god’s way of teaching people about typography

webliography updated briefly on 2004-10-04


significant studies

  • bush memo font study by david e. hailey, jr. / utah state university [pdf] — basically concludes that “all evidence” points to mechanical production, not digital production; but that in any case, based on typographic analysis alone, CBS was well within bounds of reasonability to accept the documents as important. This study was then critiqued by wizbang [9/30] [updated 10/2] [wired 10/7 covered the wizbang-inspired blog-attacks on hailey


  • funny bush memos: funny

blogging about it

mainstream media investigations

stray thoughts & listing the observations that have been illuminating:

  1. on recreating the document in MS Word: the default in MS Word does a few things that do not show up in these documents:
    • the superscript looks pretty different from the superscript in the memos. the superscript in MS word starts exactly halfway up, is smaller, and the top of the h goes just under the top of the 187. The superscript on the memo starts higher — closer to the top of the 187 — and goes way above it.
    • when you start a paragraph with a number MS Word tries to automatically create an indented paragraph.
    • on the other hand, folks on both sides of the issue have argued that it would be too much effort to turn superscripting on sometimes & not on other times. As for MS Word, MS Word defaults to replacing your “187th” with “187th” for you. However, as soon as it does that, if you do a CTRL-Z, you undo the superscript. Probably most folks don’t automatically do it and may not realize it can be so easily undone. But if you realize it’s pretty easy, without having to go into the menus to turn it off. As for typewriters, it depends on how it was done–for my money, it’s probably more likely that you would get some-of-them-super, some-of-them-not, in a typewriter scenario. Because MS Word does it automatically, and so they will all be superscripted, unless the typist de-superscripts the text. If the MS Word typist desuperscripts the text, they are probably de-superscripting ALL the text–because either they’re doing it via the menus which would just turn off the auto-superscript, or they are doing it with the CTRL-Z which is an easy fix and you get in the habit of doing it. If on the other hand someone is a typewriter typist and they do it sometimes but not other times, it’s not worth it to re-type the memo just to fix it. … By the way, this points out that despite claims to the contrary, you can NOT simply reproduce the documents in MS Word without doing anything but typing.
    • I also found this to be a reasonably persuasive argument [i think from dailykos]: that a typist gets in the habit of automatically doing the shift-upper or whatever for her own unit (the 1187th or whatever) but less in the habit of doing so for other units typed less frequently. Therefore the pattern of superscripting that you see in the memos suggests from a human perspective that they are typewritten not word processed.
  2. Thoughts about faking documents:
    • Why would someone do a fake that was generated from MS Word default settings? It sounds too stupid and makes me doubt that it’s actually a fake. In general, I think that if someone were really trying to put across a forgery, they would use a period typewriter. It wouldn’t be hard to do. So, if the memos really are forgeries, and the forgery is really amateurish, it tends to make me consider about a Karl Rove-type fake-fake.
    • On the other hand, if the technology is roughly available, then I think it’s easiest just to believe the memos are really real. Why? Because in a time of changing technology there are just naturally lots of variability. In other words, almost any historical document produced on an evolving technology could be subject to allegations & disputes about its authenticity. Most such documents are not because (a) they are not examined closely and (b) there is no reason to examine them closely & no political reason to want them to be fake or real. But if you apply the same standards to other historical documents that you apply to the Bush TANG memos my sense is you could generate controversy about all kinds of docs. Thus, unless there is a smoking gun, I’m skeptical.
  3. The typographic smoking guns that I’ve seen thus far, the simple and obvious proofs, have all been blown away:
    • no superscripting on typewriters at the time: no, the contemporaneous, unquestionably authentic bush document had a superscript at the time
    • this is a font (Times New Roman) not available on typewriters at the time: no, this is not Times New Roman, although it is similar, and in any case, TNR was available at the time
    • this is not fixed-width and only fixed-width was available: while it appears to be proportional font, there was proportional fonts available then
  4. Fuzzier typographic guns: so maybe there’s no single hardcore piece of evidence. there may still be circumstantial evidence or, while typewriters that could do (A) were available and typewriters that could do (B) were available, there were no typewriters that could do both (A) and (B). My general response here is that fuzzier logic generates less certainty & more skepticism. I’ve seen nothing that is extremely persuasive.
    • lines up exactly with MS Word default: (a) this is merely circumstantial, but would be somewhat persuasive; but (b) it’s demonstrably false (I demonstrated it just a few minutes ago on my own machine)
    • leading & kerning comments: (a) i simply haven’t seen an extremely persuasive & undisputed argument about this. i’ve seen a lot of debunked and rebutted arguments about this issue. (b) the leading & kerning seem susceptible to the problems caused by multiple duplications.
    • centering: there were center bars on typewriters when I was young & it’s not that hard to center paper exactly time after time on a typewriter
  5. Typographic evidence for typewriters: Some evidence actually suggests typewriting.
    • Floating letters: some letters float up & down a la typewriters not regular word processing
  6. Non-typographic evidence:
    • social factors: family members. I discount the family members. If a family member & a co-worker disagree about somebody’s behavior at work, then I think the co-worker (who actually observed the actual behavior, and who actually observed the typical behavior at work) is more trustworthy. People all too often present different faces at home. … Plus in this particular instance I’m suspicious: why would the guy’s son have such a good memory, 30 years later, of his dad’s glowing review of the Bush son?
    • social factors: co-worker who sort-of reneged. The story I’ve heard is that he said that CBS said the memos were handwritten by Killian, and he said that if that was so, then they reflected his feelings. This dispute is a little too fuzzy & subject to misinterpretation on both sides, and any set of interpretations could actually generate no controversy. In other words, it sounds like CBS, at most, suggested the memos were written entirely in hand instead of merely signed by hand. And the guy is saying that he gave a qualified endorsement if they were written by hand. But the guy isn’t saying that they seem likely to be untrue if they were not written by hand. So he’s not really disputing the memos’ authenticity. He is saying that he is less certain of them if they were not written by hand, in Killian’s handwriting. … Just on the face of it, this dispute neither adds nor subtracts much from the debate.
    • social factors: alleged string-puller wasn’t around at the time: retired military still has pull — surprise!
    • enough doubt on typographic issues: no, casting doubt on authenticity is perfectly plausible as a political tactic; since there’s a good reason for the controversy that may have nothing to do with the actual authenticity of the documents, the actual existence of the controversy itself cannot be used to justify the controversy. hmmm… this reminds me of Bush Sr.’s comments on Kerry’s military record: While he himself didn’t have any reason to doubt Kerry’s valor, since there was so much controversy he thought there had to be something to it.
    • who writes “CYA” as a subject of a memo anyway? as i commented on one list, that does sound unlike any memo i’ve seen or received, and some of the revelations seem conveniently worded. the memos don’t fall (in my book) on this evidence though, for two reasons: (1) my own social circles do not (currently) include enough folks in the military to say whether this would be a likely kind of memo for the TANG in the 70s; (2) in any circle, someone who is pissed off at a situation might well do a CYA file. i’ve certainly consciously created paper trails in various work situations where i felt it might be prudent. and so i don’t find it implausible that someone pressured by a superior might do likewise.

so do i think they’re fakes? maybe, maybe not. i am not in a position (nor is anybody else in the blogosphere so far as i can tell) to easily discredit these memos. and it’s virtually impossible to prove they’re not fake, because how to prove a negative?

the significance of this flap therefore lies in whether it will (a) add incremental evidence to the story that george w. bush blew off his national guard duties; or (b) make somebody (probably a kerry supporter) look bad.

and, of course, how significantly it distracts us all from the war in iraq and the economy and our disappearing civil liberties and the national disgrace that has been our treatment of prisoners.