The Evidence Against Rather

eclectic said: "It is interesting to check the samples of the alleged typewriter for overhang glyphs."

The reason for my interest in this detail is that FR can easily supply the technique needed to display an "overhang glyph" in enlarged format to Fox News along with an assertion regarding the possibility of such a thing being done on a typewriter.

Such an exercise would allow Fred Barnes to shut down Juan William's claim that some typewriter somewhere might have been able to create the document. It is important to have one, simple, graphical example which demonstrates the forgery if that is possible.

9. Margins. These look like a computer's unjustified default, not the way a person typing would have done it. Typewriters had fixed margins that “rang” and froze the carriage when typist either hit “mar rel” or manually returned carriage.

===

We have confirmed with several Selectric owners that justifying with a proportional font typewriter was not usually done as it was terribly difficult

Hmmm, the new democrat 527 "Texans for Truth" says that there aren't many reasons advanced to convince them the papers are fraudulent. Think I should invite them over here? LOL

17. Paper size problem, Air Force and Guard did not use 8 1/2 x 11 inch paper until the 1980s.

====

This is not a problem since the forger could have used any page size and cropped or expanded the image. The resulting page size which the CBS memos show in Acrobat is NOT 8.5 x 11

29. Not signed or initialed by author, typist, or clerk.

===

For some strange reason, we have found (after looking at non-CBS correspondence from them) that Killian's colleagues generally did not follow this practice.

That's an important point. Only printing directly to paper, scanning and comparing is likely valid. Screen drafts, screen print previews, print-to-PDF, etc., are all apt to have sublte but damaging differences, even vs. the fairly coarse CBS docs.

Exactly. The glyph outlines in high quality TrueType fonts like TNR (arguably the highest quality computer font in existence) are highly hinted -- distorted, really -- when displayed on a low resolution device like the screen. Once you get to a high resolution device like a 600dpi laser printer, the hinting is less extreme. For example, here's what a 12pt TNR 'S' looks like in my ancient font testing app:

This is a 12pt 'S' as it appears on a monitor. Note how the white outline is highly distorted so that certain pixels are forced on and others are forced off. Compare that to the same 'S' at three times the resolution (36pt on the screen):

Now the outline is silky smooth. Hardly any distortion at all. Also, note that the spacing between glyphs is affected by the point size of the font, too.

36. Why were these exact same documents available for sale on the Internet y Marty Heldt, of leftist web site Tom Paine, as early as January 2004? Is this where CBS obtained their copies?

===

That's a huge one! Do you or anyone have a link for this?

http://www.ibmcomposer.org/ has some samples. Yes they have overhang. High quality typography did use overhang for centuries after all, but the probability that this memo was typeset with a composer machine, rather than ordinary typewriter is pretty slim

best point

You (and Southack) are correct. See

http://www.truetype.demon.co.uk/articles/times.htm

Excellent point.

eclectic said: "High quality typography did use overhang for centuries after all, but the probability that this memo was typeset with a composer machine, rather than ordinary typewriter is pretty slim"

Let me try to outline this:

It is true that the word "interference" appears in one of the memos.

Is it true that magnification of the "fe" in that word reveals "glyph overhang"? (I believe so, but am not an expert in interpreting digitized images of photocopies of "typed" documents.)

If so, is it true that there was no tool available to the Texas Air National Guard in 1973 which could have created this specific detail? ( I believe so, but I am not an expert in how 1973 vintage typewriters accomplished proportional fonts and similar features.)

If so, then this specific example can be used to establish that particular document as a forgery.

The reason for boiling down the details this way is to eliminate any possible refutation by the liberals. It is only necessary for us to provide a single fact which is provably inconsistent with the CBS hypothesis that the document is genuine.

The goal, I think, is to provide the unbiased media with a simple convincing example in order to prevent the "he said/he said" confusion that the liberals love so much.

All applications on today's computers use kerning. What people call kerning in a word processor is simply a programmatic feature which allows the user to override the automatic settings.

It seems to me that overhang would be the kiss of death. I've verified that Word2k *always* has "fi" overhang, regardless of kern setting. In the typography world printers would use what's called a "ligature" in that case, which is a single character representing the "fi" where the "i" had no dot and the overhang of the "f" extended just a little farther so it looked like it completely overwrote the i's dot. Anyhow, there is no way a standard typewriter could produce overhang, much less a ligature character. I eagerly await news from our resident experts on what a composer machine's capabilities are. Nonetheless I think it's already painfully obvious that these documents are blatant forgeries. It only remains to show something concrete that can definitively prove it so the backlash against Dem hacks can begin in earnest.

If so, is it true that there was no tool available to the Texas Air National Guard in 1973 which could have created this specific detail? ( I believe so, but I am not an expert in how 1973 vintage typewriters accomplished proportional fonts and similar features.)

This cannot be ruled out with absolute certainty. It is just highly unlikely. The font (a variant of Times Roman) and the machine to type it (IBM composer) did exist at the time the memo was allegedly produced. But exact match to MS Word is suspicious. The poor quality of copy is suspicious. The only way to be sure is to examine the originals. Let them produce the originals!

The Swiss Cheese Defense: A Live Fisking

September 10, 2004, 03:22:50 EDT

Dan Rather's defense of himself tonight, while probably impressive to shallow observers was far from convincing. Here's a list of things he ignored, did not properly address, or concealed from viewers. Feel free to send us your suggestions to this live fisking. For the transcript, click here.

Sourcing problems

The 72-year-old anchor conveniently did not mention the fact that James Moore, one of his key validative sources, is a left-wing activist and author who has written two anti-Bush books, Bush's Brain, and Bush's War for Reelection. Rather referred to him as "author Jim Moore has written two books on the subject."
Not coincidentally, Rather also did not mention that one of its main validators, retired Maj. General Bobby Hodges is accusing 60 Minutes staff of lying to him in order to get him to say the supposed Killian memos were authentic. ABC News has the story:
      "Hodges, Killian's supervisor at the Guard, tells ABC News that he feels CBS misled him about the documents they uncovered. According to Hodges, CBS told him the documents were 'handwritten' and after CBS read him excerpts he said, 'well if he wrote them that's what he felt.'
      "Hodges also said he did not see the documents in the 70's and he cannot authenticate the documents or the contents. His personal belief is that the documents have been 'computer-generated' and are a 'fraud.'"
      The Washington Post reported earlier today that CBS considered Hodges its "trump card":
      "A senior CBS official, who asked not to be named because CBS managers did not want to go beyond their official statement, named one of the network's sources as retired Maj. Gen. Bobby W. Hodges, the immediate superior of the documents' alleged author, Lt. Col. Jerry B. Killian. He said a CBS reporter read the documents to Hodges over the phone and Hodges replied that "these are the things that Killian had expressed to me at the time."
      "These documents represent what Killian not only was putting in memoranda, but was telling other people," the CBS News official said. "Journalistically, we've gone several extra miles."
      The official said the network regarded Hodges's comments as "the trump card" on the question of authenticity, as he is a Republican who acknowledged that he did not want to hurt Bush. Hodges, who declined to grant an on-camera interview to CBS, did not respond to messages left on his home answering machine in Texas.
      Looks like jokers are no longer wild.
He deliberately ignored statements from Col. Killian's wife and son who said that he hated using typewriters, hardly ever kept notes, and very much liked George W. Bush. In today's Washington Post, CBS conceded that it had not asked his wife to authenticate the letters it claims were written by her husband. Both Killian's widow and son say that the alleged memos are not characteristic of his style, both say he had no "personal file" from which CBS's source could have obtained them, and both do not believe all of them are authentic.
Rather did not mention that Ben Barnes, the Democratic lobbyist who is now saying he helped young Bush into the Texas Air National Guard (TANG), has changed his story according to his Republican daughter, Amy. She says that Barnes is making his Bush claims in preparation for his upcoming autobiography and to build up his political profile in the hopes of getting hired by a Kerry administration, all of which he allegedly told her.
Also left out by Rather was the fact that one of the CBS documents dated in 1973 refers to pressure that then-Col. Walter B. "Buck" Staudt, had supposedly been applying on Killian to make things easier for Bush. Unfortunately for CBS's case, however, Staudt had retired in 1972.
CBS's own paid signature expert (the network featured no typographers or typewriter experts tonight or in Wednesday's report), Marcel Matley, directly undermined CBS's case several years earlier in an essay for the American Law Institute:
"Do not passively accept a copy as the sole basis of a case. Every copy, intentionally or unintentionally, is in some way false to the original. In fact, modern copiers and computer printers are so good that they permit easy fabrication of quality forgeries."
In his defense tonight, Rather admitted that "the documents CBS started with were also photocopies."
The original 60 Minutes report as well as Friday's rebuttal did not feature a single person person who was quoted as coming to Bush's defense who was not on his staff, despite the fact that it is not hard at all to find people who say they served with Bush during the period in which he is accused of being AWOL. The only person that CBS did put on camera hardly provided much support for the documents' authenticity. Rather quoted him as follows (read the rest here.
"Well, they are compatible with the way business was done at that time. They are compatible with the man that I remember, Jerry Killian, being. I don't see anything in the documents that are discordant with what were the times, what were the situation and what were the people that were involved."
Reached by the AP today, Strong was even more lukewarm toward the documents' authenticity. His former colleague, Retired Col. Maurice Udell called them fakes: "That's not true. I was there. I knew Jerry Killian. I went to Vietnam with Jerry Killian in 1968."

Document problems

Although he tried to minimize the typographical concerns raised by many critics, Rather nonetheless tried to defend himself in this area. He failed, however. On the superscript issue, which Rather tried to explain away by throwing out the red herring that "Critics claim typewriters didn't have that ability in the 70s. But some models did."
The problem with this statement is that Rather fails to list any such typewriters which might have the capability or how a measely Air National Guard office would be able to afford such expensive machines. Simply showing a photocopy of a letter in Bush's official file which originated from the Army's national office is no proof at all.
The split screen image CBS offered of an official Bush document with superscript ordinal suffix and one of its own documents was not very convincing to Sandra Ramsey Lines, a forensic document expert who edits the Journal of the American Society of Questioned Document Examiners who told the Associated Press that she "could testify in court that, beyond a reasonable doubt, her opinion was that the memos were written on a computer." She told the AP that she was "virtually certain" the CBS memos are not genuine.
Rather also neglected to mention that all of the documents which were written by Killian himself and his officers relied on simple mechanical typewriters incapable of printing in proportional fonts, let alone superscript.
Despite the fact that Jerry Killian hated keeping notes, hated typing things (see above) that National Guard offices mostly use hand-me-down equipment from the full-time armed forces, and that Killian and his Guard officers have not been observed to have ever sent documents printed with proportional fonts, there is a possibility (OK really, really small) that Bush's superviser might have had access to an expensive IBM electric typewriter.
      Assuming Killian somehow had access to an IBM Selectric Composer (or similar model), Blogger Jeff Harrell wondered what one of the CBS memos would look like if typed in one of the re-famous devices. His results, obtained with the help of the very gracious Gerry Kaplan of IBMComposer.org, are yet more evidence that the CBS docs are forgeries. The letter spacing is markedly different and the superscript (which Kaplan was able to make through a very laborious process of manually removing the typewriter's font ball and switching to a smaller ball and then backing the paper up a few points, typing the numbers, then replacing the original ball and then indexing the paper back down) doesn't line up.
      Harrell and Kaplan also note that two of the CBS documents use a centered letterhead format which lines up exactly when images of the two documents are overlaid. Kaplan notes the following about this:
      "Another point that is very suspicious is the centered heading. This is a snap to do with fixed spacing (like courier), but the text is centered using proportional spaced text, which means that the typist had to carefully measure the text prior to typing to calculate its exact center point. Typing a superscript, with all its steps, is simple compared to centering text proportionally without digital electronics."
      Since Killian's family has repeatedly said that he was no typist and did not like taking notes, it is highly doubtful he would have done such a nitpicky centering job or waited for his secretary to do it for him.
Dan also appears unfamiliar with fonts and typography. At one point in the rebuttal, he refers to the font used in the CBS documents as "New Times Roman," when the real name is Times New Roman. Rather also appears to be ignorant of the fact that Times New Roman was never used in typewriters and only came into wide use in the early 1990s when Microsoft licensed the font from the Monotype Corporation in preparation for the launch of Windows 3.0.
Even if Times New Roman had been used in proportional typewriters during the 1970s, the font then was not the same as it is today since its present form actually dates from the 1980s following some changes that Monotype made to the font.
The Sloppiness Problem. Even if Killian somehow had access to a magical typewriter with the ability to print in a manner almost exactly similar to the default settings of a Microsoft Word document using Times New Roman and equipped with an extremely rare superscript key and compatible font ball, why on earth would the lieutenent colonel have gone to the trouble of using superscript on some lines but not on others?
From our earlier interview with Garry Kaplan, owner of IBMComposer.org: "The person who produced this copy does not appear to have taken the time to properly space things out, such as 'May,1972' has no space after the comma; '(flight)IAW' has no space after the parenthesis. So, it would be hard to believe that they would take the time to produce the superscript 'th' manually."
The Apostrophe Problem. CBS's May 19, 1972 document features what is officially known as a typographer's apostrophe. These characters were not available on electric typewriters according to Jim Forbes, who runs Selectric.org and maintains a database of IBM typewriter font faces.

Logical Problems

In his Friday defense, Rather said the following:
      "Document and handwriting examiner Marcel Matley analyzed the documents for CBS News.       "He says he believes they are real, but is concerned about exactly what is being examined by some of the people questioning the documents. Because deterioration occurs each time a document is reproduced and the documents being analyzed outside of CBS have been photocopied, faxed, scanned and downloaded and are far removed from the documents CBS started with which were also photocopies."
      This statement presents a significant problem for CBS's credibility. If it is indeed true that the documents which are being analyzed on the internet are of much worse quality than the ones that CBS News has in its possession, this naturally leads to the following question: If CBS knew that it was going to come under severe criticism about the legitimacy of its documents (which Rather seemed to state Friday on the streets of New York), why would it ever provide inferior-quality documents? Why run the risk of having diminishing the credibility of the entire CBS News division and star anchor Dan Rather?
      In point of fact, it is very unlikely that CBS would enage in such a positively stupid action. One can only conclude then, that Rather's above statement is a lie.
By now, thousands of people have demonstrated that the documents can easily be duplicated using common word processing software, common fonts, and default settings. Occam's Razor, the idea that the simplest explanation for a natural event is the most probable one, dictates then that this must be the default explanation. Thus, the burden of proof is on CBS to prove that the documents were not made on a computer.
While CBS was able to get a legitimate handwriting expert to say the signatures on the documents are similar enough to each other, this matters little since CBS never had the originals. It is eminently possible that a forger could have taken authentic signatures and pasted them onto a pre-made image and no one looking at a copy could tell the difference. Without having the original (which again CBS does not have and does not ever appear even to have seen) documents, any signature analyses are moot.
Since Killian was no typist, who typed the memos then? Did Killian even have a secretary? Someone needs to ask this of one of Killian's former associates.

It's going to come down to character spacing. Someday soon someone's going to use something like Fontographer to open the font, write down the character widths, numerically figure out the width of the text according to the font, and then check the copy. Then do the same thing with the IBM composer. The forgery won't be all that hard to prove, it will just take a little bit of leg work. I worked at a printing company for many years (and even made fonts) and doing something like this is tedious but not overly difficult. I only wish I still worked there--I'd have the tools and could do a thorough analysis in a day or two.

Perhaps the whole list needs to start with: 0. The "documents" in question can be trivially recreated, with high precision, by using Microsoft Word, under default settings of the program. This places the documents in the late 1980s if not more recent.

I prefer the following:

The "documents" in question are obvious forgeries unless Microsoft's Word Design team used these specific "documents" as templates in the design of MS Word.

This is a subtle, but telling item: The samples with the "th" superscripted clearly show it to be WITHIN the line. It does not rise above the line. This is a trivial thing with inkjet and laser printers, but a huge problem for mechanical printers. Just as a mater of practicality, a mechanical device from 1972 that tried to make a fractional carriage adjustment to get the "th" above the line would never be able to return exactly and reliably to the original line.

Actually, this was tried yesterday with an IBM Composer, and it was unable to exactly duplicate the raised "th". It also failed to duplicate the letterspacing, and no attempt was made at centering, because the problem was too time consuming.

http://shapeofdays.typepad.com/the_shape_of_days/2004/09/the_ibm_selectr.html