, DC to provide guidance to Chief Counsel Robert Blakey. Thompson then recounts the origins of what would become a decades-long pursuit, referred to by some as “acoustics evidence” for a gunman on the Grassy Knoll.

 

The acoustics evidence in question refers to recordings of the radio transmissions of the Dallas Police Department (DPD) that occurred before, during and after the motorcade. The group of WC critics summoned by the HSCA had been discussing possible conspiracy scenarios (which, surprisingly, Thompson admits to having no interest in, even nodding off at one point), when the conversation turned to the Dallas police, at which point Mary Ferrell as an aside brought up the “police tapes,” and that apparently there were some out there (specifically Gary Mack) saying that one could “hear” 7 to 8 gunshots on them. And yes, the original claim was that one could actually hear the gunshots on the recordings. This is a relevant point to keep in mind as we proceed in the discussion of this topic.

 

At this point, a brief explanation is in order for those unfamiliar with the DPD radio recordings (as I was, not too long back). Although this topic was something that I had been vaguely aware of from my earlier literature review of the Kennedy assassination (e.g., the book by Larry Sabato, along with Vincent Bugliosi’s treatise, Reclaiming History), it was the claims in Last Second that forced me to dive in more, and it is quite an onion to unpeel. In 1963, the Dallas Police routinely recorded the radio transmissions of their police force. They broadcast over two FM channels, the first channel (Channel 1) being the normal channel used for day-to-day activities, the second being a supplementary channel (Channel 2) for special events such as the Kennedy motorcade. There were separate recording systems for each channel, a Dictabelt and a Gray Audograph for Channels 1 and 2, respectively. The peculiar characteristics of these two devices become important in the analysis of the recordings, notably an automatic sound activation so that the devices would record only when there was a broadcast, switching off during periods of radio silence.

 

The police radios operated in typical fashion: A police officer would press a button on the microphone to “talk” (transmit his voice) and release the button to “receive” (to hear other broadcasts on the channel). During the time period of the motorcade, the current times would be broadcast occasionally by the dispatchers to annotate the time on the recordings. The radio system also featured an automatic gain control (AGC) whereby loud sounds would be dampened (reduced) and soft sounds would be amplified, thereby maintaining a relatively constant volume (amplitude).

 

It turns out that it was not uncommon for microphone buttons to get stuck in the “on” position, and as luck would have it this happened to one of the motorcycle police officers, which led to a continuous broadcast for a period of about five minutes on Channel 1. Although Channel 2 was the channel for motorcade communications, some of the buffs hypothesized that the open microphone may have originated from a motorcycle in the motorcade located in Dealey Plaza (for some reason transmitting on Channel 1 in violation of protocol) at the time of the shooting, thereby picking up the gunshots. This led the HSCA to pursue the matter further, and one of the investigators was able to obtain “original tapes” from a former DPD assistant chief. These were then submitted for analysis, first to a team led by Dr. James Barger, a scientist with Bolt, Beranek and Newman (BBN), then on to Prof. Mark R. Weiss and Ernest Aschkenasy (W&A) at CUNY, all three being experts in acoustics.

 

The analyses that BBN and W&A performed involved matching waveform impulse patterns obtained from the Channel 1 open mic recording to test patterns (including echoes) obtained from experimental (BBN) and simulated (W&A) setups. The experimental setup by Barger’s BBN team was conducted onsite in Dealey Plaza, with an array of microphones, and gunshots (rifle and handgun) fired from two locations, the Texas School Book Depository (TSBD) Sniper’s Nest and the Grassy Knoll. Barger found a match to the DPD Channel 1 recording with the Grassy Knoll test shot with a 50% statistical confidence. But W&A found a simulated Grassy Knoll waveform match (with 95% confidence) from manual calculations based on a model of Dealey Plaza (N.B., the model was not a computer model, but rather used a hard-copy map of Dealey Plaza). Specifically, the match with the Grassy Knoll was found for a simulated open mic located at a specific location near the intersection of Houston and Elm.

 

It was the latter finding that led the HSCA to an eleventh-hour amendment to its original conclusion of a lone gunman (affirming the WC Report), to include a second gunman on the Grassy Knoll (at the location that Thompson identified previously based upon his interviews with eyewitness S. M. Holland) who fired one shot that missed the limo completely, thus concluding that President Kennedy was “probably assassinated as a result of a conspiracy.” So, while the HSCA concluded “probable conspiracy,” their scenario was based upon the acoustics evidence and was at odds with Thompson’s (and others’) scenario of a frontal impact to JFK’s head. Thompson relays to us that the HSCA’s conclusion was constrained by independent findings from Neutron Activation Analysis (NAA) that was performed by Dr. Vincent Guinn on the bullet fragments from the crime scene. Guinn had determined that it was “highly probable” that the fragments in Gov. Connally’s wrist were from the “stretcher bullet” (CE399) found at Parkland Hospital, and that the fragments from President Kennedy’s head were from the same bullet as the fragments found in the limousine, thereby providing strong evidence that only two bullets caused all the wounds.

 

In his book, Thompson does a decent job of accurately relaying all this background information to the reader. The DPD radio recording system, with its primitive recording media (Dictabelt and Audograph), automatic shutoff and AGC features, was designed for recording and monitoring routine police radio voice communications. The system was designed for recording spoken radio communications over the course of an entire day, not for high-fidelity forensic sound reproduction of a 10-second crime sequence during an outdoor motorcade.

 

As an analogy, one might think of security camera video footage, which is generally poor resolution, black-and-white, and wide angle—a setup optimized for routine monitoring for criminal activity over the course of an entire day, but not the positive identification of perpetrators. The same goes for the DPD radio Dictabelt and Audograph audio recordings, which, beyond the recorded content of spoken words, are considerably more primitive in terms of information content by comparison. The BBN and W&A teams nevertheless argued that the Channel 1 open mic recording could be used to identify gunshots in Dealey Plaza.

 

Parts III and IV

 

III: Breaking the Impasse—The 2000s

IV: The Signal in the Noise—2013–2017

 

The latter two parts of the book mark the transition from the author’s retrospective account of his involvement in the case (spanning his work up through the 1970s and the HSCA) to his subsequent renewed interest beginning in the mid-2000s, which is traced back to the 1982 National Research Council (NRC) Ad Hoc Committee on Ballistic Acoustics, along with a subsequent peer-reviewed 2001 paper published by Dr. Donald Thomas [9] in the UK journal, Science & Justice.

 

The prestigious NRC panel was commissioned to examine the HSCA findings involving the DPD recordings and included a number of recognized experts, including Luis Alvarez. Thompson spends an entire chapter in Part IV (Chapter 17) portraying what might be characterized as his ultimate triumph against the late physicist-nemesis, whose work had left him in a “funk” in 2013 (when Last Second was originally planned to be published) because of the debunking of the acoustics evidence by the NRC panel. Of course, Thompson then proceeds to impugn the ethical character of Alvarez as well as the entire panel, calling it “unvarnished advocacy,” and insinuating in Chapter 20 that they were guilty of criminal malfeasance. This sets the stage for later in Part IV, where Thompson then argues that the “acoustics evidence” has been vindicated despite the NRC conclusions.

 

By this time, Thompson had established his own Private Investigator practice, the days of being a professor long behind him. The second half of the book leads the reader to Thompson’s revised scenario of the crime, which, like the original scenario, involves an unnamed conspiracy with an unknown gunman firing from the Grassy Knoll and hitting Kennedy in the head at Z313. However, the novelty of Thompson’s revised scenario is his claim that there was a second shot to the head (from the rear) at Z327–Z328, following a frontal shot at Z312–Z313; this scenario is based upon four revised “puzzle pieces”. [10] Two previously accepted puzzle pieces (evidence) are tossed out of the box and replaced by two new pieces that conveniently fit Thompson’s puzzle as follows:

1. Tossed out: The NAA findings that only two bullets caused the wounds of both Kennedy and Connally.
2. Tossed out: The observed forward snap of Kennedy’s head in Z312–Z313 (indicative of a shot from behind) due to motion blurring of Z313 (i.e., a “reinterpretation” of these frames as a “optical illusion”).
3. New piece: A reinterpretation of the 1967 Zapruder Film measurements as evidence for a late Z327-Z328 “head shot” along with an attendant reinterpretation of nearby eyewitness testimony.
4. New piece: A resurrected interpretation of the DPD recordings as “acoustics evidence” of two shots to the head in “the last second” of the Zapruder Film sequence.

On its surface, the high-level approach of Thompson’s presentation has a certain allure given its symmetry and the metaphor of a puzzle. I too have considered the solution of this case as analogous to the solution of a puzzle, whereby the full picture comes together not by any single piece (of evidence), but rather the aggregate whole of all the pieces.

 

Other than the questioning of the NAA findings, those pieces crumble upon closer scrutiny. We will consider these points individually, with more attention given to the “blur illusion” (#2), discussed in the last section. [11]

 

The first puzzle piece to be tossed (#1) is the NAA finding that there was physical evidence of two, and only two, bullets found at the crime scene. There has apparently been some degree of legitimate dispute about the NAA findings of Guinn. However, counterarguments have since been advanced from forensic experts such as Larry Sturdivan (cf. The JFK Myths) and Luke Haag. [12] Lacking personal expertise, I shall remain, for the time being, agnostic on Guinn’s findings. Sturdivan and Haag are not to be easily dismissed; Haag has indicated that he has been looking into modern methods such as Multiple Collector-Inductively Coupled Plasma-Mass Spectrometry (MC-ICPMS), which he wrote: “goes far beyond what Dr. Guinn, or anyone else using NAA in the 1960s and 1970s” would have been able to achieve. Thus, I would not consider it justifiable for this “puzzle piece” to be tossed just yet. The case for a lone gunman does not hinge on this evidence, and despite the uncertainty surrounding the only-two-bullets conclusion, no positive NAA evidence of a third bullet has been found.

 

A Rear Shot to the Head at Z327–Z328

 

The second puzzle piece to be tossed is the one-frame Z312–Z313 forward snap of President Kennedy’s head. However, the first “new” puzzle piece that Thompson introduces (#3) replaces it, namely a reinterpretation of Thompson’s original (Thompson-Hoffman) 1967 Zapruder Film measurements. Thompson credits this “reinterpretation” to “another brilliant non-professional” named Keith Fitzgerald, who pointed out to him at a meeting in 2005 that “according to [Thompson’s original] measurements, JFK’s head moved forward 6.44 inches between frames 327 and 330,” and that apparently the head wound changes during this timeframe. When I first read this, I wondered what they could be talking about, for I did not recall seeing anything like this, either in the film or in the Thompson-Hoffman data. As Thompson himself says in the book, the “climax” of the film is the fatal shot at Z313. I do not recall any prior suggestion of another gunshot wound to the head occurring after Z313.

 

One good reason for this, of course, is that when one views the film at full speed, one simply does not see another gunshot effect after Z313. We clearly see a shot hitting both Kennedy and Connally at the same time around Z224–Z225, then the fatal shot at Z313, but there are no other indicators of gunfire wounding the limo occupants. It’s really that simple.

 

The HSCA had already considered this scenario of “the fourth shot” from the Texas School Book Depository (TSBD) sniper’s nest (based on the acoustics evidence) striking Kennedy in the head at Z327–Z328. However, they concluded, based in part upon a preliminary trajectory analysis, that “it is highly unlikely that the head wounds were inflicted by firing a bullet from the [TSBD] southeast window [sniper’s nest] that impacted [JFK’s head] at the time of Zapruder frame 327.”

 

The frames in question occur very shortly after the Z313 kill shot and Kennedy has already rebounded off the back seat after a neuromuscular spasm. His body has gone visibly limp and has slumped over sideways onto Mrs. Kennedy, who, understandably, has just begun to react in terror. When discussing these details, one should keep in mind that all this happened in under one second (indeed, this is the raison d'être for the eponymous “Last Second in Dallas”). The President’s head had been shattered by a temporary cavitation wound produced by the passage of a high-energy projectile, and its “contents” were still pulsing out due to restoring force oscillations. As Marilyn Sitzman (who was standing with Zapruder) pointed out (and Thompson quotes): “And we could see his brains come out, you know, his head opening.” Or as witness Bill Newman told Thompson in his interview with him: “When he was hit, I was looking right at him, and of course immediately I knew he was shot… then blood, gook just started bulging out...” and Thompson himself confirms: “What [Newman] was describing was exactly what the film shows.” Mrs. Kennedy, in shock, instinctively tried to climb out of the limo before Secret Service Agent Clint Hill got her safely back into the car.

 

Thompson bases this conjectured late shot on his 1967 data, which shows a multiframe “forward lurch” following the multiframe “rearward lurch,” along with “changes” to the head wound during these frames that he tries to illustrate with annotated high-resolution color prints of the frames. As alluded to above, the observed changes to the head wound are the result of the lingering effects of Z313, which occurred less than one second prior. The skull contents were literally still undulating due to the temporary cavitation wound and were thus pulsing out and falling to the floor under the force of gravity. Another gunshot wound at this point would exhibit a one-frame anomaly or impulse of some sort. Thompson claims “any subsequent impact would cause no explosion” because “the pressure vessel” of JFK’s head “no longer exists” (after Z313), but this is simply an incorrect understanding of the mechanics. The “explosion” is not due to the static pressure of the brain inside of the skull cavity (as in a pressurized cylinder of gas)—it is due to a transient pressure wave generated within the tissue by the deposit of kinetic energy (KE) from a high-speed projectile disturbance, somewhat analogous to a rock thrown into a pond.

 

Yes, another “explosion” would in fact happen from a high-speed bullet, provided that the bullet deposits its KE during passage. Thompson suggests that JFK’s aggregate forward motion over these multiple frames is due to such a projectile, but like the earlier “rearward lurch,” both the timescales and the magnitudes are nowhere near consistent with the effect of a high-speed bullet. But let us assume for sake of argument that in one of the frames there is a forward head-snap (occurring on just the head, mind you, not the entire body) of the same magnitude as Z312–Z313 (that isn’t a “blur illusion,” mind you… stay tuned). That would mean that the projectile deposited its momentum, which then means that it also deposited its KE. If it were a high-powered rifle then the KE deposit would be large, thereby creating another temporary cavitation pulse, and thus another “explosion.” On the other hand, if the projectile were to pass through meeting little resistance without causing noticeable temporary cavitation damage, that would mean that it did not deposit any appreciable KE, which in turn would mean that it did not deposit any momentum, and thus no “forward snap.”

 

Getting back to the actual Thompson-Hoffman data: When looking at any type of data, one must always bear in mind what exactly it represents, and in this case, it is simply the frame-by-frame position of the back of the President's head relative to the top of the back seat. Around the time of the fatal shot (Z313), JFK is still seated upright (albeit leaning slightly forward and to his left) and positioned lateral to Zapruder’s field-of-view. Further, there is nothing interfering with the free motion of his head on his neck. These circumstances fortuitously allow one to perform a kinematic analysis of his head in these frames (along with the top of his shoulders, as was performed by Itek Corp). By the time we get to the later frames in question (Z327–Z330), the President has gone completely limp and is slumping over sideways away from the camera onto his wife, who herself is in the process of reacting. JFK’s motion during these frames is not due to a near-instantaneous interaction of his head with a high-speed projectile, but rather the interaction of his lifeless body with the backseat, Mrs. Kennedy, and the force of gravity.

 

Figure 1 includes the Six Seconds data, as shown there, with the addition of Z330 and graphs of velocity and acceleration. There is no indication of a shot at Z327–Z328. The forward motion of “6.44 inches between frames 327 and 330” is part of a smooth curve already in progress (Fig. 1a). This multi-frame forward movement was initiated at Z322 by the rebound of JFK’s lifeless body off the rear seat (after a neuromuscular spasm), followed by his falling limp onto his wife.

 

 

Figure 1 - Motion of President Kennedy’s head (blue) relative to the back of the limo seat from the Zapruder Film (frames Z301 to Z330) based on data from Thompson’s Six Seconds in Dallas and plotted vertically in the same manner as in the original figure (op. cit.): (left) position of the President’s head relative to the back of the seat, (middle) first-order change in position, and (right) second-order change in position. The left y-axes denote Zapruder Frame number and the far-right y-axis denotes elapsed time in seconds (rounded to the 2nd decimal place). Error bars denote the measurement uncertainty estimate (propagated through the 1st and 2nd order differences). The red ellipses on the top y-axes encapsulate the frames referenced in the text where Thompson claims to see a second gunshot wound to President Kennedy’s head.

The Resurrection of the Acoustics

 

The second new “puzzle piece” is actually an old puzzle piece long thought to be discarded, namely the hypothetical acoustics evidence from the DPD recordings. And with this, we must return to peeling back the onion that I spoke of earlier. Since their emergence from obscurity by buffs in the 1970s, the DPD recordings have almost become a story unto themselves. My original disposition (circa 2018 before being introduced to Paul Hoch and others) was that the DPD recordings were long ago debunked. But for starters, a brief chronological recap leading up to Thompson’s “resurrection” is warranted.

 

As discussed above, the DPD radio recordings were utilized by the HSCA as “evidence” of a second gunman on the Grassy Knoll in Dealey Plaza, which led them to their eleventh-hour amendment of the original conclusion of a lone gunman. The buffs, who had been summoned for ideas by HSCA Chief Counsel Blakey, managed to entice him into pursuing an investigation that turned out to be hasty and flawed, leading to a last-minute “Grassy Knoll gunman” scenario. The conclusion was based upon waveform matches between simulated test shots and the Channel 1 Dictabelt recording.

 

It was a mere couple of years later that an NRC Panel (which included Nobel Prize winning physicists Luis Alvarez and Prof. Norman Ramsey, informally called the Ramsey Panel) would debunk this conclusion for a good two decades, as Thompson laments in his book. The Ramsey Panel discredited the DPD recordings as “acoustics evidence” largely based upon a finding by Steve Barber, a musician who had an ear for audio recordings. Barber (with his friend Todd Vaughan) was able to make out an instance of “crosstalk” of a Channel 2 broadcast onto the Channel 1 open mic recording during which time the HSCA claimed the gunshot impulses occurred. The crosstalk in question was broadcast by Sheriff Bill Decker about 1 minute (according to the Ramsey Panel) after the shots were fired: “hold everything secure until the homicide and other investigators can get there… [to the railroad overpass].” This particular crosstalk instance has since been called “HOLD” for short.

 

Because it is known that HOLD was broadcast on Channel 2 well after the shots occurred, that meant that the suspect impulses on Channel 1 could not have been gunshots due to the timing. Barber shared this information with the Ramsey Panel, who would eventually conclude the same after rigorously confirming that the HOLD signal on Channel 1 was indeed the Decker broadcast on Channel 2. This should have been the nail in the coffin for the “acoustics evidence,” and for all intents and purposes, it was.

 

But alas, some twenty years later, out of the blue there appeared a paper in the UK forensic journal Science & Justice, which would receive mainstream media attention, thereby overnight “resurrecting” the DPD recordings. The article’s author was Donald Thomas, a research entomologist employed by the USDA. Thompson would learn about the article in the Washington Post (26 March 2001), and 9 months later he would arrange to meet up with Thomas.

 

As with others that he agrees with, Thompson takes a personal liking to Thomas, and he paints an agreeable portrait of a gracious, thoughtful, soft-spoken man of a few words. Thomas would go on to publish a 700+ page pro-conspiracy apologia in 2010 entitled Hear No Evil: Politics, Science & the Forensic Evidence in the Kennedy Assassination, which, with its gratuitous ad hominem attacks and politicizing generously sprinkled in with “science and the forensic evidence,” does not leave this impression.

 

In his 2001 article, Thomas attempted to “one-up” the Ramsey Panel with a crosstalk of his own, spoken by DPD motorcycle officer Samuel Bellah, which would counter the timing disqualifier of the HOLD crosstalk discovered by Steve Barber and argued by the Panel. The possibility of other crosstalks existing on the Channel 1 recording itself is not controversial, but here Thomas’s claim was that the timing of this particular crosstalk would trump the HOLD crosstalk and restore the timing of the Dictabelt impulses to the time of the shots in Dealey Plaza.

 

Enter Michael O’Dell, a systems analyst and computer scientist who took a keen interest in the science of the acoustics evidence after having read the Washington Post article that reported on Thomas’s paper. Thompson makes positive reference to O’Dell several places in his book. It was Paul Hoch and Max Holland who originally connected me with him as a subject matter expert in this area of the case. O’Dell told me how, consulting for numerous government and private firms, he became adept at problem solving in a wide variety of real-world situations, and he has very effectively applied that skill set to this aspect of the case. He had knowledge, software, and tools for manipulating audio, at which point in 2001 he dove into the deep end of the pool.

 

In relatively short order O’Dell single-handedly discovered a small error in the NRC Panel’s timing (for reasons having to do with the skips on the Channel 2 Audograph record). This error did not at all negate the conclusions of the NRC Panel, but it did negate Thomas’s conclusions by establishing that the impulses based on his alternative Bellah crosstalk would still be 30 seconds too late. Thomas himself would concede at that time that his “objection to the NRC’s hypothesis is largely blown away,” but it is fully understood that this was not due to any incompetence on his part; thus, while his paper’s conclusion was rendered obsolete, it was not a blemish on his reputation as an entomologist. And that should’ve been the end of it, a brief encore of the final act of the “acoustics evidence.”

 

But alas, it wasn’t. Thomas was not done yet; as Thompson put it: “Undaunted, Thomas soldiered on, pretty much alone and largely ignored by the media,” which is a strange statement given that neither Thomas nor Thompson were ever “ignored” by all the media. Apparently, the mining for alternative crosstalks proved to be too tempting an endeavor to pass up, and within a year, he would find another “crosstalk,” this one in accordance with O’Dell’s revised timing. The alleged crosstalk in question was from a Channel 2 broadcast just before 12:30 from Deputy Police Chief N. T. Fisher, this being “Naw, that’s all right, I’ll check it”; the purported crosstalk on Channel 1 only includes the latter 3 words “I’ll check it” (“CHECK” for short).

 

The surviving members of the Ramsey Panel graciously sent Thomas a pre-publication manuscript rebutting his 2001 paper (submitted to the same journal). But Thomas would promptly write them back informing them of his recent discovery of CHECK… and also indirectly accusing them of intentionally covering it up! So much for the collegial assumption of good faith, and it is no wonder they never dignified it with a reply. And Thompson doubles down on the accusation, the posthumous resentment against Alvarez (and Norman Ramsey) not at all waning.

 

The Ramsey Panel rebuttal paper would eventually be published in Science & Justice in 2005 (Linsker et al., 2005) [13], including a full rebuttal of the claim of a CHECK crosstalk, concluding that the words in question heard on Channel 1 were not, in fact, the same as “I’ll check it” spoken on Channel 2.

 

And their conclusion isn’t out in left field. My curiosity piqued, I went ahead and “checked it” myself. On the Internet (YouTube), I was able to locate both events (CHECK and HOLD) on the Channels 1 and 2 recordings:

1. For Channel 1, I made use of Larry Sabato’s Channel 1 sequence: Assassination of John F. Kennedy -- Dictabelt Channel 1 Audio. The alleged “I’ll check it” (CHECK) crosstalk is at 13:57, and “hold everything secure” (HOLD) is at 14:08.
2. For Channel 2, I relied on a time-corrected recording that Michael O’Dell has recently made available with motorcade locations: Kennedy Assassination Dallas Police Radio Recording, Channel 2. The alleged “I’ll check it” (CHECK) is at 6:33 and “hold everything secure” (HOLD) at 8:11.

Using headphones, I carefully listened to these multiple times, and while HOLD sounds the same on both channels, CHECK subtly does not. I was left with the impression that CHECK on Channel 1 was probably another instance of someone, perhaps the same officer (Fisher), saying “I'll check it,” but it wasn’t crosstalk with the Channel 2 transmission claimed by Thomas (similar to what the Linsker et al. paper suggested). How the words are spoken is different—there is a different rise and fall in the pitch, giving rise to a different intonation of the words.

 

However, Michael O’Dell, Steve Barber, and Sonalysts, whose judgment I trust more than my own on these matters, have concluded that they are not even the same words, O’Dell and Barber believing that the words are closer to “all right Chaney” (Sonalysts believing the words were “five seven”), and that they were directly spoken by someone on Channel 1 (and thus not crosstalk from Channel 2). Of course, although our ears aren’t quantitative measuring devices, they are nonetheless well-adapted to pick up on the subtleties of human speech (especially if they are trained in music and audio recordings, as with Barber and O’Dell).

 

And in this case, it would seem that our ears haven’t failed us, for an independent study commissioned by Larry Sabato,[14] performed by Sonalysts, Inc. (Olsen and Martin, 2013 [15]; Olsen and Maryeski, 2014 [16]), has subsequently confirmed this, concluding that CHECK was not a crosstalk from Channel 2, but rather a Channel 1 transmission on another mic, as evidenced by the presence of a heterodyne.

 

Thus far, in a nutshell, the “acoustics evidence” has been discredited by multiple independent lines of reasoning, each articulated and refined over the intervening decades since the HSCA hastily introduced it into their final report. The numerous rebuttals fall into several categories.

1. Timing issues. This was the first and oldest rebuttal, the one discovered by Steve Barber with the HOLD crosstalk and originally argued by the 1981 NRC Panel. The hypothesized gunshot impulses on Channel 1 occurred approximately 60 seconds after the gunshots in Dealey Plaza. This has been reaffirmed in the paper by Linsker et al. (2005), and more recently (and compellingly) by both Michael O’Dell and Sonalysts (Olsen and Martin, 2013). The subsequent attempt to supplant HOLD with CHECK has been discredited for multiple reasons, including some not yet discussed here.
2. Open mic location assumptions. The entire case for the acoustics evidence assumes that a motorcycle in the motorcade was broadcasting on Channel 1. Furthermore, the HSCA conclusion was based upon an impulse match that required the motorcycle with the open mic to be in an extremely specific location near the intersection of Houston and Elm at the time of the shots. These assumptions are either called into question or outright invalidated because of the following (roughly in order of increasing weight):
1.  The open mic was on Channel 1. Channel 2 was the channel for “special events" (which at 12:30 on 22 November 1963 was the Kennedy motorcade). All motorcycles in the motorcade would have been using Channel 2 as standard operating procedure, thus there should not have been a mic stuck open on Channel 1 within the motorcade. [17]
   
2. DPD testimony. Related to the above, the motorcycle identified by the HSCA was that of Officer H.B. McLain, but both he and Sgt. James Bowles, key witnesses in this arena, have vehemently denied it with a number of plausible rationales that the HSCA apparently chose to ignore (e.g., see the 65 page “Endnote” #381 in Bugliosi). 
3. Photographic evidence. In 2003, computer animator Dale Myers performed a thorough epipolar geometric analysis of available photographic evidence from the crime scene (primarily home movie films taken of the motorcade) and concluded that H.B. McLain’s motorcycle was approximately 175 feet from the location found by the HSCA acoustic experts (BBN and W&A) just half a second before the first shot. 
4. Analysis of Channel 1 audio, including motorcycle engine speed. In 2013, Sonalysts analyzed the DPD Dictabelt recording (Channel 1) and rigorously determined from a number of different considerations (including the motorcycle engine speed) that the “data uniformly indicate that the motorcycle with the open microphone was not part of the motorcade” (Olsen and Martin, 2013).

Each of these arguments, taken individually, independently discredits or seriously calls into question the DPD recordings as “acoustics evidence” of multiple gunmen in Dealey Plaza (implying a JFK conspiracy, as concluded by the HSCA). However, with the exception of the first line of arguments based on timing inconsistencies (mostly involving crosstalk between Channels 1 and 2), the remaining facts argued above about the DPD recordings were mostly glossed over or ignored entirely by Thompson.

 

But my own initial objection to the DPD recordings as acoustics evidence was more elementary. As I continued my foray into this topic, there was something on the surface that bothered me.

 

I found myself immediately doubting this claim that there was a 95% confidence of a Grassy Knoll shot (Don Thomas actually elevates this “confidence” to a ridiculous number in his book), when one cannot even hear the sound of gunshots on the recording. Of course, as a scientist I’m well aware that our immediate five senses are not the last word on what we can know about reality. For that matter, one also cannot readily “see” the forward head snap on JFK’s head when watching the Zapruder Film at full speed. But, nevertheless, our senses are still “instruments” well-adapted to ascertaining information about the physical world, and it is from these that we gain intuition allowing us to formulate scientific hypotheses.

 

We are told that we can’t distinguish the gunshots on the recordings with our ears, but why is that? One reason, of course, is the AGC feature designed for damping loud sounds coupled with a loud background sound (motorcycle engines), which kicks the base level gain down. But what exactly does this mean? In this case it means, among other things, a key piece of audio information is missing from the DPD recordings, namely the amplitude information (i.e., sound volume) in the waveforms. The DPD recordings simply do not contain sufficient information to distinguish gunshots from other potential sources.

 

So, what did the HSCA investigators (Barger and W&A) claim to find? In a nutshell, they looked for and found waveform patterns from Dealey Plaza test configurations involving gunshots from two locations (the TSBD and the Grassy Knoll) that matched those on the Channel 1 recording. The problem is the waveforms on the Channel 1 recording have been significantly truncated in amplitude-space due to the AGC. Especially without this additional information about the nature of the sounds (i.e., amplitudes well beyond the AGC truncation level), there might very well be other sources that could lead to the suspect waveform patterns (including the impulses attributed to echoes).

 

And I subsequently found that I was in good company in this line of reasoning. Both Michael O’Dell and Sonalysts (Olsen, Bamforth, and Grant, 2013) were way ahead of me, having already made arguments to this effect, and they have demonstrated it using their own independent analyses.

 

Sonalysts simply analyzed the waveforms and found that the impulse patterns attributed to gunfire were not uncommon on the Channel 1 recording and could have resulted from any number of sources, including mechanical noises from the open-mic motorcycle.

 

O’Dell has gone further by clarifying precisely what was meant by W&A’s “95% confidence” (or the exaggerated 99.999% confidence boasted by Don Thomas), and it is not anywhere close to saying that “we are 95% confident there was a Grassy Knoll shooter.” We are not even in the same ballpark.

 

As Mark Twain is reputed to have said, “Facts are stubborn things... but statistics are more pliable.”

 

Statistical significance tests are extremely useful tools for determining if relationships in sample data are meaningful and not the result of blind chance, but the relationship in question must be explicitly and precisely defined, and the conclusions deduced in accordance with the specific relationship under consideration. In this case what was meant by “confidence” was simply that the matches between the impulse timing patterns on the Dictabelt with the test impulses were extremely unlikely (e.g., 95%) to have been the result of randomness.

 

That’s all well and good, but the original recording signals were not random—there were in fact many real, non-random signals, including motorcycle (engine and other mechanical) sounds, heterodynes, and of course, voice broadcasts and crosstalk by DPD officers. And even if we assume for sake of argument that the Channel 1 background signals were completely random, we have only determined that the waveform pattern matches were due to a non-random signal that still could have had any number of sources. So, we find that there was an implicit assumption that the recorded signals were either impulses associated with gunshots (including echoes), or random noise.

 

And here is the clincher: Based on his own waveform analysis of the Decker HOLD broadcast, O’Dell has plausibly argued (and convinced me) that the suspect “gunshot” waveforms were, in fact, due to the HOLD crosstalk itself.

 

That is to say, the “acoustics evidence” may in fact be nothing more than the non-random waveform signal caused by the vocalization of the words “hold everything secure until...” on the Channel 1 recording. For more details on this, the reader is referred to O’Dell’s excellent online article (hosted on the McAdams Kennedy Assassination website): The acoustic evidence in the Kennedy assassination. Specifically, Figure 4 (op. cit.) shows the exact overlap between Decker’s voice and the suspect “Grassy Knoll” shot.

 

But regardless of the suspect waveform source (whether it was the HOLD crosstalk or some other non-random source signal), there is now therefore a third (in my opinion, more fundamental) argument that should be added to the above list: In addition to the timing issues and open mic assumptions, the DPD recordings do not contain sufficient information for the suspect impulses to be uniquely identified as gunshots. In layman’s terms, the Channel 1 waveforms are not the equivalent of “ballistic fingerprints.”

 

Even if we assume for sake of argument that the motorcycle with the Channel 1 open mic was in the motorcade, and located at the requisite place and time near the intersection of Houston and Elm (in spite of photographic and audio evidence to the contrary), the suspect impulses could still have originated from any number of sources, including the spoken words “Hold everything secure” that occurs simultaneously on the Channel 1 recording (regardless of whether or not one wants to deny that those words were crosstalk).

 

This means that the best one could ever hope to claim from the DPD recordings was that the waveforms could possibly have been caused by a gunshot from the Grassy Knoll, but this was not to the exclusion of other possible causes—possibilities that both O’Dell and Sonalysts have subsequently identified as, for example, the spoken words “Hold everything secure,” or a microphone rattling on the motorcycle (among other possibilities).

 

But the “other possibilities” go even further than waveform audio sources.

 

O’Dell has also performed his own acoustical model simulation of Dealey Plaza (akin to W&A, albeit using a computer), and he actually found numerous waveform “matches” with gunshots fired from other various locations in Dealey Plaza, some of them laughable; for example, a gunshot originating at Zapruder’s pedestal produced multiple matches with open mic motorcycle locations in Dealey Plaza. Therefore, from the DPD recordings alone, one also cannot rule out gunshots from other locations.

 

That, to put it mildly, is a far, far cry from the claim that there’s a 95% confidence of a gunman on the Grassy Knoll.

 

To recap, there are three general categories of arguments that continue to discredit the “acoustics evidence,” each with their own independent lines of reasoning, “resurrection” notwithstanding:

1. Timing issues.
2. Open mic location assumptions.
3. Insufficient information-content within the DPD recordings.

Thompson all but ignores the second and third set of arguments. Although the third (in my opinion) is the most fundamental of the three, it is also the least known or argued, so we can certainly cut Thompson some slack on that. But we are not yet done discussing the first one, because Thompson attempts (with a certain dramatic flair) to mount a renewed defense of CHECK.

 

Thompson’s Defense of the “Acoustics Evidence” and Barger Redux

 

In spite of what we have since learned about the DPD recordings in the decades since the HSCA’s rash decision to insert its analysis into the JFK case, Thompson has remained committed to the spectre of a Grassy Knoll gunman. Thompson makes a final attempt to gain the upper hand on the ghost of Luis Alvarez, who has apparently lived on within the “Ramsey Panel survivors.” He devotes his concluding chapters in Part IV to a defense of the acoustics evidence against the rebuttal of the CHECK crosstalk by those survivors (Linsker et al.).

 

In support of the alleged CHECK crosstalk event, Thompson first offers two simple arguments that he arrived at with his Editor, John Grissim, who had become a “true collaborator” with him on his crusade against the Ramsey Panel survivors. They also came up with simple, easily remembered names for their arguments, namely “the unshrinkable 88 seconds,” and “move one, move all.” These are followed by what was apparently meant to be a coup de grace, namely an encore from one of the original acoustical experts for the HSCA, James Barger. We will now examine each of these.

 

The “Unshrinkable 88 Seconds”

 

Thompson is here referring to an irreconcilable time difference between the original CHECK and HOLD transmissions on O’Dell’s time-corrected Channel 2 recording versus the time difference between their suspect Channel 1 crosstalks. The time difference on Channel 2 is 98.7 s, but the difference on Channel 1 is only 10.9 s, thereby leaving a minimum 87.8 s of time unaccounted for between the two events on Channel 1. It’s called “unshrinkable” because if the Channel 2 recording has intermittent missing times (due to periods of radio silence when the Audograph automatically shut-off to conserve the recording medium) that would increase, not shrink, the discrepancy. Thus far, this is all true. Thompson concludes that this “unshrinkable” time on Channel 2 means that the HOLD crosstalk on Channel 1 is an “artifact” (i.e., it got there by some other means).

 

However, this conclusion is not only wrong, but the proper conclusion is actually the opposite of what Thompson claims. This time discrepancy merely indicates that these two Channel 2 transmissions (CHECK and HOLD) could not both be Channel 1 crosstalks, but from this information alone, we do not know which.

 

To answer that question, O’Dell points out that one needs to check for consistency with the other established time markers. This involves examining the time differences of these two with nearby established simulcasts (broadcasts on both channels) and crosstalks (viz., the “HOLD family”: “BELLAH1,” “BELLAH2,” and “I’VE GOT”), and looking to see elapsed times between the event pairs on the Channel 1 continuous recording are greater or equal to those on Channel 2 (i.e., Δtch1 ≥ Δtch2). O’Dell has performed this test and he has confirmed that the elapsed times between HOLD and other events on Channel 1 are indeed all greater than those on Channel 2 (i.e., for HOLD, Δtch1 > Δtch2 as required).

 

But what about CHECK? Perhaps not too surprisingly, for CHECK the opposite is true: Elapsed times between “CHECK” and the other reference events (BELLAH1, BELLAH2, I’VE GOT, SIMUL) on the Channel 1 recording were smaller than those on Channel 2, which is an impossibility. Thus, while the “unshrinkable 88 seconds” does indicate that something was amiss between the timings of the alleged “CHECK” and HOLD events on Channels 1 and 2, when similar comparisons are performed for other event pairs, we find that HOLD is universally consistent and CHECK is universally inconsistent. Therefore, we have just stumbled upon another independent argument that points to the fact that the alleged “CHECK” event on Channel 1 is not a Channel 2 crosstalk.

 

“Move One, Move All”

 

Thompson’s second argument is related to “the unshrinkable 88 seconds” in that it tries to address the missing time on Channel 2 due to the recording dropouts from the automatic shutoff feature. The Ramsey Panel deduced from various time-marker events that the amount of missing time was 46 seconds. Thompson questions how likely it would be that they would somehow come up with exactly what they “needed” (46 seconds), as if they were somehow back-engineering it, but they did not “need” anything. Forty-six seconds was simply the deduced missing time—there was nothing special or required about the value.

 

But Thompson further argues that this insertion of missing time would not only “move” all the earlier Channel 2 events back in time, but it would also move the events back in space, thereby moving Chief Curry in the lead car back onto Main Street when he was announcing that he was at the Triple Underpass in Dealey Plaza. Admittedly, this one had me initially scratching my head, because on the surface it seemed a reasonable proposition. But Michael O’Dell was not so easily convinced, and he explained to me how the argument was fatally flawed.

 

First, it should be noted that because the Channel 2 recording has pauses, the discovery and determination of missing time between location broadcasts (by Curry) should come as no surprise to anyone. Consequently, the insertion of time determined to be missing (approximately on the order of a minute or less) between locations on Channel 2 due to recorder pauses does not necessarily interfere with the car location (during the continuous Channel 1 broadcast) provided that (1) the rule Δtch1 ≥ Δtch2 is not violated, and (2) we do not violate any physical limits regarding the motorcade movement (i.e., the limo speed) on Channel 2.

 

In other words, there is no issue with inserting time on Channel 2 so long as the recorded time between events on Channel 2 is less than or equal to the recorded time between the same events on Channel 1, and that the time insertion does not involve impossible limo speeds.

 

But let’s assume for sake of argument that the Channel 2 recording did run continuously. Time insertion on Channel 2 still does not move events because the timeframe in question occurred following the last location call-out by Curry at the Triple Underpass—in other words, we are not inserting time between known events in the assassination sequence. This means that the Channel 2 events are not altered in the least because there is nothing “anchoring” them to any time or location—they are all free to slide together in unison and retain their relative positions within the recording timeline, without even altering the limo speed.

 

Barger Redux

 

The Linsker et al. (2005) rebuttal to Don Thomas’s alternative-crosstalk hypothesis of the Channel 2 CHECK transmission ultimately compelled Thompson to reach out to James Barger, the acoustical engineer originally commissioned by the HSCA to study the DPD recording. Thompson relays to us his encounter in his chapter entitled “Journey to the Lair of the Master.” Barger, for his part, was receptive to Thompson’s request to reexamine the DPD recordings, specifically the alleged CHECK event, and he subsequently enlisted one of his younger engineers, Richard Mullen.

 

The first question they sought to address was why Linsker et al. did not report the pattern cross-correlation (PCC) value that they presumably found between the Channel 2 CHECK broadcast and the Channel 1 suspect pattern. Don Thomas basically straight up accused the Ramsey Panel of cover-up, insinuating that they found a significant PCC value for CHECK but intentionally withheld it from the public. However, this accusation was simply never true: The Ramsey Panel did, in fact, report a PCC result for CHECK.

 

Mullen, on the other hand, was simply able to ascertain that a PCC peak for CHECK could be obtained from a significantly narrower sampling window, approximately only 12.5% of the width originally used. Thompson does not mention that the window used by Linsker et al. was for analyzing the recordings in an objective manner meant to find pattern correlations, nor does he question the justification for the using a smaller window, other than the purported crosstalk (“I’ll check it”) was of a shorter duration. Instead, he took this as a new development as if out of “a movie—a scientific detective thriller,” while eagerly anticipating the PCC results for the HOLD crosstalk using the much narrower sampling window. Thompson shares his excitement with us, and I can relate to it from my own experience. He flew out from the West Coast to get the news in person from Barger and Mullen at Raytheon BBN in Cambridge, and there we learn (not too surprisingly in light of the lead up) that Mullen found a stronger PCC peak for CHECK over HOLD when using the smaller sampling window.

 

The PCC graphs are shown in Thompson’s Figures 22-5 and 22-6, and while the CHECK peak indeed has a higher absolute value over HOLD (0.5 versus 0.26), the background levels for each were 0.25 and 0.17, thereby leaving “net peaks” of 0.25 and 0.09, respectively. Thompson claims checkmate at this point, but I wasn’t convinced by the graphs that CHECK was thus confirmed. (As a side note, as with his first book, Thompson’s provision of graphs, data and methods in Last Second is to his credit). One reason for this is that if one considers the fractional peaks (i.e., dividing the net peaks by the total signal: net ÷ total), one gets 0.25 ÷ 0.5 = 50% and 0.09 ÷ 0.26 = 35%. When looked at this way, one can see that although CHECK still has a higher relative peak (using this sampling window), the difference from HOLD isn’t as large as one might be led to believe. When using the broader sampling window (as used by Linsker et al.), the PCC peak for HOLD is notably higher as shown in Thompson’s Figure 22-7, with a net peak of 0.16 (0.24 total peak minus 0.08 background), or a fractional peak of 67%, which is greater than CHECK by about the same margin.

 

So even if I wasn’t myself convinced at this stage that HOLD was invalidated (especially given all the other evidence that attests to its validity), CHECK nevertheless did show a significant PCC peak, apparently contrary to the Linsker et al. findings. Michael O’Dell pointed out that Mullen’s PCC study, like Thompson’s “unshrinkable 88 seconds,” was premised on CHECK being a valid crosstalk. However, this premise is almost entirely based upon the transcript by DPD Sgt. Jim Bowles, who had denoted the Channel 1 transmission as a crosstalk. This was the justification used by Mullen to narrow the sampling window in his PCC analysis. But Bowles’ transcript was based upon his ear, which is subjective and not definitive. This means that there never was an objective justification for using the smaller sampling window.

 

Thompson provides no justification for his treatment of the PCC peak value as a “score” that means one match is better than the other. O’Dell also pointed out to me that while the PCC test provides a measure of similarity between two signals, the test itself does not prove they are identical, and it otherwise makes no sense to compare scores from entirely different samples and decide which is the “better” score.

 

O’Dell has identified three additional fundamental problems with Mullen’s PCC analysis that I shall only briefly summarize here.

 

First, the PCC peak is found to occur for time-stretch factors in the neighborhood of 1.2, which is far too large (in terms of audio pitch) than the expected value of 1.

 

Secondly, O’Dell has also generated spectrographs of the purported CHECK broadcasts (“I’ll check it” on Channel 2 along with the alleged crosstalk on Channel 1) and has found that they are not the same.

 

In the end, all the timing problems that Thompson (and Don Thomas) have laid out go away simply by rejecting the hypothesized CHECK crosstalk.

 

Barger and Mullen offered Thompson one additional argument against the validity of HOLD: The presence of multiple hum frequencies on the Channel 1 recording suggested that the HOLD crosstalk was somehow overdubbed during the creation of a multigenerational copy. However, they based this on the Ramsey Panel copy, which was already known to be a multigenerational copy, so there was nothing surprising about the presence of these hums. The only way Barger could solve the timing problem was to conclude that there was overdubbing, and the only alleged direct evidence of overdubbing is the extra hums.

 

If a mysterious overdubbing were the origin of the HOLD crosstalk, then this hypothesis can be tested against earlier copies: Not only should earlier copies not only not contain the hums, but also, they should not include the HOLD crosstalk. And O’Dell, once again, has convincingly demonstrated that this is not the case. Verify for yourself that the HOLD crosstalk is present on all three known versions of the recordings (the Ramsey Panel’s, the FBI’s, and BBN’s); therefore HOLD is not a dubbing artifact: Dictabelt Hums and the "hold everything secure" Crosstalk.

 

The Acoustics Non-Evidence

 

As mentioned earlier, the storied history of the acoustics evidence in the Kennedy assassination case is a multilayered one that has almost taken on a life of its own, evolving through the decades into a nine-headed Hydra. It was not without some time and effort (researching the topic with the guidance of Michael O’Dell and Paul Hoch) that I was able to make any sense of it. And what I have found (to my own satisfaction) is that the DPD recordings cannot be considered legitimate acoustics evidence in this case.

 

To distill it down for the reader, a credible conclusion that it has been found with 95% confidence that there was a shot from the Grassy Knoll (as claimed by the HSCA and by subsequent proponents endorsed by Thompson), really should have been based upon the following premises and findings:

1. It was established that there was an open mic within the motorcade at the time of the shots, and audio information from the crime scene was recorded by the Dictabelt at the time of the assassination.
2. The Dictabelt recording contained sufficient information (within the waveforms) that could be reasonably attributed to gunfire to the exclusion of other possible (non-random, non-gunshot) ambient sources.
3. Subsequent acoustical analyses were performed in an effort to isolate the location(s) of the identified gunfire by obtaining test waveforms (featuring echoes) from rifle shots performed at all hypothetical sniper locations in Dealey Plaza.
4. Impulse correlations between the test and suspect waveforms indicated with 95% confidence that, for a transmitting mic at a known location, one of the suspect gunshot waveforms (including the identified muzzle blast, shockwave and building echoes) matched a test waveform from a Grassy Knoll shot to the exclusion of all other locations.

That, my friends, would have constituted iron-clad, dispositive evidence... which is what is needed for such an extraordinary claim.

 

Instead of this, what we have is the following.

 

Premise #1 has, over the decades, been invalidated by multiple independent panels, experts and researchers, including Ramsey et al., Linsker et al., Dale Myers, Sonalysts, Barber, and O’Dell. This has been much of the focus of the "anti-acoustics" arguments because it wholesale rules out the DPD recordings: The open mic was not in Dealey Plaza and the recorded audio was not concurrent with the assassination.

 

However, all along Premise #2 was implicitly missing, as were parts of #3 and #4. In place of this, an implicit assumption was made, namely that the only signals contained in the Dictabelt recordings were either random noise or echo patterns from gunshots at two specific locations. Hence, the ensuing echo correlations with test shots were taken to be prima facie evidence that (1) the Dictabelt did in fact contain gunshot impulses from Dealey Plaza, and (2) one of those gunshot impulses originated from the Grassy Knoll. However, this was based on a false dichotomy. There were, in fact, other possible ambient audio sources of the waveforms found on the Dictabelt— correlations with echoes by themselves are not “evidence” that they were, in fact, gunshots. Thus, because the Dictabelt recordings do not otherwise contain waveform signals uniquely attributable to gunfire, other sources cannot be discounted as possible causes of the so-called “echo patterns.”

 

With these fundamental limitations of the Dictabelt recordings in mind, the location of the transmitting mic was not found to be in the specific place it needed to be (as established by Myers), nor in the motorcade (as established by Sonalysts), nor did the suspect impulses occur during the assassination timeframe (as established by Ramsey, Linsker, et al.), and the match was not to the exclusion of all other locations (as established by O’Dell). These subsequent facts confirm to us that the Dictabelt waveform patterns (including those attributed to “echoes”) had a non-gunshot source.

 

From my own perspective, it was the lack of Premise #2 that I found to be particularly problematic, because in my opinion it fundamentally limited what one could have ever expected to conclude from the recordings. Even if all the other problems were to go away (e.g., if the bike with the mic were proven via photographic evidence to be at the specific location needed and HOLD were not a crosstalk), the most anyone could have ever hoped to claim was that “based upon our echo pattern correlations, we were not able to rule out a shot from the Grassy Knoll.”

 

Of course, such a finding might still have been considered corroborative if there had been other physical evidence of a shot from that location. But outside of that we are left at an inconclusive dead end: The DPD recordings do not contain acoustics evidence of a Grassy Knoll gunman.

 

The “Blur Illusion”

 

This topic has been deferred to its own section, primarily because we will be delving into it in more detail, and I will be covering some moderately technical concepts pertaining to remote sensing and photogrammetry. Although I originally hoped to avoid a deep dive into this obscure topic, in the end I felt it was better to err on the side of thoroughness in addressing what I consider to be the most absurd of Thompson’s “puzzle pieces,” the “blur illusion.”

 

Overview

 

With the subsequent “resurrection” of the DPD recordings as “acoustics evidence” and the rejection of the NAA findings, Thompson found himself facing “an impasse”: How does one reconcile his “simultaneous shot” hypothesis with both the autopsy evidence and the acoustics evidence? The answer would manifest itself in rejecting a key finding of his own 1967 book, namely the measurement of an approximate 5 cm (≈2-inch) forward head snap observed in Z313.

 

The forward head snap was discovered prior to Thompson and is clearly observable to anyone with the still frames. Thompson’s main contribution was that he was the first to measure quantitatively (working with Haverford physics student William Hoffman) the change of position in Kennedy’s head. Thompson was primarily concerned with the “back and to left '' motion seen while watching the movie at full speed, but by taking measurements he provided us with extremely valuable data amenable to quantitative dynamical analysis, especially the 5 cm forward snap of Kennedy’s head.

 

But Thompson would find a way around the “impasse” of the Z312–Z313 forward head snap from a systems analyst by the name of David Wimp, who presented his unusual takes on the Zapruder Film as early as 2002 . Referencing Wimp, Thompson proposes that what we observe in the Z312–Z313 frames is merely an “illusion” brought on by the blurring of the image evident in frame Z313. Thompson asks, “Why didn’t I think of this?” and makes odd claims that it’s “elementary” and “so obvious and so simple.” I say “odd” because, first of all, the notion that a blurred frame in a movie film could create the illusion of motion when there was none is anything but “obvious” — indeed, if it were so “elementary” then why hadn’t anyone else “thought of it” as well? This not only includes Thompson (who’s no dummy), but also his student assistant Bill Hoffman (one of “the best” students in the Haverford physics department), two Nobel Prize winning physicists (Prof. Richard Feynman and Prof. Luis Alvarez), Francis Corbett and his team at Itek Corporation, Kodak film expert Roland Zavada, Ken Rahn, Robert Groden, David Lifton, Gerald Posner, Larry Sturdivan, me, and practically every other researcher who has studied the still-frames, WC critic and apologist alike.

 

Of course, part of the reason for this is that you can plainly see for yourself that the back of JFK’s head does, in fact, move forward by a measurable distance from frames Z312 to Z313 [See: Animated GIF #1].

 

 

Animated GIF 1 - Motion of President Kennedy’s head between Z312 and Z313.

 

But importantly, it is not only the back of his head (where the “blur illusion” is said to take place) that moves forward, but rather it is the entire head that pivots forward and downward on his neck, and even appears to be distorted in shape (which is precisely what happens from the force of the cavitation pressure wave).

 

And, although this is a grim topic to dwell on, one must also keep in mind that by the time of the shutter closing at Z313, the concept of “JFK’s head” involves more than simply the intact portion of it, because a good portion of his head mass is at that moment found within an explosive cloud of blood, tissue and bone seen forward of the remaining intact part. [19] Furthermore, as pointed out in the 2018 Heliyon paper, this forward motion is observed only with JFK's head: It is not observed on any of the other limo occupants, nor even JFK’s own torso, wherein lies his center-of-mass (CM). The isolated motion on JFK’s head rules out the notion that there was a fictitious force at play due to a decelerating reference frame, which would have acted equally on the CMs of all the limo occupants. Thus, there is absolutely no getting around the fact that the center-of-mass of JFK’s head moved forward in Z313, this unequivocally the result of an isolated real force that pushed it forward.  [20] 

 

So, what precisely is meant by the “blur illusion”? Further digging shows that the phenomenon alluded to by Wimp and Thompson is a bit more nuanced, and that the Devil is in the details… It was Paul Hoch and Max Holland who first brought my attention in 2018 to Wimp’s writings on the subject (along with Thompson’s embrace of it and Rahn’s critique). Hoch basically told me that I ought to look into it and not blow it off (as I was inclined to do). I subsequently researched Wimp’s writings, which are found online in three separate documents (referred to in Thompson’s book). My pursuit was performed somewhat grudgingly at first, especially given that Wimp’s writings are not clear, but I think I was finally able to decipher what the argument is.

 

And what I have found is that while the phenomenon that Wimp and Thompson allude to is a real phenomenon, the application is invalid, and the conclusion is therefore wrong—the forward motion cannot be dismissed as a “blur illusion.”

 

Although my education in geophysical sciences has trained me reasonably well in applied classical dynamics (which I used extensively in the Heliyon paper), my actual research specialty is in environmental satellite remote sensing, which by its very nature involves image analysis. Remote sensing science is closely aligned with photogrammetry (the technique of acquiring measurements from photographic images), but it relies more generally on principles of radiative transfer over the entire electromagnetic spectrum and is not limited solely to the solar spectrum visible to the naked eye. In recent years photogrammetric techniques have been applied in numerous fields of study, including engineering, geophysics, and forensic science. So it is through some odd twist of fate that I found myself reasonably well equipped for studying the very same Zapruder Film that Josiah Thompson had a small part in getting broadcast on national TV, exposing its graphic content to an ordinary kid in a small town in upstate New York, which would be the starting point that ultimately led to this review article.

 

And it turns out I am in good company. Luis Alvarez was among the first to discuss the blurring of various frames in the Zapruder Film in an effort to ascertain the timing of gunshots by analyzing “jiggle” evident in various frames. As Zapruder panned his camera to keep the limousine in his viewfinder, there were moments where he “jiggled” the camera, these being potential indicators of a gunshot startle response. But not all “blurred” frames were due to startle responses—many of them were simply due to “panning error.” While Thompson and other “critics” (e.g., Thomas) believe that the image blur at Z313 was due to a startle response, it was most certainly due to panning error, given that (1) it was almost entirely horizontal, and (2) the limousine was orthogonal to Zapruder’s field-of-view, which meant it required the greatest degree of rotational panning to keep it centered in the viewfinder. Couple that with the fact that the limo slowed down around Z313, and you get what’s observed, namely a horizontal panning error where the camera got a bit ahead of the limo. The effect of this error is to create a mostly horizontal front to back smear of the image, which we colloquially refer to as “blurring.” 

 

Error Bars and Uncertainty Estimates

 

While the consideration of scene smearing on the analysis of images is legitimate, Thompson (and Wimp) seem not to acknowledge the fact that image smearing was, in fact, already accounted for in the measurements independently carried out by Itek Corporation for CBS News in 1976 (Thompson references the Itek study in Chapter 12). The effect of smearing on an image is simply to create greater uncertainty in feature positions on the 2-D Cartesian grid due to the convolution with adjacent picture elements, leading to an irreversible loss of resolution. But let’s be clear: All the position measurements reported by both Itek (1976) and Thompson-Hoffman (1967) were subject to some degree of uncertainties. In the Thompson-Hoffman measurements, a single uncertainty value was reported (0.13 inches) for all the frames they measured (Z301–Z330), but the Itek data provides individual uncertainty estimates for each frame they examined (Z312–Z317).

 

To confirm my understanding of Itek’s measurements and uncertainty estimates, I consulted with the Itek Principal Investigator, Francis Corbett in early 2019 and again early 2021, and he told me:

Our work was rigorous and we went to great lengths to ensure the impact of all possible variables were accounted for, and we produced true and accurate results of the body motions within the limousine..

So, the error or uncertainty quoted in our report, I believe resulted from the repeated, multiple measurements by independent observers. And to answer your question, I believe the uncertainty we quoted would have accounted for, or included, all sources, since it was basically a heuristic analysis, including that due to motion blur too.

When we look at the Itek data, what we find is that the uncertainties in the “blurred” frames are significantly higher than in the non-blurred frames, but they are still easily within the necessary precision for resolving an approximate 5 cm (≈2”) movement.

 

This can be seen in Figure 2, where I have plotted Itek’s data along with their error estimates in the same manner as before. In terms of the observable kinematics, the middle and right plots are indicative of speed and acceleration. From Newton’s 2nd Law, we may deduce the forces (at the relevant timescales) on JFK’s head and body as follows. First, there was an initial large forward impulse on the head that occurred just following the shutter closure at Z312 — this was due to the direct momentum transfer of the projectile during collision with the skull and passage through the soft tissue. Immediately following this was a delayed, but equally large rearward acceleration of the head prior to Z314, this being due to the indirect jet recoil effect brought on from the cavitation wound left in the wake of the projectile. Finally, there was a further-delayed, 2-frame rearward acceleration of both the body and head, which given the timescale and magnitude could only be due to a neuromuscular spasm. For more details on the dynamical analysis, the reader is referred to my paper Gunshot-wound dynamics model for John F. Kennedy assassination; more details on the neuromuscular reaction and gunshot ballistics can be found in Larry Sturdivan’s book, The JFK Myths: A Scientific Investigation of the Kennedy Assassination.

 

 

Figure 2 - Similar to Fig. 1, but showing motion of President Kennedy’s head (blue) and shoulders (red) from the Zapruder Film based on data from Itek Corporation and plotted vertically in the same manner as Fig. 1: (a) near horizontal distance per frame, (b) calculated speed in Imperial units (ft/s), and (c) calculated acceleration (ft/s2), showing initial dominant accelerations on JFK’s head (impulse and recoil effects), followed by identical accelerations on both his shoulders (upper torso) and head (indicating forcing on the higher-mass torso). Error bars denote the frame-by-frame uncertainty estimates (propagated through the 1st and 2nd order differences) provided by Itek.

The Kodachrome II Outdoor Reversal Film

 

However, back to the image blur, it was nevertheless incumbent upon me to make sense out of Wimp and Thompson’s argument. Although difficult to discern from their writings, the basic claim appears to be that because the film has a non-linear response to light exposure, asymmetric displacements can occur in the sharp edges between light and dark boundaries. Wimp effectively illustrates this by showing frames near the end of the Zapruder Film sequence, which Thompson reproduces in his book, specifically zoomed portions of Z405 and Z409 (shown as Photos 14-1 and 14-2). Z405 shows a relatively sharp (unblurred) frame with a lamppost centered in the image, whereas a few frames later, Z409 shows a severely blurred image of the same lamppost. Many of the frames in Zapruder’s film are severely smeared after the fatal shot at Z313—these were likely due to Zapruder’s shock and horror of what he just saw and were much more haphazard than the simple panning and startle errors earlier in the film.

 

What is interesting in these two images is that in the blurred image the lamppost shows a distinct discontinuity between the pavement and the grass turf in back of it. Over the brighter pavement, the lamppost has visibly shifted toward the left in the photo. Although these two frames don’t follow one after the other, if one were to align them one might incorrectly think that a portion of the lamppost moved, and that this would be an “illusion.” Note of course that this neglects the fact that there were several frames in between these two and that the lamppost had in fact moved in parallax due to the panning of the limousine. Note also that we would have physical justification to claim that the discontinuity in the lamppost is not real and thus instead some sort of artifact in the photo, and furthermore the smearing in frame Z409 was so extreme (and the lamppost’s edges so blurred) that Thompson-Hoffman and Itek would likely have omitted it as unusable for measurements. But for the sake of argument, I decided to ignore all these disqualifying factors and make an attempt to understand Wimp’s argument, especially given these illustrative frames near the end of the film, and whether or not the argument would apply to Z312–Z313.

 

So, to study the matter further, I found that I had to brush up a bit more on specifically how film cameras generate photographic images. 

 

For starters, cameras (both film and digital) generally use optical lenses to focus a scene onto a focal plane. But rather than use a focal plane array of detectors, a film camera such as Zapruder's 8 mm Bell & Howell uses a light-sensitive film that is processed (i.e., “developed”) after the entire film is exposed. We are not so much interested in the details behind the chemical processing involved in developing the film into photographs, but rather a physical understanding of the relevant remote sensing principles. Generally speaking, camera films from the past required an adequate amount of light exposure to capture an image properly. One could adjust the shutter speed on their still-camera to control the amount of exposure, but for dynamic (moving) scenes there was not a lot of flexibility given that longer exposure would lead to greater blurring, and thus a movie camera like Zapruder’s would usually have a set shutter speed. For dim indoor lighting conditions, one was forced to use either a flash for a still-camera or a light-bar (or similar flood light) for a movie camera.

 

Film Response

 

And yet, there was one additional consideration that every serious cameraman from the pre-digital era (e.g., Abraham Zapruder) had to worry about, and that was the film speed. A photographer would need to purchase film optimized for the lighting conditions that he or she expected to encounter. Note that film speed is not the same thing as shutter speed. For dimly lit outdoor conditions (e.g., overcast skies) one would select a higher-speed film (indoor conditions also requiring the light-bar), and for brightly lit outdoor conditions (e.g., bright sunshine) one would use a slower film. So, what exactly is meant by film speed? Briefly, the film speed refers to the rate at which the film responds to light exposure to create the photographic image of the scene. It is this aspect of the film behavior (i.e., the film response) that is essentially the basis behind the ad hoc “blur-hypothesis” promoted by Thompson in his book.

 

This led me to research the type of film Zapruder specifically used and what its characteristics were. I found through secondary sources that the film was Kodachrome II outdoor reversal film. This type of film was not high-speed and thus the response time was optimized for bright lighting conditions (e.g., direct sunlight) thereby preventing overexposure. But to learn more, I consulted directly with Kodak photographic expert Roland Zavada (both by email and over the phone, again in early 2019 and early 2021). Zavada was a 40-year veteran at Eastman Kodak headquarters in Rochester, New York, who was called upon to authenticate the Zapruder Film for the ARRB in 1998. In his own words, Zavada had the following to say in response to my query about the Kodachrome II movie film:

KODACHROME II film is intentionally non-linear. Three different emulsion response characteristics describe the toe, mid-scale and shoulder for each of the three-color sensitivities (nine emulsion components). The goal was to reproduce a scene close to the human eye response. The only loss of toe response in the Zapruder film was primarily in the reflections of the windshield, glass or bumper [of the limousine].

By “toe response,” Zavada is referring to the flattening out of the curve near the high end of the exposure response (as will be seen in Fig. 3, discussed more below). Upon my request, Zavada graciously provided me with technical specifications of the Kodak Kodachrome II film. Specifically, on page 2 of the document was the figure that I was seeking, namely the “modulation transfer function characteristic curves” which I shall refer to in this review article simply as the film response function (FRF), this being descriptive jargon I am more familiar with. As its name implies, the FRF simply describes (quantitatively) the response of the film to light exposure. I was able to digitize the data from this figure, which then allowed me to use it for purposes of studying what kind of impact the film response might have on what we observe between Z312 and Z313.

 

Figure 3 shows the digitized data along with some processing steps that I have undertaken to facilitate the analysis. The upper left plot shows a reproduction of the original plot in the Kodak film specs sheet, which shows the film response (i.e., “density” on the y-axis) versus the “log exposure” (in lux-seconds on the x-axis). Translated, this plot basically shows the film opaqueness (i.e., “density”) versus the logarithm of the cumulative light intensity over the exposure time. The exposure time of a camera such as Zapruder’s is over a short time interval (i.e., 0.025 s), so under most circumstances the collected light is relatively constant throughout the scene, leading to a sharp, accurate 2-D image of the scene. The upper right plot shows the same data but plotted on a linear x-axis; here the non-linearity of the camera response is more readily apparent. So far, so good.

 

Next, we can note how the red, green and blue curves all exhibit the same degree and magnitude of nonlinearity. Therefore, to simplify things, I converted these curves to grayscale by taking the mean, recognizing that white is the combination of these three primary colors. But I also wanted to get away from the “density” unit used by the film engineers. Basically “density” translates to “opacity,” and “opacity” translates to “darkness” (i.e., the higher the density, the opaquer the film, the less light can pass through it, and thus the darker it appears when projected onto the movie screen). So, what I did was take the “negative” of the grayscale density—thus, dense (opaque) areas are now indicated with smaller numbers and vice versa; this is shown in the lower left panel.

 

As a final step, I wanted to get away from the remnant density magnitudes, shown as real numbers between 0 and 4 on the previous three plots, and create a more generic variable called “brightness” that would have a more straightforward physical interpretation as being a unitless real number between 0 and 1 (much like “reflectance” or “transmittance” in remote sensing). Zero brightness (0) is pure black, and unitary brightness (1) is pure white. This is shown in my final plot in the lower right, which constitutes the final film response function (FRF) that I shall be using from here on out.

 

Before continuing, it is worth pointing out a few salient features of the FRF in the lower right panel of Figure 3. First, note how it is difficult to “saturate” the film, that is cause a “white-out” (i.e., brightness = 1) of the film, just as we might expect from a film designed for use in direct sunlight. Indeed, the data points do not even go to one, and at the higher-end range of exposure, the curve flattens out. On the flip side, note how at the low end of the exposure range there is a region ≤ 0.165 where the brightness drops to zero (0), that is, “black-out.” While film density here may not have been completely opaque (given we do not have data below that exposure range), it nevertheless was quite close. [21]

 

 

Figure 3 - Kodachrome II color reversal film response function: (a) original color film density data as a function of log exposure, (b) same except linear in exposure, (c) grayscale negative of density versus linear exposure, and (d) grayscale brightness, where “brightness” is a real number between 0 and 1, with 0 representing pure black and 1 representing maximum white. Data source: KODACHROME 25 Movie Film (Daylight)/7267, Technical Data, Color Reversal Film H-1-7267, April 2002 (Courtesy of Roland Zavada).

Using Information from Z312 to De-Blur Z313

 

Equipped with the Kodachrome II FRF data, I was able to use the information contained in the sharp, un-smeared frame Z312 to study quantitatively the potential impact of panning error on the following frame, Z313. This was achieved as follows.

 

First, I converted the original color images (Z312 and Z313) to grayscale numeric “brightness” arrays of type REAL8 [0,1], denoted B312 and B313; these are displayed in Figs. 4a and 5a. The Z312 brightness array was then transformed to “exposure” using the inverse FRF (cf. Fig. 3), to which horizontal smearing was applied by taking a boxcar filter that approximated the same magnitude of camera panning error. The actual panning error was very slightly oblique, but this can be neglected given that it is on the order of ≈1°. The smeared exposure array was then transformed back to “brightness” using the FRF, thus yielding a simulated blurred Z312 image that accounts for the film response (Figs. 4b and 6b). Neglecting noise, a “smearing point-spread function” (PSF) for the scenes in both Z312 and Z313 can be estimated from the blurred Z312 by dividing it by the original non-blurred Z312. A “deblurred” Z313 brightness image (which will be denoted Z313*) can finally be obtained by dividing the original blurred Z313 brightness by the “smearing PSF” estimated from Z312 (Fig. 5b).

 

Note how real displacements in the deblurred Z313* are seen as ghost images that were introduced by applying the PSF derived from the Z312 “blur” onto displaced figures in Z313 — these “ghosts” are identified in the red circles in Fig. 5b.

 

But more importantly, the deblurred Z313* can be used to estimate the effect of the Z313 blur on motion artifacts by examining the brightness differences between the two images, ΔB. Because we have been working in real numbers, we can simply subtract the Z312 brightness array from both Z313 and Z313*. Brightness differences between Z312 and Z313 will contain real changes in the scene (i.e., movement of objects) along with noise and artifacts brought on by the image smearing in Z313. However, because we have now “deblurred” Z313*, we should expect the differences between the two frames caused by smearing to be reduced, while leaving only the real changes between the frames.

 

This is exactly what we observe in Figure 7, where the left-hand top and bottom plots (d) and (e) show the ΔB values between frames Z312, and Z313 and Z313*, respectively. (For reference, Z312, Z313 and Z313* are shown in Figs. a–c, respectively). In Fig. 7d we see the edges of recognizable figures from the original images (e.g., the President and First Lady, Gov. Connally, the chrome window frame, the back of the seat), many of these being artifacts caused by smearing of the edges.

 

However, some of the differences are real changes between the two images, the most obvious being the gruesome effects of the cavitation wound left in the wake of the high-energy projectile, namely the explosion of the President’s head. Interestingly, this image dramatically reveals that the vast majority of the mass from the explosion is forward and upward, this being consistent with a high-powered rifle shot from behind. We can also clearly see a bright band on the back of his head, indicating brighter pixels in Z313. This bright band is the result of his head moved forward (≈5 cm), revealing the much brighter background scene behind it, as is obvious when viewing the two aligned Zapruder frames (Z312 and Z313) side-by-side, or in a back-and-forth toggle (as in Animated GIF #1).

 

Or is it? Could the bright band merely be an “illusion” due to the smearing effect, as is the case for the chrome window frame or Connally's forehead, as claimed by Thompson and Wimp?

 

 

Figure 4 - Zapruder frame 312 rendered in grayscale brightness: (a) Z312, and (b) blurred Z312 obtained from horizontal smearing (boxcar filter) applied to the exposure array simulating the effect of panning error while accounting for the film response. “Brightness” is defined as a dimensionless real number ranging from 0 to 1 (0 representing pure black, 1 representing pure white), thus changes in the grayscale shade in the picture elements represent real changes in the original photograph film densities.

The Blur Non-Illusion

 

To address this claim of Thompson’s, we now turn to the lower right plot (Fig. 7e), where we can see that the edge blurring artifacts have been significantly reduced. The remaining artifacts (e.g., associated with the bright reflections off the chrome window frame) are mostly horizontal in nature, this due to the fact that the actual panning error was very slightly oblique, whereas I used a purely horizontal boxcar filter (as described earlier).

 

However, we may now take note of the fact that the bright band on the back of JFK’s head is still present. If the bright band in Fig. 7d were the result of blurring, then it should no longer be present in the deblurred image (Fig. 7e). Because the head would have been stationary, there would be no difference between these two frames, and thus no bright band remaining. This quantitatively indicates that the apparent movement of the back of JFK’s head between Z312 and Z313 is, in fact, a real movement, and not merely a blurring artifact brought on by the nonlinear film response. [22] Q.E.D.

 

 

Figure 5 - Zapruder frame 313 rendered in grayscale brightness: (a) Z313, and (b) a “deblurred” Z313 (denoted Z313*) obtained from the point spread function (PSF) estimated from the blurred Z312 (Fig. 4b). Real displacements of objects between the two frames are visible as “ghost images” (highlighted within the red ellipses) that were introduced by applying the PSF derived from Z312 onto displaced figures in Z313.

Of course, in the end, this would only seem to be common-sense. But what about the glint off the shiny metallic chrome window frames, and the lamppost seen in later Zapruder frames? Why is it that blurring can cause the appearance of translational “motion” with these features?

 

The reason for this goes back to the Kodachrome II film response, along with the specifics of the scene being filmed.

 

In Z313, the scene behind JFK's head is a sunlit grass lawn along with the brighter pink suit of Mrs. Kennedy and a bit of the street curb behind. All three of these are what we call diffuse reflectors, meaning that they scatter light (from the sun) into a multitude of directions, thereby “diffusing” it.

 

This is not the case for the metallic chrome window frames on the limo. These act more like specular reflectors, meaning that they reflect much more like a mirror, thus directing a small image of the solar disk directly into our field-of-view (which can be quite blinding, as most people are familiar). While the Kodachrome II film is designed for bright outdoor lighting under direct sunlight, it is not designed to handle mirror reflection of the solar disk—under such conditions, the film saturates, and this causes the distinct streaks that we see in the blurred image.

 

 

Figure 6 - Zapruder frame Z312 rendered in grayscale brightness and zoomed on limo occupants: (a) as in Fig. 4a, and (b) with horizontal smearing applied to simulate blurring as in Fig. 4b.

And what about those later frames with the lamppost?

 

To answer this, I carefully examined the full-resolution (1-GB) images (courtesy of Max Holland and the Sixth Floor Museum at Dealey Plaza), and what’s going on there is something similar, namely that the smooth background asphalt has a glare, something that we have all experienced when driving a car into the late afternoon or early morning sun… before we reach for our polarized sunglasses.

 

The same goes for these later frames. The background asphalt has a horizontally polarized forward reflectance that increases under this particular view geometry (i.e., at large incidence and reflectance angles), which results in an enhanced “shiny” reflectance of the incident light from the sun that can be quite blinding. The Dealey Plaza lamppost, on the other hand, was practically black due to the dark-colored paint along with the scene illumination geometry that put the side facing Zapruder's camera completely in shadow. Couple these (asphalt glare and a dark-colored lamppost in shadow) with the far wilder two-axis smearing in these later frames brought on by Zapruder’s terror, along with the much faster moving limousine, and we end up with the highly distorted image of a discontinuous lamppost.

 

 

Figure 7 - Zoomed Z312 and Z313 brightness comparisons (rendered in a faux-sepia colormap): (a) Z312 brightness, (b) Z313 brightness, (c) Z313* (deblurred) brightness, (d) brightness difference between Z313 and Z312, defined as ΔB=B313 - B312, and (e) brightness difference between deblurred Z313* and Z312 (Fig. 4b). Lighter shades in (d–e) denote areas where Z313 was brighter, and darker shades denote places that were darker. Note that the edges in (d) resulting from the blur in Z313 are reduced after deblurring in (e). However, note also that the bright edge on the back of JFK’s head remains in (e) even after “deblurring” Z313, thus indicating the same degree of forward movement.

Parting Comments

 

The foregoing discussion on the “Blur Illusion” was meant to be a technical (and quantitative) response, translated to the best of my ability into plain English, to an ad hoc hypothesis, namely that the forward snap of President Kennedy’s head, a key finding in the key Zapruder Frames Z312–Z313 that is consistent only with a shot from behind, [23] coincidentally just happens to be an “illusion.” While this is not technically “alterationist,” it is what I would still consider a soft form of the same category, with the same ends in mind but employing a different means to get there. [24]

 

When I first heard about the “blur illusion” in 2018, I had hoped that Thompson would come to his senses and not publish this spurious material (and I had hoped that I wouldn’t have had to go through this). But ultimately, much like the “acoustics evidence,” my attendant study of the Zapruder camera and Kodachrome II film ended up becoming an interesting academic pursuit in its own right. The results confirm the stark reality of a sudden, 5 cm (2 in) forward motion of the President’s head from Z312 to Z313, reinforcing the fact that President Kennedy was shot in the head from behind.

 

Should the reader have any lingering doubt (perhaps they trust “common sense” more than the technical mumbo-jumbo of another scientist), then they need not worry. They may simply see for themselves that the Z312 blurring causes no “illusion” of isolated forward motion on JFK’s head [Animated GIF #2], and JFK’s head still snaps forward in the same manner as before [Animated GIF #3].

 

 

Animated GIF 2 - Comparison of Z312 (original) and Z312 (with horizontal smearing).

 

Animated GIF 3 - Comparison of Z313 (original) and Z312 (with horizontal smearing). Note forward motion of President Kennedy's head.

 

Summary and Conclusions

 

Like his classic 1967 Six Seconds in Dallas (SSID), Josiah Thompson’s new book Last Second in Dallas is, to quote Luis Alvarez (on SSID), “beautifully printed, with excellent photographs and carefully prepared graphs,” containing high-quality (albeit brutally graphic) color images, including Zapruder frames referred to in the text. Years in the making (since well before 2013), Thompson’s autobiography and revised Dealey Plaza assassination scenario may be summarized, at least in part, as a compendium of contradictions. Whether or not this was intentional is not fully clear, but the reader is nevertheless given a glimpse of a unique individual who has lived a fascinating life. Regardless of my criticisms in this review article, Thompson remains at the forefront of a dwindling number of rational WC critics who have remained resolute in a futile 55+ year search for a Ghost on the Grassy Knoll.

 

Thompson tells us that he started as a rather apolitical person from a conventional family, serving two years in the U.S. Navy before becoming a professor of philosophy at Haverford College, during which time he would evolve into a part-time political activist in the 1960s. Through a series of serendipitous encounters (beginning with Vincent Salandria, an ACLU attorney and first-generation buff whom he had met after getting arrested for distributing anti-war literature), he would become a nationally recognized critic of the Warren Commission after the publication of his original book, Six Seconds in Dallas. Intrigued by the investigation of President Kennedy’s assassination, he would eventually abandon the academic profession altogether, admitting to never “believing” in philosophy, thereby leaving behind an otherwise comfortable career as a tenured professor to enter the world of private investigators, a career path precious few can lay claim to, if any. As a private eye, Thompson would embark on a multi-decade career pursuing similar such controversial cases, including even, for example, the defense of Oklahoma City bombing terrorist Timothy McVeigh. But eventually he would come full circle and return to “his case,” the Kennedy assassination, which has led the reader and me to this juncture.

 

On the one hand, Tink Thompson can be quite rational and grounded, underscoring the primacy of evidence in the case and recognizing the limitations of eyewitness testimony. He understands the futility of denying key evidence, whether it be the autopsy materials or the Zapruder Film, elsewhere effectively demolishing Zapruder Film “alterationist” bunkum. However, on the other hand, it is simply baffling that he would deny his own key contribution to the case, namely a critical measurement from the Zapruder Film that would provide kinematic data amenable to dynamical analysis, eventually serving as an additional independent piece of evidence that the shot at Z313 originated from behind.

 

Rather than embrace that conclusion at face value, he has fallen for an esoteric ad hoc hypothesis that the Zapruder Film photographs in question (Z312–Z313) don't in fact show what they show. Thompson heavily bases his “revised scenario” on the DPD recordings as acoustics evidence, but we have seen that there have actually been three general categories of arguments (accumulated over the decades) that have thoroughly discredited the “acoustics evidence,” multiple times over.

 

Thompson is curiously uninterested in concrete scenarios about who the alleged conspirators were. Although he is gracious to independent investigators like Michael O’Dell and Steve Barber, he ignores what they have discovered on the subject of the DPD recordings. Thompson has no problem “appealing to authority” when it suits him (e.g., there's even a chapter titled “Journey to the Lair of the Master,” with a good page and a half devoted to the CV of the Master, James Barger), but then goes full ad hominem against anyone else, accusing brilliant physicists of “cherry picking data” and political bias, and all but accusing an NRC Panel of criminal malfeasance.

 

Numerous times throughout the book Thompson refers to “simplicity,” “clarity,” and “elegance” (could there be a trace of the philosophy professor still lurking in there?), never seeming to grasp the fact that his “revised scenario” is a convoluted mishmash of multiple unknown gunmen firing from different directions, escaping the crime scene unseen, and leaving no trace of physical evidence. The victim is shot twice in the head to fit the narrative of a Grassy Knoll gunman, relying entirely on a set of extremely crude DPD radio Dictabelt recordings that contain no discernable gunshot sounds, with the Zapruder Film itself containing arcane “illusions.” Distinguished scientists who have brought reason to bear on the problem are guilty of “complicity.” By comparison, the scenario hashed out in 1964 by a Presidential Commission, while less sensational, exhibits far more simplicity, and to this day is the only scenario that explains all the verified crime scene evidence.

 

Throughout Last Second in Dallas Thompson seems to adopt the approach of a defense attorney who, on the behalf of a client, tries to persuade jurors by creating doubt in the prosecution’s case. I am well aware of the fact that an inverse accusation has been leveled against famed prosecutor Vincent Bugliosi in his book Reclaiming History as being a prosecutor’s case. In the U.S. justice system this approach of prosecution versus defense is the best we have (in the real world) to balance the task of the protection of individual rights and liberties with the task of enforcing equal justice under the law. There were many times during my readings of Last Second where I wanted to stand up and yell “objection!” (and have basically done so in this review article), but the approach to any fact-finding case must be anchored “in the cold light of science,” where established facts are by their very nature objective and thus independent of the politics or personality of the investigators.

 

With the only suspect in the Kennedy assassination long deceased, and because the protection of individual rights belongs only to the living, we should now only be concerned with establishing facts. And the best tried-and-true method for establishment of facts is the scientific method, detached from personal interests, opinions, and politics. But unfortunately, because people inevitably come with competing biases and self-interests, it is precisely the impartial aspect of science that can lead to “controversy” when conclusions inevitably run counter to our partialities. In our own day and age, we continue to see people (both individually and collectively) rejecting science and denying objectivity because they don’t like the conclusions or implications. But in the end, it is up for us to conform ourselves to reality, not vice versa… that is, if we wish to continue to coexist successfully with that reality.

 

Acknowledgments

 

I would like to express my sincere appreciation to Paul Hoch, Michael O’Dell, Max Holland, Francis Corbett, Roland Zavada, Martin J. Kelly, and Louis Girdler for their expert feedback and assistance during the development and writing of this review article. I also wish to thank Dale Myers for reviewing the draft and agreeing to feature it on his website. It has been a privilege for me to have worked and consulted with these professionals in the study of this monumental case. The Sixth Floor Museum at Dealey Plaza is acknowledged for providing access to the full-resolution Zapruder Frames Z405 and Z409.

 

Finally, I’d also like to acknowledge Josiah Thompson for the contributions of his early sleuthing on this case, particularly the Zapruder Film measurements. The Warren Commission got the basic facts of the crime correct, but it is increasingly becoming all too apparent that circumstances beyond their control prevented them from tying down all the loose ends, including the motive of the assassin. I’ve come to believe that many of the early buffs were sincere in pursuing these loose ends (Thompson included), and as a result they have contributed to a more complete understanding of this historical tragedy, allowing us to learn so as not to repeat it. [END]

 

 

--------------------------------------------------------------------------------------

 

Footnotes

 

1. Jacqueline Kennedy Onassis - Wikiquote

2. For more on ballistic shockwaves and the confusion they can cause in identifying the source of a distant gunshot, see Larry Sturdivan’s excellent book, The JFK Myths.

3. Why the Zapruder Film is Authentic - Josiah Thompson

4. National Nightmare on Six Feet of Film: Mr. Zapruder's Home Movie and the Murder of President Kennedy

5. For more on this as it pertains to the JFK case, I highly recommend John McAdams’s book JFK Assassination Logic: How to Think about Claims of Conspiracy

6. Gunshot-wound dynamics model for John F. Kennedy assassination, Heliyon, 4(4), (2018), doi:10.1016/j.heliyon.2018.e00603. Unrelated, but for those curious about the difficulty of the shots taken in Dealey Plaza, I also recommend Sniper Target Tracking Analysis of John F. Kennedy Assassination, J. Assoc. Crime Scene Reconstr., 22 (2018), 11-21.

7. In Newtonian mechanics, a real force is distinguished from a fictitious force, that latter being an apparent force resulting from a non-inertial frame of reference (e.g., a centrifugal force is a fictitious force). Thus, if we take the earth’s surface to be an inertial frame, the force you feel pushing you forward in a braking automobile is a “fictitious force,” whereas the force of gravity holding you in your seat is a “real force.”

8. See Gunshot-wound dynamics model for John F. Kennedy assassination and Corrigendum to “Gunshot-wound dynamics model for John F. Kennedy assassination” [Heliyon 4 (2018) e00603]: Heliyon Figures 7 and 8.

9. Donald Thomas is a research entomologist with the US Department of Agriculture (USDA).

10. In Thompson’s words, “I came to see that I was trying to put together a puzzle in which some of the pieces did not belong.”

11. I have some expertise on this topic (image analysis), and in my opinion the “blur illusion” is the most egregious contention of the four “puzzle pieces.”

12. Lucien C. Haag, AFTE Journal, 48(2), Spring 2016.

13. Synchronization of the acoustic evidence in the assassination of President Kennedy 

14. Excerpt: The Kennedy Half Century – Sabato's Crystal Ball

15. The Kennedy Half Century Acoustical Analysis of November 22, 1963 Dallas Police Recordings

16. Further research analysis and commentary on the Dallas Police Department recordings of November 22 1963.pdf (thekennedyhalfcentury.com)

17. According to Officer H. B. McLain (riding the suspect “bike with the mic”): “We were ordered to be on Channel 2. I [McLain] was on a Channel 2 assignment that day.” Vincent Bugliosi (in his Reclaiming History Endnotes, p. 177) comments that “The importance of McLain being ordered to be on Channel 2 and remembering many of the Channel 2 transmissions during the motorcade cannot be overemphasized. It literally, by itself, destroys the HSCA’s acoustic conclusion of a fourth shot—at least if the committee is basing it on McLain being the cyclist with the open microphone.”

18. Cf. 38a. The nonissue of blurring by Prof. Kenneth Rahn. Rahn, an atmospheric chemist, had independently drawn conclusions similar to my own about President Kennedy’s reactions in the Zapruder Film after the fatal shot, albeit he used simpler methods that were more pedagogical in nature.

19. Incidentally, this is why there had to have been a “jet” recoil effect—because it is a necessary consequence of momentum conservation for the observed directional mass expulsion in this particular case… nitpicking of melon/coconut demonstration experiments notwithstanding.

20. Cf. Gunshot-wound dynamics model for John F. Kennedy assassination.

21. Incidentally, this is one of the reasons why the back of Kennedy’s head looks black in some of the Zapruder frames; it’s not because it was “altered” to “cover-up” a “gaping wound” on the back of his head. On this point I suspect Thompson would again agree, but nevertheless this conjecture of Zapruder Film alterationists has become a commonly repeated myth. Roland Zavada pointed out to me that the ambient lighting conditions were not ideal given that the lighting source (i.e., the sun) was in front of the camera rather than behind it, thereby creating shadows on the sides of the subjects exposed to the camera. According to Zavada: “The non-ideal scene illumination accounts for the black patch on the back of JFK’s head.” In addition to these considerations, 3-D modeling has also shown that the shadow on the back of JFK’s head is fully consistent with the geometry and ambient lighting conditions at the crime scene [see Farid, 2010: A 3-D lighting and shadow analysis of the JFK Zapruder Film (Frame 317)].

22. Not to reiterate the fact that it is his entire head that moves forward, not simply the back edge.

23. For the rationale behind this statement, see Gunshot-wound dynamics model for John F. Kennedy assassination.

24. While not entirely clear, Wimp and Thompson appear to have admitted that there was, in fact, a forward head snap, but that it was significantly less than 2 inches (i.e., about half as much). This hair-splitting was not important to the essential question as to whether blurring causes illusory motion, but apparently Thompson believes that this remaining forward snap can be attributed to the braking of the limo. However, the limo braking had precisely zero to do with the isolated motion observed on JFK’s head between Z312 and Z313. As pointed out above, there was no detectable motion on the other limo occupants at this moment. A fictitious force brought on by braking would cause all the limo occupants to react in the exact same manner, no different from car crash test dummies. That’s why seatbelt laws apply to all motor vehicle passengers (whether they are asleep or awake, healthy or sick, swooning or non-swooning…).

 

NOTE: Figure 2 was updated on December 30, 2021, to correct mislabeling found in the original figure; this change does not involve data and does not impact the conclusions of the article