Debugging Bugliosi

by Don Thomas

INTRODUCTION

Photo of the actual dictabelt on which the gunfire in Dealey Plaza was recorded
Photo of the actual dictabelt on which the gunfire in Dealey Plaza was recorded.

Vincent Bugliosi's treatment of the acoustical evidence is turgid with factual errors, many more than can be enumerated in a ten page essay, so I will cite only the most significant. Acoustical experts for the HSCA in 1978 determined that the assassination gunfire was captured on a police recording of a broadcast emanating from a motorcycle cop's radio microphone, and that the shots included one from the infamous grassy knoll. Although Bugliosi cites from the technical reports he clearly has no grasp of the technical issues. On such details as the manner in which the motorcycle microphone became stuck open, on the manner in which fragments of broadcasts cross-talked from one channel to another, and the manner in which ambient sounds are distorted by the radio circuitry, Bugliosi's account is substantially wrong. The facts and logic that support the acoustical analysis are misconstrued beyond recognition. Although Bugliosi relies heavily on interviews to get his points across, it is perhaps telling that Bugliosi's interviewees did not include any of the scientists that worked on the acoustics, on either side of the issue. Instead he cites extensively from interviews with persons he considers to be the experts on the acoustics in the JFK case: Dallas cops and a Rock musician named Steve Barber. Like a moth drawn to a flame Bugliosi seems bedazzled by any critical opinion that supports his own.

Most crime investigators consider forensic evidence, that is, scientifically analyzed physical evidence, to be the "best" evidence, and would seldom give equal weight to an eyewitness account, but Bugliosi does. For example to refute the acoustical evidence Bugliosi cites from the memory of the policeman, H.B. McLain, whose stuck microphone picked up the sounds that are on the police recording. McLain's memory of events is not in accord with the sequence of events evident in the recordings. Thus, Bugliosi declares of McLain's memory "It literally, by itself, destroys the HSCA's acoustic conclusion…"

BUGLIOSI AND OFFICER MCCLAIN

Model AT2C Dictaphone Belt Recorder like the one used by DPD to record channel one radio traffic
Model AT2C Dictaphone Belt Recorder like the one used by DPD to record channel one radio traffic.

The DPD was using two radio channels on the day of the assassination. The acoustical evidence of gunfire was recorded on the primary police channel designated as Ch-1. Asked in testimony, without warm-up or warning, if he could remember what channel he was tuned to on the day of the assassination, Officer McLain responded that he would normally be on Ch-1 but that he really didn't remember (5HSCA630). The HSCA left it at that. But in subsequent interviews, having learned why the HSCA was so interested, McLain now insisted that he could not have been on Ch-1. Bugliosi insists that McLain was somehow "duped" and "suckerpunched" in his earlier testimony, and that it his refreshed memory that absolutely refutes the forensic evidence. Bugliosi is picking gnat-shit out of pepper.

Filmed evidence shows that McLain's was the only motorcycle in the right position to have captured the sounds of gunfire manifested on the recording. But rather than rely on the films, Bugliosi insists that McLain's memory of where he was during the shooting is sufficient to refute the acoustical evidence. The acoustical evidence requires that the open microphone was on Elm Street at the time of the fatal head shot. McLain remembered that he was on Houston Street about half way between Elm and Main when he heard the gunfire, and that he stopped at that point (5HSCA629). However, one of the salient events in the immediate aftermath of the shooting, one that McLain claims to have seen from his position on Houston Street (in Sneed p. 163), was Mrs. Kennedy's brief sojourn out onto the trunk of the limousine. Between 3-5 sec after the fatal head shot, a sequence shown in the Zapruder film, Mrs. Kennedy ventured briefly on to the trunk to retrieve a piece of her husband's skull. Given the crowds of people and the structures on the south side of Houston Street, it seems unlikely that McLain could have had a clear view of the event from where he claimed to have been, as opposed to the unobstructed view that he would have had on Elm Street where the acoustical evidence places him. More importantly, another motorcycle officer, Jimmy Courson, also saw Mrs. Kennedy on the trunk of the limo, but recalled that he was on Elm street, actually just making the turn on to Elm from Houston at the time. In point of fact, newsreels of the motorcade immediately before and immediately after the shooting show that Courson was several car lengths behind McLain, and thus Courson's account regarding his location directly conflicts with that of McLain. Both officers cannot be correct, and while this doesn't prove that McLain's memory was wrong, it does show how unreliable eyewitness memories of events can be, especially coming scores of years after the event.

However, in this instance a film taken by Elsie Dorman shows a motorcycle officer turning the corner of Elm and Houston at exactly the same moment that Mrs. Kennedy was on the trunk of the car. Robert Groden originally, and Dale Myers subsequently, assumed that the officer was McLain. But the film lacks any critical detail that would allow one to identify the officer as McLain or Courson though the circumstances of the film (the timing and the other vehicles) are such that it has to be one or the other. So, the truth of the matter is, the filmed evidence is exactly in accord with Courson's account, if it is Courson, and directly contradicts McLain's account no matter who it is. Bugliosi does not discuss this issue, perhaps because of the direct conflict to his star witness. For the acoustical evidence to be true, the open microphone had to be on a police motorcycle just ahead of the mayor's car when the Dorman sequence begins. In the Dorman film, there is no motorcycle between the Mayor's car and the cop seen rounding the intersection. If the officer in the Dorman film is Courson, the only place McLain can be at that time, is just ahead of the Mayor's car, exactly where the acoustical evidence places the motorcycle with the open microphone.

Aerial view of Dealey Plaza
Aerial view of Dealey Plaza, with TSBD building at top center, grassy knoll to its left, and Elm Street curving under both.
Courtesy Stewart Galanor.

Ignoring the filmed evidence is also necessary to crediting Bugliosi's version of the siren issue. Bugliosi cites McLain's memory that he turned on his siren immediately when he raced off to Parkland Hospital. Sirens can be heard on the Ch-1 recording but not until a full 2 min after the shooting, whereas, McLain can be seen in films beginning his run for the Hospital approx. 25 sec after the shooting. Bugliosi rejects any suggestion that it is even remotely possible that McLain's memory on this issue was also faulty. Moreover, Bugliosi, raises the old argument, first raised by Anthony Pellicano, that if McLain's were the open microphone, it should have picked up the sirens from the lead vehicles which certainly did turn on their sirens before heading for the hospital because they are audible on Ch-2. On this point the acoustical experts, whose opinions are eschewed by Bugliosi, explained that because of the distance and direction (the lead vehicles were at least 300 ft in front of McLain), and with the motor noise dominating the ambient sound levels at the microphone, the recording would not have captured discernible sounds of the sirens. Bugliosi therefore accuses the acoustical experts of sophistry, not because of any better acoustical knowledge on his part, but because Bugliosi insists that McLain was only 120 ft behind the limousine.

But the 300 ft estimate was based on the hard filmed evidence, not on sophistry. Firstly, the acoustical analysis placed the open microphone 141 ft behind the limousine at the time of the fatal head shot, not 120 ft. At that time the limousine was on Elm Street exactly 203 ft west of the corner of Houston street and 250 ft east of the underpass at the west side of the plaza. At the time of the shooting, the lead police car was around 150 ft ahead of the limousine approaching the underpass. In fact, there is a broadcast on Ch-2 wherein the chief of police in the lead car announced that he was "approaching the triple underpass." Based on broadcasts common to the two channels to synchronize events, the sounds acoustically identified as gunfire were deposited on the Ch-1 recording at that same moment. The Zapruder film shows that the lead car then came to a slow roll within the underpass and waited for the limousine to catch up. The limousine reached the underpass 8 sec after the fatal head shot. The limousine pulled alongside the lead car somewhere in the underpass and the two cars were in this position as they exited the underpass. A photograph of this scene was taken by Mel McIntyre who was standing at the entrance to the Stemmon's freeway. After a brief conference, surely no more than a few sec, the two cars began their race to the hospital. Thus, Chief Curry broadcast the orders to the motorcade, "Go to the Hospital officers, Parkland Hospital." This broadcast on Ch-1 comes 20 sec after the earlier "approaching the triple underpass" broadcast. Logic and circumstances dictate that at this point, beginning the run to Parkland Hospital, is when the lead car or the limousine, or both, would have turned on their sirens, and in fact, a siren can be heard in the background of Curry's order to go to the hospital.

Photo taken by Wilma Bond, depicting officers McLain and Courson on Elm Street in front of the grassy knoll
Photo taken by Wilma Bond, depicting officers McLain and Courson on Elm Street in front of the grassy knoll.

And where was McLain during all this? An important position marker is the parked motorcycle of officer Bobbie Hargis. Hargis was positioned behind and to the left of the President and thus when his motorcycle was struck by brain matter, according to Hargis nearly knocking him over, he perceived that the shot had to have come from his right front. Dismounting and parking his motorcycle, Hargis with gun drawn, advanced on the grassy knoll. Hargis' motorcycle is parked very close to where JFK was struck, that is, about 250 ft west of the Underpass. McLain and Courson several car-lengths behind him, appear in several films beginning about 15 sec after the head shot. In the Bell film McLain traveling slowly reachs and passes Hargis's motorcycle, Courson catching up and now close behind him. Gary Mack matched the Bell film to the Weigmann and Z-films to show that McLain reached the Hargis motorcycle at 21 sec after the head shot – at approx the same instant that Curry made his broadcast ordering the motorcade to go to Parkland Hospital, and presumably would have turned on his siren. At that moment McLain is at least 300 ft behind the lead car and the limo, exactly as the acoustical experts had argued – not 120 ft as Bugliosi insists – and on the opposite side of the Underpass.

Another point is relevant. McLain remembers hearing Curry's order to go to the hospital (5HSCA635). Because that broadcast emanated from Ch-2, Bugliosi cites it as evidence that McLain must have been on Ch-2. It is thus perhaps relevant that at the moment Curry made his broadcast, McLain was idling in the immediate vicinity of, and in between, both Hargis' and Courson's motorcycles, either or both of whom could have been tuned to Ch-2. In fact, just seconds after passing Hargis with Courson now pulled up beside him, the two raced off in tandem for Parkland Hospital. The Ch-1 recording does not have a siren sounding at that point, but it does have the motorcycle engine revving up.

On another issue, McLain also appears in a film that was taken by Robert Hughes shortly before the shooting, turning from Main Street on to Houston Street. Without providing any supporting details, Bugliosi cites the work of researcher Richard Sprague who estimated that this segment of film corresponds to a time of only 3.5 sec before the first shot. Because McLain was at that point about 170 ft away from the acoustically determined position of the open microphone at the first shot he would have had to travel at an implausibly high rate of speed to make up the distance. But Sprague's estimate was only a guesstimate. There is no connection, direct or indirect, between the Hughes film and the Zapruder film, or any other film, that would allow an accurate estimate of the time interval between when McLain appears in the Hughes film and the first shot. The estimate depends entirely on the relative positions of the motorcade vehicles in the Hughs film and the later Zapruder film and extrapolation of the vehicle speeds during the time when they are not visible. The different vehicles in the different films are traveling at different speeds. Depending on which speed one chooses to impose for the extrapolation, McLain had as little as 3 sec (too little time) or as much as 7 sec (plenty of time). But this soft evidence metamorphoses into hard evidence under Bugliosi's pen.

BUGLIOSI AND JIM BOWLES' "REBUTTAL"

Bugliosi also relies heavily on an anti-acoustics diatribe written by Jim Bowles, stating (on p.178),

"[Jim] Bowles...was never called to testify before the HSCA. His lengthy 1979 rebuttal offers such a wealth of information that had the HSCA been privy to his insights, it might have saved itself a great deal of time, money, and embarrassment."

Yes, Bowles as a supervisor of the Dallas police communications department did have a wealth of useful information concerning the police radio system and its operation and would have been a helpful witness to the HSCA. Most important was his knowledge of the broadcasts, the officers in the motorcade, and especially the crosstalk, of which the acoustical experts were unaware. But as far as the acoustical data, Bowles' opinions were less than expert. For example, with reference to the seeming contradiction that the gunshots are indiscernible but that at one point a person can be heard whistling on the police recording. Bowles remarks,

"So you can hear someone whistling but you can't hear a rifle shot. Who's going to believe that?"

The answer is, anyone who understands what Automatic Gain Control does. The AGC enhanced voice communications by automatically dampening very loud sounds, such as gunshots. Of course it would not dampen sounds at the amplitude of human voices such as someone whistling. If it dampened normal sounds it wouldn't be able to transmit voices!

Similarly, Bowles cites as evidence that the open mike could not have been in Dealey Plaza the fact that there are no discernible crowd noises during the motorcycle segment. The HSCA experts explained that the motorcycle motor noise drowned out the crowd noise. To which again, misinterpreting this explanation to mean that the acoustical experts were contending that the radio system couldn't pick up crowd noise, Bowles counter-argued that there were crowd noises audible on the Ch-2 record. But of course the Ch-2 record was not dominated by the sound of a motorcycle motor!

Along the same lines, at Endnote 157:

"...shouldn't we expect to hear the very loud gunshots on the recording..."

This is a critical misunderstanding. The HSCA experts never said that the sounds are not audible. Just that the sounds are not discernible as gunshots.

"To the human ear, the tapes and Dictabelts contain no discernible sounds of gunfire." (HSCA Final Assassinations Report, p.67)

Discernible is not the same thing as audible. An analogy would be trying to see a white moth in a snow storm. It is not that the moth is invisible, it is just hard to distinguish. The AGC of the police radio system automatically dampened loud noises so that they were about the same loudness as the other noises in the recording. At footnote 160 Bugliosi cites Johann Rush and Jim Bowles that the gunshots should be clearly audible because they had heard gunshots on other recordings. Apparently Bugliosi expects the reader to accept the inexpert and untested opinions of Rush and Bowles over the results of the experiments conducted by the acoustical experts. BBN obtained the same equipment that the police used in 1963 – radios, amplifiers, receivers: and played gunshots into them and measured the dampening effect of the AGC on the amplitude and waveform of the sound to confirm what their knowledge and experience had taught them to expect.

Bowles' rebuttal is replete with such non-scientific arguments. Among them that he could tell the speed of the motorcycle by its sound, and even the type of motorcycle by its sound. He offered no clues as to what the defining acoustical characteristics were in support of these claims. Bowles then used these magic powers in his argument saying,

"When the motorcade was on Main Street the motorcade was frequently reduced to walking speed, barely able to go faster than 3 to 3½ mph, with a top speed of 5 mph. Yet during the same period the speed of the motorcycle with the open microphone was determined to be operating at 25 to 35 mph, averaging approximately 30 mph."

The HSCA experts never determined the speed of the motorcycle at any time and Bowles offered no evidence for the unlikely claim that the motorcade was traveling at 3 mph on Main street. All of these values cited by Bowles were just made up by him and repeated by Bugliosi. Like a Rudyard Kipling story, it was just so!

COURT TV AND ACOUSTIC MATCHING

On page 172 Bugliosi cites the study of Robert Berkowitz who announced on Court TV that his analysis of the police recording demonstrated that there was no match between the grassy knoll test shot and the suspect patterns on the police recording as the HSCA's acoustical experts had reported. Had Bugliosi interviewed him, Berkowitz would have admitted that he had flubbed the comparison by using the wrong recorder speed. It is not uncommon that an advocate of a position will accept without question any evidence that accords with their preconceptions. But it appears that Bugliosi chose not to question anyone familiar with the technical issues in this case. Instead, Bugliosi relied on his own interpretation of the technical reports. Not surprisingly then, Bugliosi grossly misrepresents the acoustical analysis, on both sides of the issue, though in all cases the misrepresentation falls on the side of his position.

For example, during field tests the acoustical experts fired test shots in Dealey Plaza and compared them to six suspect sounds that were on the police recording. A total of fifteen test shots matched to five of the suspect sounds. Bugliosi's interpretation of this result appears on page 163:

"Although all fifteen patterns were deemed significant, it was clear to BBN scientists that some of them were invalid: that is, they did not represent true gunfire from the book depository or the grassy knoll."

Bugliosi thus confounded two separate issues. One issue faced by the scientists was the possibility of a false positive in the sense that one or more of the suspect sounds on the recording might not be true gunshots, just as Bugliosi writes. But the second issue was, that if any or all five of the suspect sounds were truly gunshots, their origin might not have been correctly determined. It was clear to the BBN scientists that some of the 15 detections were invalid with regard to this second issue. But Bugliosi misrepresents the reasoning of the BBN scientists when he applies this degree of clarity to the first issue. Consider this. Four test shots matched to the suspect sound eventually determined to be the first shot. The matching test shots were recorded at two separate (but adjoining) microphones. Moreover, while three of the matching test shots were fired from the book depository, one was fired from the grassy knoll. Applying the strict logic that only one combination of shooter position and microphone position can be correct, then, three of the four matches are invalid, but only in the sense of the second issue. Bugliosi completely misconstrued this concern not only to mean that some of the sounds were not gunshots, but that the acoustical experts knew it! Apparently Bugliosi missed the part of Barger's testimony where he explained that his array of microphones was a net set to capture a motorcycle, not the location of the shooter. This is a common misunderstanding among conspiracy and anti-conspiracy buffs alike to think that the experiment was designed to locate the shooter.

Map showing locations of microphones at 18-foot intervals in Dealey Plaza during the 1978 acoustic tests
Map showing locations of microphones at 18-foot intervals in Dealey Plaza during the 1978 acoustic tests. The figure also shows the 5 microphones which matched echo patterns.
(view larger version)

With regard to the critical first issue, that the matches could be false positives in that they might not actually be true gunfire, but something else, the definitive evidence was the order in the matching data. BBN had set an array of 36 microphones in the knowledge that the position of each microphone relative to the echo producing structures was the primary factor determining the echo pattern associated with each shot, the position of the shooter being of only secondary influence. In designing their experiment the BBN scientists had calculated that the odds of getting a match between a test shot and a sound other than a Dealey Plaza echo pattern was small, less than 5%. But the problem was that a large number of test shots were being tested (36 microphone locations and two shooter positions = 72 distinct patterns). So it is not unreasonable to expect that some matches would occur by chance, even if the odds were against any one test shot matching to a suspect sound. That is why the matches in and of themselves are not proof that the sounds were gunshots. But, herein was the significance of the experimental design. If any of the suspect sounds were not in fact Dealey plaza echo patterns, then regardless of the odds of a match, and regardless of what the sounds were, there would be an equal chance of matching at any given test microphone position. And thus if the sounds were not Dealey Plaza echo patterns and the matches spurious, then they would match in random or nonsense order. But that is not what happened. The matches lined up in the exact and only chronological and topological order required of gunshots recorded on a microphone on a motorcycle traveling in the same direction and speed as the motorcade. The BBN report states specifically that this order in the data was the evidence that conmpelled them to conclude that the suspect sounds on the police recording was the assassination gunfire. Bugliosi never explains this logic or evidence in his book.

BUGLIOSI AND THE GRASSY KNOLL SHOT

It is very likely that Bugliosi did not understand the logic or the facts in this analysis because it led to a further half-baked argument against the acoustics, although to give credit where credit is due, Bugiliosi probably got this argument from the NRC-CBA report. This is the claim that the method used by Weiss & Aschkenasy to confirm the grassy knoll shot was unproven. Bugliosi, and the NRC-CBA panel before him, argued that the method should have been applied to the Book Depository shots as a control. Bugliosi's argument, though correct in concept, was wrong in application.

Rifleman's view from behind the grassy knoll fence
Rifleman's view from behind the grassy knoll fence, at the spot where acoustics evidence indicates the fatal shot was fired.

The match to the grassy knoll shot had the same two issues attached to it as the rest: was this suspect sound truly a gunshot, and if it was a gunshot, did it truly come from the grassy knoll. The BBN experiment was designed to be accurate with regard to the microphone location but not the shooter location. That is why Barger hedged about the results of the analysis with regard to whether or not he had detected a shot from the grassy knoll. The test conducted by Weiss and Aschkenasy using sonar principles was designed to address the shooter position issue. Weiss and Aschkenasy designed their test to be accurate with regard to the shooter location by factoring in such variables as temperature, motorcycle speed, recorder speed, and using shorter distances between microphone positions. With regard to these factors, it is an established fact and requires no verification, that the speed of sound in air is a function of air temperature. Furthermore, a valid comparison of the suspect sound to the test shot, as Bob Berkowitz learned to his chagrin, requires that the playback speed match the original recording speed. These principles do not have to be proven any more than one needs to prove that gravity exists. In doing this analysis Weiss & Aschkenasy applied the same principles and used the same data that BBN had used: the same test shot and the same suspect pattern. The fact that it did match did not prove that it was a gunshot, but only, as Barger explained in testimony, if it is a gunshot, it came from the Grassy Knoll. The order in the matching data, not the matching itself, is the evidence that compels us to conclude that assassination gunfire is on the police recording. Unlike the situation with the grassy knoll, there is hard non-acoustical evidence that gunshots emanated from the sixth floor window of the book depository, and it is also known that test shots from the book depository match to the suspect sounds on the police recording. Applying the sonar analysis to these sounds cannot change either of these facts. Although the matching procedure was a critical step in the analysis, it is not proof nor was it claimed to be the proof that the assassination gunfire is on the police recording.

Another example where Bugliosi misrepresents the acoustical evidence is on page 167.

"Weiss and Aschkenasy disagreed with every other impulse sound that Barger had identified as a gunshot."

Absolute nonsense. Bugliosi now confuses the acoustical matching, in which Weiss and Aschkenasy had affirmed the work of BBN, with the subsequent audio-video synchronization. JFK received the fatal bullet at Z-frame 312/313 and because of the rearward head movement, logic and the evidence dictate that the fatal shot came from the grassy knoll. But the HSCA was insisting that the knoll shot had missed, so they did the synch both ways: with the grassy knoll shot matched to Z-313 and with the later shot matched to Z-313. Of course the alternate synchronizations also offsets the time of the other shots with regard to the Z-film. Somehow, Bugliosi managed to misconstrue the different alignments of the shot sequence as evidence that Weiss and Aschkenasy had disagreed with BBN as to which sound patterns were identified as gunshots.

BUGLIOSI AND BRUCE KOENIG'S SIGNAL ANALYSIS

Bugliosi cites the FBI Signal analysis unit as disagreeing with the HSCA acoustics analysis. The FBI, actually S.A. Bruce Koenig, claimed that gunshot echo patterns are not unique and that one of the alleged gunshots matched to a recorded gunshot from Greensboro North Carolina and this misinformation is parroted by Bugliosi. Even the NRC-CBA panel pointed out that Koenig's assertion was wrong. Koenig's performance in the Greensboro case is revealing of his expertise. At the Greensboro trial Koenig testified that he had determined that the first shot had been fired by a Klan member. As a result the Klan member was convicted of murder. However, the verdict was overturned on appeal and a second trial was held. At the second trial Koenig reversed himself, now testifying that a communist party member had fired the first shot. At this point the court brought in a real acoustical expert, Prof. Harry Hollien of the University of Florida, and author of the textbook, Criminal Acoustics. Hollien discovered that Koenig had done it wrong both times. Specifically, Koenig had failed to correctly locate the position of the microphone, an essential ingredient in an acoustical reconstruction. Koenig's performance in the Greensboro case, as well as his report on the Kennedy case, reflected that Koenig's opinions were not well grounded in acoustical principles.

THE CROSSTALK ISSUE

The challenge to the acoustical evidence that is most "hyped" is the apparent asynchrony between the suspect sounds on Ch-1 and the time of the shooting on Ch-2. Indeed, taken at its face value there does seem to be a contradiction. About one minute after Chief Curry ordered the motorcade to Parkland Hospital, county sheriff Bill Decker, in the same car, broadcast an order to surround the grassy knoll and, "Hold everything secure." In other words, Decker's order to look for a shooter on the grassy knoll is distorted by Bugliosi as evidence that there was no gunshot from the grassy knoll. A fragment of Decker's broadcast leaked from Ch-2 onto Ch-1, a phenomenon called "cross-talk," resulting in an accidental simulcast between channels. On Ch-1 the Decker broadcast occurs virtually simultaneous with the suspect gunfire. If the suspect sounds were truly the assassination gunfire then they would have been deposited on the recording at the exact time of the shooting, not more than a minute later, and so concluded the NRC-CBA panel.

But, if the cross-talk synchronization was reliable then the matching of the sounds to the test shots, and the order in the data must have been all just freaky coincidence – a series of false positives. So also would be the close match between the sequence of gunfire and the sequence of wounding and blurs in the Zapruder film (a fact not mentioned by Bugliosi). It does not seem to have occurred to the NRC panel or Bugliosi that it was the crosstalk and not the acoustical analysis that was unreliable. It takes but little effort to prove that to be the case because the Double Decker was not the only crosstalk between channels during the motorcycle segment.

There were five simulcasts in the minutes after the assassination, and in every single instance the intervals between them is different on the two channels. That simple fact means that some recording anomaly was causing the simulcasts to be offset on one or both channels. The resulting change in juxtaposition will inevitably result in apparent asynchrony of events between channels. Even the simulcasts themselves do not synchronize with one another when any one pair is used as the tie point. Moreover, the degree of offset among the simulcasts is more than enough to account for the apparent asynchrony between the suspect sounds and the assassination.

Table 1.- Speed corrected playback time intervals between simulcasts on the Dallas Police recordings (in seconds). Data from O’Dell (2003)

_______________________________________________________________________

  INTERVAL CH-1 CH-2 Δ

_______________________________________________________________________

  CHECK (1) to HOLD 10 99 89
  HOLD (2) to YOU 174 143 31
  YOU (3) TO ALL (4) 15 12 3
  YOU TO ATTENTION (5) 114 90 24

_______________________________________________________________________

1 = "I’ll check it", 2 = "Hold everything secure", 3 = "You want me to hold traffic on Stemmons", 4 = "I’ll check all these motorcycle radios", 5 = "Attention all emergency vehicles."

The only way to salvage the NRC-CBA hypothesis would be to determine where the anomaly has imposed the offsets and the amount of time involved. Although several candidate phenomena have been nominated, there is no direct evidence to select from among them, nor is there any valid reason to suppose that all potential phenomena have been thought up. Nonetheless, the NRC-CBA proposed, and then accepted as fact, that the cause was stoppage of the recorder on Ch-2. The two problems with that hypothesis are that 1) there is no evidence that the recorder stopped, 2) there is evidence that there is no time missing on Ch-2, at least not during the critical time in question: the first few minutes after the shooting.

Bugliosi declares on page 210 that these arguments, raised by yours truly, a mere entomologist, are "demonstrably false." He claims that the NRC-CBA found that the Ch-2 recorder "paused five times." Adding, "One wonders how all of this escaped Thomas's attention, since he obviously must have had a copy of the CBA report." I do indeed have a copy of the CBA report and it definitively does not say that the recorder paused five, or any number of times. The CBA found five places where there is no recorded speech for intervals of around four seconds, which in theory might have triggered the sound actuation feature on the recorder to shut off. There is no evidence that it actually did so. Contrary to Bugliosi's claim that the NRC-CBA playback is the best record of the Ch-2 recording, at least some of the speechless sections in the recording are artifacts from flaws in their playback process. There are at least four broadcasts on the Bowles playback that are missing from the NRC-CBA playback. Some are due to skips, in spite of the false claim that their playback was free of skips, while others are missing because of a volume problem with the FBI's playback machine causing silent periods instead of speech. More importantly, even if the recorder did stop, there was no direct way to determine for how long it had stopped. But there was an indirect way.

The radio dispatchers as part of their protocol would append a time notation to their broadcasts, for which purpose they each had a digital clock on their console. On Ch-2 there is a one-to-one agreement (mathematically 0.99) between the dispatchers time notations and the actual playback time, as determined by the NRC-CBA panel for the six minutes after the shooting. This means that not only is the NRC-CBA hypothesis not proven, and that there is no evidence to support it, but that it is mathematically unlikely. Bugliosi, on page 210, asserts without any counter calculation that this math is "demonstrably wrong." In reply to this problem Linsker et al argued that the apparently perfect agreement between the time notations and the playback time does not quite attain the 95% level of certainty that there isn't thirty seconds missing, as their hypothesis requires. The fact that their hypothesis may not be provably false hardly meets the burden of proof demanded by their assertion that the crosstalk proves that the acoustical identification of gunfire is wrong.

The further problem for the CBA hypotheis and Bugliosi's argument, is that the simulcasts do support synchrony between the assassination and the suspect sounds. Neither the CBA panel or Bugliosi ever informs the reader of the juxtaposition of the assassination on the Ch-2 recording, a point which would seem to be critical to an argument over whether the sounds are synchronous with the asssassination. The CBA used Curry's broadcast "Go to the Hospital" as their marker. Obviously though, the shooting must have occurred some time before that broadcast. In fact, 20 seconds earlier Curry announced that he was "approaching the triple underpass." It is known from the motorcade films that the assassination occurred when the president's limousine was in the mid-section of elm street and that at that moment the lead car was approaching the triple underpass.

The instance of crosstalk closest to that event was a broadcast two seconds earlier by Deputy Chief Fisher, saying, "Naw that's all right, I'll check it." The last words, "I'll check it," crossed over to Ch-1, just two seconds before the first acoustically identified gunshot. Hence this simulcast establishes close synchroneity between the shooting and the deposition of the suspect sounds on Ch-1.

In denial of this evidence Bugliosi cites Steve Barber's misinterpretation of the spectrographic comparison (voiceprint analysis) of the Fisher simulcast by Linsker et al. (2006), claiming that the pair failed the test. The exact opposite is true. Bugliosi and Barber are easily bamboozled, and Linsker, Garwin and colleagues willing bamboozlers.

First of all, a standard voiceprint analysis is inadequate in this situation because of the high noise to signal ratio. The dominant sound on the Ch-1 recording is noise, mainly the motorcycle motor. The proper method to compare the signals is in the manner done by the NRC-CBA panel in 1982 and again by Linsker et al. in their report. The computer driven analysis not only compared the frequency pattern of the broadcasts to one another, it also compared several seconds of sound on either side of the Ch-1 broadcast to the Ch-2 broadcast as a check. For example, in the case of the Decker broadcast, the noise on either side of the utterance on Ch-1, when compared to the broadcast from Ch-2, gives a correlation coefficient of 0.1 (of a possible 1.0), meaning about a ten percent similarity. This represents random or chance similarity. But when the utterances are lined up, the similarity spikes, only to about 0.3, but the fact that it spikes or peaks is the indication that the signals are at least significantly similar with regard to their frequency spectrum. When the Fisher broadcasts were compared the same thing happened, the computer comparison generated a peak indicating that they are significantly similar – the same as the other crosstalks.

Listen to Audio Recording of Deputy Chief Fisher

Linsker et al dismissed this peak as "accidental," in other words, arguing that it was a false positive. It couldn't be a true simulcast, they argued, because it is out of synch with the other simulcasts. In science we call this circular reasoning. It amounts to saying that the test results cannot be correct because it contradicts our hypothesis. All of the simulcasts were out of synch with one another. The second reason cited by Linsker et al as to why the spike was "accidental" was because the peak occurs at a speed warp different from expected. But the other simulcast comparisons also peak at speed warps different from the expected. Besides, a comparison at the wrong speed cannot cause or explain the spike. Whereas Linsker et al. were in denial, Bugliosi was in the dark. The words on Ch-1 are clearly "I'll check it," not "all right Chaney" or "all right Jackson." That is why whenever I lecture on this subject I play the recordings for the audience so they can judge for themselves. I have even been accused of "enhancing" the recording. When I debated Richard Garwin on this issue he argued that there are multiple places where the words "I'll check it" occur. That argument was also false.

CONCLUSION: SCIENCE VS. STATUS

One last word. In what I hope is not typical lawyerly fashion, Bugliosi makes light of my profession as an entomologist and implies that I am therefore unqualified to have a valid opinion on the acoustical issues. Entomologists (including Aristotle, Charles Darwin, Alfred Kinsey and Walter Reed to name a few) have a long history of contributing solutions to societal problems, a historical fact of which Bugliosi is unaware. Bugliosi's style of argument serves to point out the difference between scientists and lawyers. In science a theory stands or falls on the merits of the facts and logic behind it, and not on the status of the individual who made the theory. But Bugliosi is not even a very good lawyer because if this standard were to be applied in this case then the only persons entitled to an opinion on the acoustics issue would be Barger, Weiss and Aschkenasy. None of the authors of the CBA study, all fine scientists in their own fields, were acoustical experts. Moreover, since the Daubert decision, "bona fide expert" opinion is no longer sufficient to establish the results of forensic studies in a court of law. Such evidence is now challengeable on its scientific merits, as it should be. Bugliosi's arguments concerning the forensics in the JFK case would not withstand such a challenge.


For more information, please see Thomas' essay Overview and History of the Acoustical Evidence in the Kennedy Assassination Case and this interview with Thomas and accompanying resource page.

© Mary Ferrell Foundation. All Rights Reserved. |Site Map |MFF Policies |Contact Us