This is a story from September of 2004. We don't yet know all the details. We do know that, on September 8, 2004, 60 Minutes Wednesday ran a segment in which Dan Rather presented four memos ostensibly from the commander of George W. Bush's Air National Guard unit. The memos (links below) appeared to contain remarks critical of George W. Bush's service record, hinting that he has be derelict in his duties. After a few of days of analysis by bloggers, it became apparent that the memos as released by CBS were modern forgeries, not documents typed in 1972. The fake documents were an attempt to deceive (successful for a while) and are rightfully called a hoax. This hoax had as its ultimate target all of us - it appeared to be an attempt to influence the upcoming election. Before that deception could occur the hoaxer had to fool CBS News. Seems that was not difficult.
The firestorm of accusations of forgery arose even before the broadcast ended. The controversy concerned the typography of the memos. CBS conveniently posted pdf copies of the scanned memos so we could see them. The issue was that the appearance of the type indicated that the memos were modern forgeries, not memos typed in 1972. We saw an enormous amount of blather about this. All kinds of possibilities were floated to explain how something that looked like the memos could be produced in 1973. Remember that IBM's first PC was introduced in August of 1981.
After the ensuing debacle for CBS, their management assembled a group headed by Richard Thornburgh and Louis Boccardi to investigate the affair and produce a report. The resulting 223 page document appears to cover every possible detail about the affair. On page 222 they summarized the problem.
"Behind all of the Aftermath's missteps lay the fierce conviction of some at CBS News that the story was true and there was a refusal by some to consider that it might be false. That unwillingness led CBS News to ignore mounting evidence - detailed throughout this Report - that there were problems with the documents and the adequacy of the original reporting." (Report of the Independent Review Panel)
We must consider the possibility that "the will to believe" was operating here. CBS staff wanted so much for the story to be true that they ignored the warning signs that were popping, like this one:
"... Second, a respected typewriter expert, Peter Tytell, contacted Miller and Howard and explained in detail why he believed the Killian documents were likely fakes. His views were not pursued or analyzed in part because 60 Minutes Wednesday was searching only for experts who would defend the September 8 Segment.This behavior might also be described as "Confirmation Bias," a well-known pnenomenon of ignoring contrary data." (Report of the Independent Review Panel)
There was a lot of blather about a supercript "th" that took up time. It did give Peter Tytell a clue. Comparison of the memos that CBS posted with the output of Microsoft Word for the same text shows an exact match, character appearance and spacing. They are identical. Add to this the fact that CBS located the National Guard secretary from Col. Killian's office, the one who would have typed the original memos. She said that her typewriter was an Olympia manual typewriter.
Peter Tytell told the panel:
"On September 9, Tytell had downloaded the Killian documents from the CBS News website and told the Panel that he began 'working on the case out of curiosity.' Tytell told the Panel that he told Miller on Friday, September 10, that the documents aired on the Segment were prepared in Times New Roman, a typeface available on modern computers but one that did not exist on typewriters in the 1970s." (Report of the Independent Review Panel)
Appendix 4 of the Report sums up Peter Tytell's finding nicely.
"Tytell concluded, for the reasons described below, that (i) the relevant portion of the Superscript Exemplar was produced on an Olympia manual typewriter, (ii) the Killian documents were not produced on an Olympia manual typewriter, and (iii) the Killian documents were produced on a computer in Times New Roman typestyle . Tytell acknowledged that deterioration in the Killian documents from the copying and downloading process made the comparison of typestyles 'to some extent a subjective call.' However, he believed the differences were sufficiently significant to conclude that the Killian documents were not produced on a typewriter in the early 1970s and therefore were not authentic." (Report of the Independent Review Panel)
What we can do is establish what the CBS memos were not. To do this all we need are facts established above.
We finish this evidence with a comparison to MS Word.
Image from Wikimedia. Public domain.
This animated image is by Charles Johnson showing one of the memos CBS used in alternation with what was produced with MS Word with default parameters. Word produced an exact reproduction of the CBS memo. This can happen only if the characters in the font match exactly in size and shape and the space increments also match exactly. If the character widths and/or space increments did not match, the difference would be obvious.
Notice that, with this evidence, it is not necessary to consider the actual content of the memos. Nursery rhymes would do as well. CBS personnel ignored or made excuses for all of this.
Concluded that the CBS memos were forgeries, what else can we say? Some, but not all, details are reasonably well-known.
There are still important details that are not known.
Message for students: a journalist/reporter falling for a hoax of this scale and refusing to acknowledge evidence of a hoax is at serious risk.
If you would like to see the fake memos for yourself, here they are.
For deep details of the typographical analysis, see Dr. Joseph Newcomer's excellent dissection of the typography of the memos. He demonstrates quite scientifically that the memos were forgeries. He actually outlines, then uses, the scientific method to do it.
The Washington Post studied the memos and found numerous reasons to conclude that the memos were fakes (thelink is now gone).