INDC was the blog that found Bouffard.
UPDATE on Dr. Bouffard's Forensic Opinion
I spoke with Dr. Bouffard about his equivocation in the NY Times, and first let me say that I get the feeling that the poor guy wants to go into hiding after all of the attention that he's getting from the media. His cooperation, professionalism and patience in this matter is most appreciated. And make no mistake, the mainstream media is pursuing this story. CNN, NYT, WaPo, regional papers, they're all calling him.
There is also an animated discussion taking place within the community of forensic document anlaysts, with the large majority that have spoken to Dr. Bouffard leaning towards or essentially convinced that the documents are forgeries. Bill Flynn, a well-respected document examiner from Arizona with a specialty in computer word processing, is certain that the documents could have only been created with a computer. On the other hand, Dr. Bouffard received an e-mail from a document examiner named Lynn Huber that attempted to bolster CBS's case.
Is Ms. Huber the secret analyst that CBS won't name? I'm not sure, but the documents that she provided were:
* Samples of Times Roman-like fonts from an IBM Composer machine, a rather bulky typsetting machine that was capable of producing a Times Roman-like font that could resemble the CBS documents, but still has notable differences.
* An Air Force memo that indicates that the military was testing the IBM Composer and considering a purchase in 1969. No further documents buttress the concept that the machine was ever purchased (though it's certainly plausible, perhaps likely), or how it would be possible for a large typesetting machine to wind up typing a personal "CYA" memo in the office of a Texas Air National Guard Lt. Colonel, however.
This new information is the cause of Dr. Bouffard's equivocation to the NYT. Based purely on font analysis and given the poor quality of the documents, it is conceivably possible that a somewhat similar document could have been created by an IBM Composer in 1972, though other elements yet to be determined still could render it impossible (see kerning below). The superscripted "th" casts additional doubt on the veracity of the document, but does not slam dunk the case for forgery. It would have been possible (but highly unlikely) to order specialty balls and keys that would create just about anything, including a superscripted "th," but it's important to consider the fact that ordering such a key would also demand additional specialty balls and keys for the superscripted "nd," etc., as well as separate balls for the italic versions of all the characters. It would have been a custom, mean, souped-up typing machine.
What does this all mean? It means that the possibility that this memo was not forged based on typographic evidence alone is contingent upon the possibility that an Texas Air National Guard Lt. Colonel typed a personal "CYA" memo on an ultra-modern, highly-customized typesetting machine that was typically used for professional or high end applications that demanded camera-ready documents for use in printing. And even so, Dr. Bouffard still noticed "a dozen or so design differences" between the CBS document and the possible type of such a rare machine, differences that seem apparent but cannot be confirmed to a 100% threshold because the of the poor quality of the photocopied document.
But even so ...
... what about kerning? Kerning is the process by which letters are placed closer together in the space on the page in a way that automatically adapts to the order in which letters are placed next to each other. For example, in the word "my" in certain fonts, you may notice how the bottom tail of the "y" curls under the letter "m." As Powerline has mentioned, the memo has kerning, something widely believed to only be possible with modern word processing programs that algorithmically adapt to the sequence of letters.
So could an IBM Composer kern? Dr. Bouffard can't say with 100% accuracy, but he does not think it's possible that early machines had the computer memory necessary to complete such a function. The only possible way to find out would be to obtain an IBM Composer circa 1972.
Long story short?
I'm just presenting the analysis one expert forensic document examiner that specializes in typefaces, so, ignoring the statements by Killian's family that disavow the documents, ignoring the inconsistencies in tone, format and the active military status of individuals mentioned in the memo, soley based on forensic typographical analysis, it is highly unlikely that the documents are real, and if someone can verify that a 1972 IBM Composer cannot kern (or auto center, for that matter), then it will be completely verified that the documents are frauds.
Once again, I am not commenting on the myriad of other angles that question this story, merely providing an in depth summary of Dr. Bouffard's findings. To quote:
"Because it takes such a stretch to come up with all of the remote possibilities involved in creating the (CBS) document, it is much more likely that it is a computer generated document.
...
I can't imagine the Composer would have enough memory (to kern)." |
