Scientists Cast Doubt on Kennedy Bullet Analysis

[Michael Hogan]

By John Solomon

Washington Post Staff Writer

Thursday, May 17, 2007

In a collision of 21st-century science and decades-old conspiracy theories, a research team that includes a former top FBI scientist is challenging the bullet analysis used by the government to conclude that Lee Harvey Oswald alone shot the two bullets that struck and killed President John F. Kennedy in 1963.

The "evidence used to rule out a second assassin is fundamentally flawed," concludes a new article in the Annals of Applied Statistics written by former FBI lab metallurgist William A. Tobin and Texas A&M University researchers Cliff Spiegelman and William D. James.

Full article: http://www.washingtonpost.com/wp-dyn/conte...7051601967.html

[Pat Speer]

By John Solomon

Washington Post Staff Writer

Thursday, May 17, 2007

In a collision of 21st-century science and decades-old conspiracy theories, a research team that includes a former top FBI scientist is challenging the bullet analysis used by the government to conclude that Lee Harvey Oswald alone shot the two bullets that struck and killed President John F. Kennedy in 1963.

The "evidence used to rule out a second assassin is fundamentally flawed," concludes a new article in the Annals of Applied Statistics written by former FBI lab metallurgist William A. Tobin and Texas A&M University researchers Cliff Spiegelman and William D. James.

Full article: http://www.washingtonpost.com/wp-dyn/conte...7051601967.html

Tobin's a bit of a hero. He's a retired FBI man who knew NAA testing for bullet lead was junk science, and has been successfully campaigning to improve it or kill it for years.

[J. Raymond Carroll]

By John Solomon

Washington Post Staff Writer

The "evidence used to rule out a second assassin is fundamentally flawed," concludes a new article in the Annals of Applied Statistics written by former FBI lab metallurgist William A. Tobin and Texas A&M University researchers Cliff Spiegelman and William D. James.[/b][/indent]

Full article: http://www.washingtonpost.com/wp-dyn/conte...7051601967.html

Tobin's a bit of a hero. He's a retired FBI man who knew NAA testing for bullet lead was junk science, and has been successfully campaigning to improve it or kill it for years.

A JFK researcher named Emory Brown contacted Western Cartridge Company back in the mid-1970's and made the same discovery about the metallurgy of bullet manufacturing that Tobin made only recently. Emory Brown's article was published in The Continuing Inquiry, which Rex Bradford is planning to make available online at the Mary Ferrell website.

Even before Tobin went public with his concerns, and even before the National Academy of Science issued its report, the courts had recognized that the "methodology" used by Vincent Guinn (and adopted by the FBI) for calculating probabilities has no scientific/mathematical basis: [Correction: Tobin had gone public before the Mikos decision came down

Forensic Science International Volume 127, Issue 3, 17 July 2002,]

Quote on:

Summaries Of Opinions

In Federal Death Penalty Cases

United States v. Mikos

No. 02 cr 137 (N.D. Ill. Dec. 5, 2003)

Court granted in part defense motion to exclude government expert

testimony regarding comparative bullet lead analysis. Such analysis

involves the comparison of two bullets to determine if they are chemically

indistinguishable and then to opine whether they came from the same batch

of lead.

The trial court determines that such expert testimony can be

separated into distinct areas. The first is an actual analysis of lead

bullet comparisons which is a question of chemistry. This was not

challenged by the defense.

The second area, however, "is essentially a

statistical conclusion" which the government concedes does not lend itself

to a scientifically valid estimate. The court rejects the government's

argument that 30 years of real-world experience and the admissibility of

this evidence in other courts overcomes the inability to accurately

measure the probability that two bullets with similar chemical

characteristics came from the same batch.

The court further finds that there is no evidence that the samples which

make up the government's database were gathered in a scientifically

valid manner so as to represent the bullet population as a whole. Therefore,

the court holds that the expert cannot testify that the bullets used to kill the

victim and the bullets found in the defendant's car came from the same batch of lead.

QUOTE OFF

This latest effort by Tobin, et al appears to be misguided, if the Washington Post is accurately describing Tobin's latest argument.

G.Robert Blakey promoted Guinn's theory to the HSCA in an attempt to prove that the Connally wrist fragment came from CE399 and that a fragment from JFK's head wound came from a large bullet piece (also known to be Carcano ammo) found in the presidential limo. Blakey was trying to prove that the Magic Bullet Theory was true and, above all, that CE399 and the limo fragments WERE NOT PLANTED, as Josiah Thompson and other critics had suggested.

The Federal Court decision in Mikos (plus the Behn decision in New Jersey) struck a deadly blow to Guinn's entire HSCA argument. As reported by Bill Kelly on this forum in November, 2005, Blakey has now admitted that Guinn's entire argument is "junk science". I am reasonably certain that if you spoke to Blakey today he would concede that there is no scientific evidence to disprove Thompson's argument that CE399 and the limo fragments were planted.

If the Washington Post article is accurate, then Tobin is way behind Blakey in appreciating the significance of comparative bullet lead analysis (CBLA) in the JFK assassination. Tobin may be an expert in metallurgy, but as far as I know he is no more qualified in statistics/probability than Guinn was. There is no need for further tests on the trace elements in the JFK bullet fragments. It is not a question of whether there was "a second gunman."

The real question now is whether the trace elements discovered by Guinn constitute proof positive that CE399 and the limo fragments were planted.

Anyone who studies the chart of trace elements that Guinn presented to the HSCA ( C.F. Volume 1 HSCA) will plainly see that the Connally fragment does not "match" CE 399, and the JFK fragment does not "match" either of the Limo fragments. The question to be determined now is whether these differences constitute scientific proof that CE399 and the limo fragments were planted.

Edited May 20, 2007 by J. Raymond Carroll

[Pat Speer]

Ray, you are correct in that Stu and Tom Pinkston and William Tobin have been focusing on the question of whether or not bullet lead analysis is 100% accurate or even 90% accurate and that there are others who focus on whether or not the fragments were planted. I doubt this because in my analysis of the NAA, Guinn not only came to an improper conclusion, he came to the WRONG conclusion! It is almost certain that the wrist fragments DID NOT come from CE399.

From patspeer.com

"But what of the Neutron Activation Analysis, which nuclear physicist Vincent Guinn testified demonstrated that the bullet fragment removed from Connally’s wrist most probably came from the magic bullet found on a stretcher in Parkland Hospital? Doesn’t this prove that the bullet hitting Connally in his wrist was the same bullet that fell out of his leg, and support the single-bullet theory.

Well, without even going into the substantial circumstantial evidence indicating that the bullet supposedly found on Connally’s stretcher after falling from his leg was, in fact, found on someone else’s stretcher (as per the hospital employees who discovered it—Darrell Tomlinson and Nathan Pool), was never seen until at least an hour after Connally had been rushed into the hospital (as per the nurses and orderlies who removed Connally’s clothes and wheeled away his stretcher—Doris Nelson, Ruth Standridge, Jane Wester, and R. J. Jimison) and never lodged in his leg (as per Connally’s doctor, Dr. George Shires), there is reason to doubt Guinn’s results proved what so many believe. (An alternative explanation for the bullet’s presence on the stretcher appears with the next slide)

If one looks at Guinn’s results, one realizes there is a surprising lack of uniformity in the make-up of Mannlicher-Carcano bullets, both from bullet to bullet and box to box. This is because the type of ammunition used in the gun believed to be Oswald’s was made from the melted-down lead of other bullets. When one looks even closer at Guinn’s analysis, one finds that his interpretation of his test results leaves even more to be desired. Since Guinn believed that similar counts in parts per million of certain elements could leave an identifiable fingerprint of exact bullets, and that Antimony, Silver, and Copper were the most reliable of these elements, let’s make a comparison between three sets of bullets on these elements, and Guinn’s subsequent conclusions.

Numbers reflect the counts of the two samples in parts per million.

A vs. B. 647-602 Antimony, 8.6-7.9 Silver, and 44-40 Copper.

C vs. D, 833-797 Antimony, 9.8-7.9 Silver, and 994-58 Copper.

E vs. F, 732-730 Antimony, 15.9-15.3 Silver, and 23-21 Copper.

So which two samples were described by Guinn as being from the same bullet?

Well, that's actually a trick question, as A vs. B actually represents FOUR samples, a fragment found in Kennedy's brain, two fragments found on the floor of the limousine, and the nose of the bullet found on the front seat. And yet notice how uniform they seem to be. One might actually conclude they are probably from the same bullet. And Guinn did. Well, since they were so uniform and since Guinn also concluded the wrist fragments came from the magic bullet, then E vs. F must be the comparison between the magic bullet and the wrist fragment, right?

WRONG. E vs. F is a comparison between 6001B and 6003A, test bullets taken from separate batches of ammunition from separate years. Subsequent tests showed them to be quite dissimilar.

Which leaves C vs. D as the wrist/magic comparison. Since the Silver and Copper ranges are substantial, it's safe to say Guinn's conclusion came purely from the similarity on Antimony. He ignored everything else and focused on those two numbers...833-797. And yet, when one looks at the test results, one finds that 6002 A2 was at 869, and 6001 B4 was at 791, within 36 ppm of the magic bullet and the wrist fragment, respectively, and this out of only 40 tests beyond the magic bulet and wrist fragment. This translates to there being a 5% chance for the wrist and magic fragments to fall within 36 ppm randomly. Of the 14 different bullets tested from assorted boxes of Western Cartridge ammunition, in fact, 3, 6000a, 6001d, and 6001A, were within 15 ppm on antimony, even though they were from different years and different batches. This reduces the 833-797 numbers to nothing near the relevance Guinn and such disciples as Kenneth Rahn attach to it. When one takes into account the other six elements tested, in fact, the logical deduction is amazingly the opposite of Guinn's ...that it's highly probable the magic bullet and the wrist fragment ARE NOT related.

A comparison of ranges of the 4 fragments found in the limousine vs. the magic bullet/wrist fragment on the 7 elements tested by Guinn:

Antimony: 4 fragments 647-602, magic/wrist 833-797

Silver: 4 fragments 8.6-7.9, magic/wrist 9.8-7.9

Copper: 4 fragments 44-40, magic/wrist 994-58

Aluminum: 4 fragments 5.5-1.1, magic/wrist 8.1-0

Manganese: 4 fragments 0.1-0.01, magic/wrist 0.09-0.07

Sodium: 4 fragments 134-9, magic/wrist 120-5

Chlorine: 4 fragments 59-22, magic/wrist 257-19

Since the range of 2 related samples should be smaller than the range of 4 related samples (7 out of 8 times), and since the range difference should usually be significant, it's clear that Manganese is the only element that suggests the magic bullet and wrist fragment are related, and that Antimony and Sodium are also consistent with that analysis. It's equally obvious that the other 4 elements tested are strongly suggestive there was NO relation at all between the two, as the range of the 2 samples is many times that of the 4. The proper conclusion then should be that the magic bullet and the wrist fragment are most probably not related. This conclusion is supported by the additional fact that CE 399, while missing some lead, is not believed to have lost any size-able amount of copper. As both Connally’s coat by his exit wound and the wrist fragments themselves were found to contain inordinate amounts of copper, one should conclude he was struck by a separate bullet whose jacket had been badly damaged. In short, anyone whose argument for the single bullet theory relies on Guinn's analysis has clearly never studied Guinn's results with an open mind. His conclusion was wrong; whether he sincerely believed his testimony or was asked to lie is open to conjecture.

While I had not planned on engaging in such conjecture, recent developments in bullet lead analysis have alerted me to much that is suspicious with Guinn’s analysis, beyond his incorrect conclusions. On September 1, 2005, the FBI announced they would discontinue the use of bullet lead matching. Their decision was spurred on by a February 2004 report by the National Academy of Sciences questioning the value of bullet lead analysis, particularly in light that it had never been tested by scientists outside those whose careers depended on its presumed worth, including Vincent Guinn. Surprisingly, this study was performed by the Academy on behalf of the FBI itself, after a former FBI metallurgist named William Tobin began writing articles critical of the probative value of bullet lead analysis. Among the reports written by Mr. Tobin and members of the Academy, I found at least three good reasons to be even more suspicious of Guinn.

1. Although bullet lead analysis has been used for over 30 years, the FBI would not allow its employees to testify beyond that a bullet (usually found within a body) was likely to have come from the same box of bullets as was found somewhere else (usually in the home of a suspect). Guinn’s testimony that it was “highly probable” the wrist fragments and the magic bullet were parts of the same bullet is perhaps the only time in history someone has testified to such degree. Since the National Academy has now found that “The available data do not support any statement that a crime bullet came from, or is likely to have come from, a particular box of ammunition,” and that the possible existence of coincidentally indistinguishable bullets “should be acknowledged in the laboratory report and by the expert witness” it would seem apparent that Guinn’s expert opinion went above and beyond what was warranted.

2. While Guinn said his opinion was based on the results of three elements, Antimony, Silver, and Copper, the FBI at that time was using Antimony, Copper, and Arsenic. Even when Guinn expanded his test to seven elements, Arsenic was not included. This forces one to consider the possibility that Guinn tested Arsenic, found it did not match, and excluded it from his results. Since Silver, which the FBI started using as one of its seven elements in 1990, is reported to have little value, as most bullets are within a small range in parts per million, and are considered to match, its propping up by Guinn as the second most valuable element is intriguing. Perhaps, faced with the fact that Copper failed to match, and being aware of how bad it would look if two out of the three elements he tested failed to match, Guinn simply picked an element that would help him make his case. I asked a prestigious metallurgist who’d helped me in the past if he knew of any good reason Guinn would use Silver instead of Arsenic, and have yet to receive an answer. The lack of value of Silver as a determinant that two fragments have an identical source is made obvious by Guinn’s own results, where more than half of the test bullets matched the wrist fragment in Silver, with many of them closer in parts per million than the “magic” bullet determined by Guinn to be identical.

3. It seems Guinn himself was skeptical of any conclusions based on only three elements. In 1970, a report for the Atomic Energy Commission prepared by Guinn and three other scientists concluded “two bullets with the same pattern of only three identification points are not usually definitively identified as having a common source, Matching concentrations of all three elements does not indicate that two bullets came from the same lot.” Since the FBI began using seven elements 20 years later, and since it was necessary for a bullet to match on all three elements tested up until that time, and all seven elements afterwards, before the FBI would even find that a bullet came from the same box as another bullet, it seems clear that, due to the problems with Copper, in no time in its history would the FBI have testified that the wrist fragments and the magic bullet matched. In fact, when given the opportunity to do so, in 1964, the FBI ruled their tests inconclusive and kept them from the public. The question then is not only why did Guinn testify in the manner he testified, in contradiction to his previous reports and the accepted standards of the FBI, but whether the FBI was deliberately removed from the process.

While I still can’t answer that, I have uncovered a possible innocent explanation for Guinn’s mistake, one that moves him from the category of deliberate xxxx to mere screw-up. In early tests of bullet lead, Guinn and others discovered there was an apparent conformity between bullets in the same box, and sought to find practical applications for their discovery. Over time, the courts came to accept the value of bullet lead analysis and the FBI began testifying that one bullet most probably came from the same box as another. This allowed prosecutors to convict suspects even when no gun was found. The problem, as outlined by William Tobin and the National Academy of Sciences, was that little research was done on how bullets were actually made and distributed, and that, when one studied these things, one could only conclude that virtually indistinguishable bullets were likely to end up in boxes of ammunition on opposite sides of the country. In his research on Mannlicher-Carcano ammunition, however, Guinn found that the bullets within the same box had no apparent conformity. This led him to believe that the slight conformity between the wrist fragment and the magic bullet had significance, as other bullets from its box would be unlikely to match on Antimony. The problem was that there was NO REASON to assume the wrist fragment bullet came from the same box as the magic bullet. Quite literally, then, Guinn was thinking inside the box when he should have been thinking outside the box!

Or maybe he was just lying. I hesitate to pass judgment on the man. All I know for sure is that his conclusions about the magic bullet before the HSCA were wrong. In July, 2006, Dr.s Erik Randich and Patrick Grant published an article in the Journal of Forensic Sciences.

Their conclusion reads: “We therefore assert that, from our perspectives of standard metallurgical practice and statistical assessment of the fundamental NAA measurements a conclusion of material evidence for only two bullets in the questioned JFK assassination specimens has no forensic basis. Although collateral information from the overall investigation might very well narrow the choices, as stand-alone primary evidence, the recovered bullet fragments could be reflective of anywhere between two and five different rounds fired in Dealey Plaza on that day. Only the near-complete mass of CE-399, the stretcher bullet, precludes the conclusion of one to five rounds. Moreover, the fragments need not necessarily have originated from MC ammunition. Indeed, the antimony compositions of the evidentiary specimens are consistent with any number of jacketed ammunitions containing hardened lead."

Forunately, Guinn’s tests were not the last to be performed on the fragments. In 1998, further tests were performed, this time to determine the nature of the organic material found on the fragment discovered on the front seat of the car. To most everyone the results of these tests, conducted on behalf of the National Archives by the FBI and the Armed Forces Institute of Pathology, Dr. Finck’s old stomping grounds, were disappointing. The Justice Department employee who inspired the tests evidently was hoping that the material would turn out to be fibers, and that these fibers would be traced back to Connally’s clothing, which would demand a re-examination of the single-bullet theory and re-open the case. (After all, how could this fragment, which was previously determined to be from the same bullet as the fragments in Kennedy’s skull, also have been the bullet to hit Connally in the back?) Others were disappointed when the tests revealed the material was human tissue, but couldn’t determine whose tissue. When I recently read about these over-looked test results, however, I found the confirmation for my theory I’d doubted would ever come. For the test results showed that 3 of the 4 fragments of organic material found on Exhibit CE 567 were made up of SKIN. The significance of this skin will ultimately be explained."

[Douglas Caddy]

By John Solomon

Washington Post Staff Writer

Thursday, May 17, 2007

In a collision of 21st-century science and decades-old conspiracy theories, a research team that includes a former top FBI scientist is challenging the bullet analysis used by the government to conclude that Lee Harvey Oswald alone shot the two bullets that struck and killed President John F. Kennedy in 1963.

The "evidence used to rule out a second assassin is fundamentally flawed," concludes a new article in the Annals of Applied Statistics written by former FBI lab metallurgist William A. Tobin and Texas A&M University researchers Cliff Spiegelman and William D. James.

Full article: http://www.washingtonpost.com/wp-dyn/conte...7051601967.html

To read the full research paper as published in the Annals of Applied Statistics, click on the link below and go to the second to the last article listed:

http://www.imstat.org/aoas/next_issue.html

By the way, the NBC Nightly News with Brian Williams tonight devoted a segment to the research paper and interviewed one of its authors, who challenged the one-bullet theory and advocated that further research be conducted on the existing JFK bullet fragments that are in the possession of the National Archives.

[John Geraghty]

There has been a considerable amount of favourable publicity for what we already know of late. The Hunt affair may be gathering steam, this research paper, David Talbots book and a few favourable articles in the Times and in other newspapers.

It is encouraging that we are amking somewhat of an impact.

John

[J. Raymond Carroll]

"There has never been a comprehensive or forensically meaningful statistically

based sampling study of bullet composition by manufacturer, location, or epoch."

Spiegelman, Tobin et al.

This fact was noted by the Court when ruling that the expert's opinion in United States v. Mikos has no scientific basis and is therefore inadmissable. The court would have had to rule the same way, and for the same reasons, if the the witness had been Vincent Guinn in the JFK case.

Even if we had such a background study, it could not tell us how many bullets were involved in the assassination, since we know nothing about the composition of bullets that might have missed (e.g. James Tague).

Until such a background study is done, it will remain impossible even to guess at what kinds of bullets, or even what caliber of bullet, gave birth to either the JFK or Connally wound fragments. As Randich and Grant noted in the article quoted by Pat Speer, "the fragments need not necessarily have originated from MC ammunition. Indeed, the antimony compositions of the evidentiary specimens are consistent with any number of jacketed ammunitions." I would add that it is by no means clear that the ammunition had to be "jacketed."

Even without such a background study, however, it may be possible to learn something meaningful from analyzing all the bullet lead in evidence.

We already know quite a bit (and we may be able to learn more) about the composition of the wound fragments and about CE399 and the limo fragments (and MC bullets in general). While we will not be able to say that they have a common origin, we may be able to eliminate CE399 and the limo fragments as possible parents of the wound fragments. We may be able to determine, even without a background study, that CE399 and the limo fragments were planted.

I submit that inquiry along such lines is the logical way forward in this area of the case.