Forensic Typography and the 2004 Campaign

Written by Scott Stowell on September 10, 2004

At this year’s TypeCon in San Francisco, Matthew Carter told me about a court case that he was involved in as an expert witness. I won’t give away the details for fear of upsetting the legal process, but let’s just say that his familiarity with type design (and particularly one of his own typefaces) and typesetting (and the machines used to do it) enabled him to easily prove or disprove the authenticity of some historical legal documents.

Cut to the discovery this week by CBS’ 60 Minutes of some very interesting documents that tell the story of George W. Bush’s National Guard service — or lack thereof. But the authenticity of the documents has been called into question due to certain typographic anachronisms. They’re set in what looks like Times, with proportional spacing. There are superscripts used in the text. And to my eye they exhibit the casual ugliness that comes from a very special source: Microsoft Word.

But don’t take my word for it. The Cyberspace News Service gathered opinions from a few typographic experts, including Allan Haley from AGFA Monotype (“It was highly out of the ordinary [in 1972] for an organization, even the Air Force, to have proportional-spaced fonts for someone to work with”) and John Collins of Bitstream (“[the superscript] would not be possible on a typewriter or even a word processor at that time”).

If these documents are fake, who faked them? One theory in our office is that the GOP did it to cast doubt on the Kerry campaign. As Richard Polt, a Xavier University philosophy professor and typewriter aficionado was quoted in the Weekly Standard, “I’m a Kerry supporter myself, but I won’t let that cloud my objective judgment: I’m 99% sure that these documents were not produced in the early 1970s.” He took the words right out of my mouth.


  1. David says:

    That’s what happens when you try and cut corners and do it yourself!

    I hope this is a lesson to people who think they can build a website, make a logo, typeset a newsletter or forge a military record to an acceptable standard just because they have a computer. The software doesn’t come with skills.

  2. optimus says:

    I have no doubt that GW wasn’t much of a National Guardsman, but I have to concur with those who call the memo a fake.

    After all, if you look at White House documents — at the presidential level, no less! — of the same era, they don’t have the advanced typographic features (superscripts, proportional-spaced fonts) that this low-level USAF memo had.

    That said, is everybody sure about this no-superscripts-in-’73 thing? I’ve seen an overlay of the document with one produced in Word, and the superscripted ‘th’ aren’t at the same height. Didn’t the Selectric II allow for superscripts?

  3. Miriam Frost says:

    I haven’t tried it myself but from what several others have said, the th height is different between the screen and printer font; printed, then scanned back in, the height is the same between the Word reproductions and the memo.

  4. John Butler says:

    One theory in our office is that the GOP did it to cast doubt on the Kerry campaign.

    I too have theories. I have one that states that by running around in circles around the North Pole and crossing the International Date Line over and over again, I can travel backwards or forwards in time.

    I have another one involving the use of fax machines to transmit beer.

    I would like to see CBS’s “expert document analysts” sign their names to their conclusions. Or just appear on the air and say “Yup, that’s real alright.”

    Proportional typewriters from the early 70s (cutting edge and mechanically amazing) were probably not casually used for everyday memos. Besides, isn’t all military correspondence still stenciled?

  5. Alastair Keady says:

    No doubt in my mind that it’s a fake (and god knows I’d love to claim otherwise).
    Aside from the superscript issue (You could do superscript as an offset with the Selectric II, but only at the same point size), and the similaritys with the default settings in Word, there’s the small matter of the apostrophe in ‘Bush’s’. Even if there was an IBM composer in their office at the time, why would a military guy or his secretary indulge in the fine art of typesetting a proper apostrophe when the dumb apostrophe was much handier and quicker?

    Just saddened that the lowest, bigoted, dirtbox of a ‘blog’ gets some profile from this story.

  6. Type It Up says:

    Actually, the IBM Executive offered proportional spacing, and was first marketed in the 1940s. And the Times New Roman font, far from being a new innovation, was created in 1931. So, yes, this could have been done on a typewriter in the early 1970s.

    More info here.

    Oh, and there’s also the IBM Selectric Composer, a desktop compositor, which was tested by the Air Force in 1969. So, why is it totally far-fetched to think Killian’s secretary could have had one of these machines a couple of years later?

  7. Alastair Keady says:

    I don’t doubt that you could have used a Composer to create the letter (not at all sure you could reproduce it on a regular Selectric), but it’s completely implausable that a secretary would go to the trouble of using proper apostrophes, centre some of the text, and juggle golfballs for the superscript element.

    It’s not Times New Roman used on the letter in any case.

  8. Type it up says:

    I agree it’s not Times New Roman. That alone should raise red flags about the “experts” quoted by the news media.

    In terms of what a secretary would do, I disagree compeletely.

    First, the IBM Executive was a typewriter designed to make typed work look typeset. While I can’t yet confirm this, I believe it may have had dedicated keys for ordinals. So, it wouldn’t have been extra work to do this. Also, manually centering text is a cinch; I learned how to do it in Junior High typing class.

    Additionally, there is evidence that this was typed by someone trained to touch-type on old typewriters. For example:

    1. The lower-case “l” is used for “1”.
    2. There are two spaces after each period. That used to be standard practice, but I think that most style guides now suggest using only one.
    3. Word wrapping appears to have been done manually. Take a look at the line starting “MEMORANDUM” in the August 4, 1972 memo. A computer would have wrapped after “Texas,” creating a widow. A good secretary would catch that, and hit the carriage return earlier, as shown here.

    Another clue that leads me to believe this was done on an old machine: Look at “P.O.” in the headers. Modern word-processing programs kern pretty aggressively after a period; try typing P.O. in Word. In this document, however, there’s a fair amount of space after the first period, before the O. I’d chalk that up to the non-digital technology used in machines like the IBM Executive.

    This document marks the first time typography has become front-page news in a long-time. Let’s really check this out before agreeing with the so-called “experts” who’ve already concluded this is a forgery.

  9. John Butler says:

    Better yet, let’s start a pool. $200 says it is indeed a forgery. Any takers?

    (We’ll have to settle it somewhere legal. There’s an Indian reservation not far from where I live.)

  10. Type It Up:

    The reason it doesn’t look quite like times is that it’s been run through a fax or something. I’m convinced that it is some version of Times. I’ve made a comparison image. This was made in Illustrator by typing some of the memo in Times New Roman. I scaled all three pieces of text by matching the length of the longest line. As you can see, there is a very close correspondence to where the characters fall. It’s also clear that the general design of the characters in the memo (i.e., placement of serifs, overall letterforms, etc.) are consistent with Times, although distorted by a fax or whatever.

    As to whether the typist used lowercase L’s in place of 1’s, I seriously doubt it. As you can see here, typing it with L’s (top line) looks very different, while the standard Times 1’s are a good match.

    Inspite of years of typographers telling people to stop using double spaces after periods, most people still do it, and Word doesn’t care.

    Centering lines with a proportionally space font will not work using the method you learned in typing class because the letters are not all the same width. In any case, if this was indeed typed on a more sophisticated IBM typewriter, you wouldn’t use that method anyway.

    Also, I’ve yet to see a word processing program that supports automatic kerning (I turned it off in Illustrator for these examples).

    Finally, although I do think these are fake, it is true that the IBM Composer had “Times”, exept it wasn’t called that, it was called “Press Roman” and was for practical purposes identical.

  11. Hrant says:

    > $200

    Give it to the Indians; it would be a small part of the reparation you personally owe them.


  12. Roger Black says:

    The IBM Executive had no fonts (or type bars) that were Times Roman, or even the Press Roman of the Selectric Composer. But it is vaguely possible that these documents were printed on some kind of line printer or dot-matrix printer in the 70s, running on an army word-processing system. Possible, but doubtful.

  13. jlt says:

    Although you don’t need to be no fancy typographitator to see that this is for sure the geniune article.

  14. that this is for sure the geniune article

    I needed a good laugh and that did it. Actually, that crack about reparations did it too.

    I’m following this thread with fascination. All this talk of ancient typewriters (why do I now want an IBM Selectric Composer?), character comparisons, apostrophes, superscripts and typographically conscientious secretaries just makes me, well … you-know.

    I’m a little confused about who’s “winning” at this point, however.

  15. N. T. Pryzer says:

    It appears from the following webpage that the Times New Roman font used in Microsoft Word, a program which I believe began to see use in the 1990’s, was based on a refinement achieved in the 1980’s.

    Thus, if the font in the 1972 questioned memoranda exactly matches that of Microsoft Word, wouldn’t that be a double anachronism? (I.e., anachronistic relative to both the 1980’s refinement and Microsoft’s later improvement of same.)

  16. James says:

    Sadly, even though I am a long time democrat and Kerry supporter after looking at these documents they are pretty clearly bogus. Well, I hope this whole thing just blows over the next couple days.

  17. Christopher Day says:

    Would anybody here care to comment on this analysis?

  18. Felix Deutsch says:

    I think you may have been conned.

    The “Cyberspace News Service” (actually “Cybercast News Service”)is a rightwing PR organisation.

    Excerpt (read the whole thing; get the daily pass):
    “In the release, Creative Response promoted a Web site called Cybercast News Service, one of several groups directed by Brent Bozell, a longtime right-wing activist who has devoted years to attacking the “liberal bias” of the mainstream press. His Media Research Center and other similar efforts have been heavily funded by conservative billionaire Richard Mellon Scaife.”

    Or are you in on it?
    Quoting the Weekly Standard in addition seems to suggest that.

  19. Guest says:

    I thought the 8’s in the CBS document looked unusual; the top loop being smaller than the bottom. Can anyone tell me if that has any significance, or am I imagining things?

  20. Edward Mendelson says:

    This story at the PC Magazine web site may be relevant to the question: Does the fact that the disputed documents can be reproduced (more or less) in Microsoft Word indicate that they’re forgeries?

    The answer is that it indicates nothing of the kind. The evidence is on that page. (I wrote the page, but I think the evidence presented there speaks for itself.)

  21. I don’t know what this guy at is talking about. The IBM Selectric typewriter had only two kinds of “fonts”: 10 pitch and 12 pitch, meaning either 10 or 12 characters per inch. These fonts were monospaced, not propotional, meaning that every character is exactly the same width, including the spaces. The sample documents use a proportionally spaced font. You don’t need to be an expert to see that the M is much wider than the I, for example.

    The IBM Executive was available at the time and did have proportional fonts, but, as Roger Black pointed out above, nothing similar to Times Roman.

    The IBM Composer did have a font that looked like Times Roman called Press Roman, but these machines were low-end typesetting machines based on the same mechanism as the Selectric. They were not typewriters–they were cheap typesetting machines. The machine cost $150 a month to rent in 1971 and if you wanted to own it, the base price was $4,400. (This information comes from an article that appeared in The Mother Earth News in 1971 about the feasibility of going to business as a typesetter in your home.) People did not buy these to type memos.

    I don’t have any theories as to why anyone would fake these documents. If it was Kerry supporters, it was pretty stupid. If it was Bush supporters, it was very clever. It does seem though that a lot of people commenting on them don’t know what they’re talking about.

  22. Sophie says:

    I would be interested to hear an expert skeptic’s response to this interesting argument about the memos (also at Daily Kos). It points out:

    1) Character baselines ‘float’ up and down – typewriter style.

    2) It’s a close relative, but it’s not Times New Roman

    3) The superscript ‘th’ was available on typewriters produced in that time period.

    4) Proportional spacing was available at the time.

    5) The IBM Executive typewriter could well have incorporated all of the features that have been called into question.

    6) In the 1960’s, IBM asked Stanley Morison to adapt his Times Roman font for their typewriters, and “One of the differences between the Times New Roman as implemented on the IBM machines, as opposed to Microsoft Word? The IBM machines apparently had the alternative ‘4’ character that matched these memos, while Microsoft Word’s TNR does not.”

    7) “Now, would the 111th Fighter Interceptor Squadron have extravagantly purchased typewriters that contained the th superscript key?”

    8) The writer quotes someone who said:

    Kevin, I worked in the IBM Office Products Division field service area fixing typewriters in NYC for over 13 years in the ’70s. I can tell you that the Model D can produce those documents, not only did it do proportional spacing, you could order any font that IBM produced AND order keys that had the aftmentioned superscripted “th.” Also you could order the platen, thats the roller that grabs the paper, in a 54 tooth configuration that produced space, space and a half and double spacing on the line indexing, this BTW was popular in legal offices. The Model D had to be ordered from a IBM salesmen and was not something that was a off the shelf item, typical delivery time were 4-6 weeks. Also, typewriter keys were changed in the field all the time, its not that hard to do. I wish I had saved my service and parts replacement manuals to backup this claim but I’m guessing a call to IBM with a request for a copy of their font and parts replacement manuals would put this to rest ASAP. Posted by: BillG NYC on September 10, 2004 at 12:24 PM

  23. Cherie Benoit says:

    Watch the CBS follow up report. Focus on the bigger issue here. It’s not about “are these documents fake?” The whole story remains that Dubya has not been honest about his guard service or lack thereof.

  24. One comment mentioned to me about the apostrophe problem was that if a proper typographer’s apostrophe was not available on the keyboard, one could roll the paper back a notch and type a comma. Not saying it’s probable, but it’s possible.

  25. Lee Mori says:

    Are you sure no type similar to Times Roman was available for the Executive? At least one former IBM employee claims that the Executive C or D could produce proportional type…in any font IBM produced. And IBM did make a Times Roman type font. You had to special order it, but I gather that wasn’t uncommon for the military. The brass liked to have the latest, greatest, and most expensive. Some people are also claiming that the military sometimes got special versions of equipment, such as typewriters, made just for them.

  26. Are you sure no type similar to Times Roman was available for the Executive?

    Well, that’s what Roger Black said. If you could get something like Times for the IBM Executive, that would make these documents more plausible. I would be curious to know exactly what fonts were available for it. Also, my understanding is that the Executive did not have interchangeable fonts, the way the Selectric did.

    I wish there was more information on these old machines on the web. I know type very well, but I have only a little first-hand knowledge of the IBM Executive typewriter.

  27. Hrant says:

    Hey, did you guys know that Centaur was once produced as a typewriter font?

    (Uh, sorry for the interruption. Back to your regularly scheduled distraction-tactic nitpicking.)


  28. Hrant, I agree. This whole thing is stupid. It just drives me crazy when people who know nothing about type act like they do. That’s my excuse, what’s yours? :-)

  29. Hrant says:

    You mean my excuse for participating? I’m just the sniper here. Like the low-wage troublemaker ones at the Mat·haf crossing in Beirut during the war. Aaah, those were the days.


  30. John Butler says:

    And yet, still no one has accepted my wager. I ask again, who will wager $200 that the documents are not straight out of Microsoft Word? Put your money where your mouth is.

  31. Sophie says:


    How about $10,000?

  32. Sophie says:


    I sure wish I had one of these!

  33. Gerald Lange says:


    Haven’t seen one of those in about a decade. I used to work next to a business that was selling surplus office equipment and he was getting about $50 for IBMs. Then Apple’s promise came true and the bottom dropped out of the market. I remember helping him haul all his remaining stock to the scrap yard. After that he just threw them in the dumpster whenever one came in.

  34. Don Munsil says:

    Take a look at this 1953 ad for an IBM Executive typewriter. I would call the top type sample (IBM Bold Type) a Times-like face. I wish they had a better scan of the ad, though.

    1953 ad

  35. Gerald Lange says:

    There is an IBM Composer on eBay for 99 cents. Not a bad investment if this would prove the point and the fellow offering the reward would pay.

  36. Sophie says:

    The brouhaha could revive interest in this classic title…

    author = “Christina Gardner”,
    title = “Typesetting with {IBM} Executive typewriters”,
    publisher = “Gardner Printing”,
    address = “Santa Rosa, CA, USA”,
    pages = “180”,
    year = “1967”,
    bibdate = “Mon May 20 11:04:41 MDT 1996”,
    acknowledgement = ack-nhfb,
    annote = ““Sources for further information”: p. 178.”,
    keywords = “Typesetting machines.”,

  37. Gerald Lange says:

    By the way, in case CBS is monitoring, I have a 8x Anchromatic Magnifier and Scale for Measuring Type Sizes that I will sell for a mere $5,000. They don’t even have to fly me out there, I’ll just ship it and let them claim discovery.

  38. Hrant says:

    John, if you’re patient enough the $10,000 your owners seem to be willing to throw around will trickle down… down…… down……… and help your peonic $200 become respectable.

    All this betting bravado is petty grandstanding, in the classic cowboy style. Do you actually think this could ever be resolved with some kind of Proof, in any “non-partisan” sense of the word? This sorry affair has already achieved its goal: to pull your eyes away from the flames of war engulfing you.

    As if anybody with half a brain can even think that any of these politicians hasn’t lied his way to where he is in the first place. A single forged-or-not memo indeed.


  39. Beltedswiss says:

    As someone who worked on newsletters and tech pub production starting in the early ’70s, I personally worked on IBM machines that had the proper font and could produce proportional spacing. And those machines weren’t new then, either. Yes, superscripts too, though it took some doing. “No Times Roman available,” my left earlobe.

    Friggin’ armchair typographers.

    And what in the name of God’s green earth are you doing depending on anything from Cybercast News Service? Do a Google search on the search terms “Cybercast News Service” and you find it listed as a conservative news resource. Halfway down the page. What Felix Deutsch said above — CNS was founded in 1998 by Brent Bozell.

    In June 1998, Mr. Bozell launched the Conservative Communications Center (C3) to provide the conservative movement with the marketing and public relations tools necessary to deliver its message into the 21st century. C3’s online news division, the Cybercast News Service at, has become a major internet news source with a full staff of journalists in its Washington, DC metro bureau, and operates bureaus in London and Jerusalem, with other correspondents around the world.” That quote is from Brent Bozell’s bio at

    Please don’t let yourself be conned any further.

  40. As one of the few registered Republican typoholics, I have to say that I was really excited to see what everyone had to say about this. I’m sorry, I just couldn’t resist. Some of the explanations here are just spectacular. Someone should call Dan Rather’s experts over and we can have a party.

    Realistically, how many people here really think that 70’s era Texas army secretaries used ‘smart’ quotes in their typing? Honestly, this is the coolest thing that’s happened in weeks.

    Let’s just say, that as an unabashed partisan and as a type nerd, I’m gleeful to see all the coverage on this. Even CNN is getting in on the action.

    What’s more there are “secret sources” involved. The alligations are flying everywhere. Who cares if they are true or not; this is just pure election fun.

  41. Lee Mori says:

    I’m curious about the experts that the media have consulted. The ones who claim this is Word. (To me, it’s clear that it isn’t. It may be word-processed, but it is not MS Word’s version of Times New Roman. There are clear differences.)

    Anyway, judging from Google, one of these experts seems to be UFO investigator when he’s not doing type forensics, the other is a declared Republican partisan.

    Marcel Matley, CBS’s expert, says the documents are legit, and that others are being fooled because they are looking copies of copies of copies, that are severely degraded. He does seem to be a well-known forensic document examiner, but it looks like he might be more of an expert in handwriting than type.

    So, does he have a reputation in this field? Does he know what he’s talking about?

  42. Sam Pratt says:

    The funniest comment I’ve heard on this topic was that “this is God’s way of teaching people about typography.” (Not that I believe in gods or anything.)

    This whole absurd episode only teaches us that the media and the general public are utter design illiterates — and will obsess over any red herring that Rove throws their way instead of dealing with the possibility that the President withheld information or acted in a less than honorable manner.

    The pathetic “experts” offered by news organizations such as the Washington Post did nothing to clarify the issue, because they failed to investigate 90% of the issues raised in this one web thread. Yeah, I’d pay cash for seats in any courtroom where Matthew Carter testifies on forensic typography — but that’s not the issue of national importance here.

    Here’s a very helpful graphic from Julius Blog. It’s an animated GIF, so be patient.

  43. The superb ScrappleFace has the issue covered as always.

  44. I think you may have been conned.

    The “Cyberspace News Service” (actually “Cybercast News Service”) is a rightwing PR organisation.

    Or are you in on it?

    Quoting the Weekly Standard in addition seems to suggest that.


    I thought the last paragraph of my post would have made that clear.

    Watch the CBS follow up report. Focus on the bigger issue here. It’s not about “are these documents fake?” The whole story remains that Dubya has not been honest about his guard service or lack thereof.

    Sure. But fake documents won’t get Bush out of office–and I doubt any talk about his service will either. We’re discussing these documents here since this site is about typography. But elsewhere there are even bigger issues on which to focus. Remember the economy, civil rights and the environment? If you want to defeat Bush how about we all talk about those things, not attacks on his past that are too little too late.

  45. C. Cone says:

    There’s an IBM Selectric Composer for sale on eBay.

    And cases of golfballs for it were auctioned at the end of August and beginning of September.

  46. P J Evans says:

    I worked at an electronics compy in the late ’70s — we used an IBM Exeecutive for making labels. No golfballs, but it used film ribbon and had multiple space bars, which could give the effect of proportional spacing (and made it really hard to backspace decently). We didn’t use superscripts, and I don’t remember if that typer had anything unusual on the keyboard.

  47. Barbara says:

    I’m not a typographer, but–

    The Navy had proportional font typewriters, which I used in 1978. They were IBM Executive models, and the ones we had were several years old when I used them. Here’s a site that gives some dates IBM typewriters were made. It appears the Executive model was available in 1971. The Selectric Composer, also proportional, was available in 1966. We had one of those in 1978 as well, but a Composer would’ve been much more expensive and less likely to be used in that command office.

    My use of a typewriter like this was justified because I typed technical manuals. I’m sure there would’ve been other reasons to allow purchase of such a typewriter by the command office in question, if they wanted one. It’s certainly possible, though somewhat unusual.

    After reading about this controversy, I also experimented with the superscript function in MS Word, and confirmed the superscripted “th” in the 04 May 1972 memo was not created in the version of MS Word I have. It’s placed too high. However, in the office where I worked in 1978 (not 1972, mind you) we could get “rub on” characters on glassine sheets to add unusual symbols such as degree signs. The IBM Executives also had removable type bars to provide alternate type, but I don’t know what was available, and the “th” in the memo looks crooked, so I’d guess it was a rub on.

    The font doesn’t look like Times Roman to me. Times Roman is more compact. It could be a less compact modern font, such as Bookman Old Style, but it reminds me a lot of the type we produced on those IBM Executives. However, it’s been years, and again I’m no expert in typography.

  48. Sam Pratt says:

    Anyone who is still persauded by these bogus forgery allegations, please — just read this piece at Daily Kos first before posting.

  49. About the Daily Kos stuff: I love how everyone who has ever worked in a type related field, or has any design training is all of the sudden a forensic document expert.

    Here’s the best blog entry I’ve found so far. He actually found someone with the Selectric Composer to test it out. The results are as I assumed. The face (Press Roman) does, in fact, look a lot like TNR, but the metrics aren’t the same as those in the infamous docs. The comments are also more illuminating than usual.

    The one thing I take from this is that you can’t underestimate the power of political/cultural identity in shaping thought. In all of the blogs, news stories, newspaper articles, and cable ‘shout shows’ I’ve seen in the past couple days (and believe me, I’ve seen a lot of them), almost never did anyone support a view that crossed their team affiliation. People will sometimes grudgingly change their view, but it takes a true preponderence of evidence.

  50. John Butler says:

    Mr. Pratt, if you’re so sure, place your bet. Still no takers.

  51. alastair keady says:


    If it’s any consolation, I’m as anti-Bush as they come, and I thought the letter was a fake from the get-go, and said so on various blogs.

    It IS a great insight as to the veracity (or lack) of information put out in the media in general. Buyer beware!

  52. In all of the blogs, news stories, newspaper articles, and cable ‘shout shows’ I’ve seen in the past couple days (and believe me, I’ve seen a lot of them), almost never did anyone support a view that crossed their team affiliation.

    Wait a minute! Didn’t anyone read my original post?

  53. typenerd says:

    As early as 1964, when I was in 5th grade and obsessed with office machines, our school had a few IBM Executives. They were typebar-based machines, had proportional spacing, used film ribbons, and produced documents that fascinated a little kid who was a typewriter nerd. While they were the “high-end” machines in the offices, it should be noted that if a small NJ prep school was using several of these in ’64, it is very likely that the TANG would have been using them by ’72.

    The CBS documents could have been, and very probably were, produced in 1972 on an Executive. My recollection of the standard typeface was that it was slightly more extended than contemporary Times New Roman (I remember “O’s” being almost perfect circles); it was a serif face with many Times characteristics.

    The idea of custom typebars for superscripts is not the least bit farfetched. There were other typewriters in use at the same school, in science offices, that had the “degree” symbol, and subscript numbers for chemical formulae. And, every secretary was able to center headings on the Executives on a regular basis. I myself, when I was finally lucky enough to get to use one of these Executives, was particularly impressed with my own nerdiness as I figured out how to do this.

    Any attempt to discredit these memos on the basis of typography is destined to failure. The memos, exactly as presented, are not only a typographic possibility but in fact are a typographic likelihood.

  54. Rather than resorting to anyone’s recollections or general feelings, can anyone actually produce a typewriter made before 1972 that is capable of rendering typography consistent with that seen in the memos? I have never known of such a machine, and would be profoundly surprised to find that one existed.

  55. Matthew Carter weighs in at Design Observer, and true to his rational form, dispenses with guessing or speculation despite his experience.

  56. I’m surprised that people keep refering to the IBM Selectric Composer. It was marketed as a typesetting machine, not a typewriter. Sure, you could use it to type a letter. You could have used Chartpak rub-down type, too. You could even get Times Roman from Chartpak. I just don’t think the Composer is was likely to be used for a typewriter back then any more than someone would by an Avid video system as a general office computer now. Sure, you could do it, but it would be a waste of money.

    The IBM Executive is a different matter, but nobody seems to be coming up with any cogent details about it.

  57. Paul Gough says:

    Comments are considering the IBM selectric composer and selectric as possible alternatives to the proportional-spaced executive.

    My experience at that time with secretaries who used the IBM Executive — and they generally were the executive secretaries, the best ones working for the boss not the run of the mill ones — they were skilled and pretty proud of the way they typed things so they might have worked to get superscripts and subscripts. Also, bear in mind that many times — even now — memos to file were dictated by the boss and typed by the secretary.

    The probability of a NG unit using an IBM composer is pretty small, so small that I feel attention should focus on the executive and its capabilities in the hands of a skilled user

    I do wonder, however about typographic quotes. I doubt many secretaries would have thought about them when typing.

    Look at much work composed on computers and you see stuff produced now often fails to have typographic quotes — even in national magazines and on billboards!

  58. Fontboy says:

    A pioneer of computerized typesetting makes his detailed analysis. Verdict? Fake.

  59. joshua says:

    Idea that this is a “Dirty Trick” by GOP is extremely doubtful, unless it is as an obscure Red Herring to sidetrack (you) truthseekers. Bush already admitted that he was a druggie who didn’t exactly serve his country with con mucho gusto — and then he got elected in 2000. When the GOP does a dirty trick, it is clear and strong. Fact is, for people who vote for a president who supports increasing mercury in their drinking water — and no offense meant toward the right-minded typography set — this IBM typewriter nail-biter ain’t gonna be the thing that turns them red (maybe it’s the mercury? Now there’s a conspiracy!).

  60. John B. says:

    Nicely put, Joshua.

    I wouldn’t put it past the TNG to have an absurdly expensive typesetting machine. Bear in mind that the government also puts either far too much money into a project or not enough.

    Also remember that this was the Champagne Unit, organized for “sons of privilege” to avoid risking their lives overseas. As such, I’m sure they had the best of everything.

  61. jlt says:

    Everyone’s infected by and acting pursuant to war psychology anyway, ignoring the facts and defending the party line. Every day spent talking about Bush’s military service (or lack thereof) is another day that the campaigns ignore the issues and the American people get screwed that much harder.

    As long as the people who get hurt the most refuse to vote, we’ll continue to get what we deserve: screwed over by rich people who just don’t give a fuck about anyone but themselves.

  62. Enough of your hate-mongering anti-Bush diatribes! Can’t you people keep your political views out of a typographical forum? Go to a political forum to spew this crap and keep typographica free of it, or drive away people who might, on the off chance have come here for the first time ever — just when typography was a hot topic — for a rational expert discussion on the subject.

  63. Don Leeper says:

    Well folks, we have walked in the vale of typographic ignorance here. I started in the printing business in 1974. In 1982 I started a typesetting company called Alphabet Express with my sister Nancy. Our primary client was the University of Minnesota Press, which had been using Selectric Composers in-house to typeset its books and wanted to switch to phototypesetting. In setting corrections to older books we often had to match Selectric Composer type. In addition, my sister was skilled on the Selectric Composer herself.

    Some observations:

    1. The documents are entirely consistent with setting on a Selectric Composer, including the supposedly anomalous superscript “th”.
    2. There is a curious amount of anachronism in the commentary. Some commentators have said that it wasn’t possible to center a line on a typewriter in the early ’70s; not true. Others have suggested that a secretary wouldn’t have done things like swap typeballs in producing a memo. Hello? Secretaries were doing a lot more complex things than that to put out clean-looking office documents. I think people’s fingers have forgotten what it was like to type a clean document on a typewriter — or perhaps they weren’t around those days.
    3. Selectric Composers were all over the place in the ’70s: student newspapers, any place that produced newsletters, universities, low-end ad shops. To think that a government agency wouldn’t have bought a fancified typewriter like the Selectric Composer this is giving more credit to government agencies than is deserved.
    4. In sum, the documents are completely consistent with documents I knew from the era. The typographic argument for their being forgeries is mistaken (when it isn’t just plain partisan). If they are forgeries they are remarkably subtle ones.
    5. Finally, once again the right has displayed a mastery of diversion and distraction. We’re talking about typographical trivia instead of facts like record government spending (nonmilitary as well as military), record deficits, global climate change, being dependent on an energy resource controlled by despots, and having our military tied down in a theater not germane to our genuine war with militant Islam.

  64. Alastair Keady says:

    anti-Bush diatribes?

    Where did you see these Martin?

  65. Stephen, I suggest you disable further comments on this discussion before it gets into a flame war and discredits typographica. My views are from the right, and I feel like flaming with diatribes of my own to balance the anti-Bush and anti Rep. crap posted here. Don Leeper’s comment above was perfectly valid until point 5, which was nothing but political opinion, so – at least in my opinion – it invalidates all the points he made before. Do all of you want a flame war political “discussion” blog here, or can you lefties see past your agendas?

  66. Alastair Keady says:

    The only ‘diatribe’ I’m seeing is the one of your making Martin.
    At least Don actually addressed the points of teh debate. I fail to see how his points prior to .5 are ‘invalidated’ by his political opinions.

    I’m neither an American citizen, nor a Democrat, but I hold my own opinions as the the merits of Bush as president. I think that the letter is a fake, and pointed out why I hold this belief. What have you contributed to the debate beyond a political hissy fit and a referral of a Bush flagwaving site.

  67. The unspoken and unwritten rule prevailing here seems to be that an anti-Bush opinion thrown into one’s post is the enlightened position and deserves no comment.
    In contrast, by merely pointing out that there’s an inflammatory sentiment there that has no place in a typographical discussion, and that this default anti-Bush postiition is offensive I’m accused of a hissy-fit.
    Be scientific, take off your blinders and get rid of this Pavlov’s dog, automatic leftist agenda.

    (And if you don’t find the offerings at ScrappleFace funny, in spite of his clear, right wing agenda then you have no sense of humor. His work is just as good as the stuff on The Daily Show with Jon Stewart, or more neutral stuff like The Onion, both of which I am a big fan of.)

    See – we’re kindling the flame war now…

  68. Alastair Keady says:

    Um, Martin,

    No-one commented on the political sentiments, because the posts were on topic and had some sort of input to the issue at hand. I’m the ONLY person to comment on your post, and only to point out that you’re not actually contributing to the debate, and that you’re seeing ‘diatribes’ where none actually exist.

    Hissy fit seems an appropriate comment. I’m also in posssesion of a fully functioning sense of humour. That guy just isn’t funny. Sorry.

  69. John Butler says:

    A thread relating to the President turning into a political discussion? Never saw that one coming.

    Posting in a comment thread is a completely voluntary act.

    At least twelve people posting in this thread seem to think Dan Rather’s prized documents were made on something other than Microsoft Word. But not one of them has accepted my wager.

  70. For myself, I just want to know what the damned documents are made on, but I’m not sure why. After the initial excitement of letterform comparisons (such as they were, in their Loch Ness Monster form) and speculation about how much care a secretary may or may not take in setting type on a fancy machine, I can’t quite figure out why I care. It’s not really political (although I’m probably more left wing than the lot of you), it’s more just the resolution of a mystery.

  71. dzd says:

    The entire argument is completely irrelevant until and unless CBS produces the original documents. How convenient that they don’t actually have them.

  72. Don Munsil says:

    For what it’s worth, I’ve just posted a long typography-heavy set of thoughts about these documents on another blog, coming to the conclusion that we still don’t have any definite proof one way or the other.

    Comments on Talent Show Blog

    My comments start about halfway down the page.

    I have to admit that when I first saw these documents I was immediately suspicious because I assumed all typewriters were monospaced. Having read more about the IBM Executive and seen examples of its output, I don’t think it’s quite so clear.

    I am looking forward to someone getting some kind of definitive answer, and it’s certainly enjoyable to see so many people getting interested in the minutia of typography. I don’t think I’ve ever see so many people trying to understand kerning before. :-)

  73. bahamaha says:

    Well I don’t know if the documents are fake but the link
    POSTED BY: Fontboy | Sep 12, 2004 06:04 PM

    Convinced me that the docs are not done by Word. Check out the shape of the numbers. Number 8 especially.
    Any how I think we should remember that the secretaries were trained personel and typing was a valued and in demand skill. Amount of paper produced by government agencies for sure required reliable and top of the line equipment just like it does today.

  74. Steve Merryman says:

    This may shed some light on the topic (copied from another blog as the Dallas Morning News requires excessive registration information):

    Dallas Morning News: Former Secretary Says She Didn’t Type Memos.
    The paper tracked down Jerry Killian’s secretary, Marian Carr Knox, who provides a definitive assessment of the CBS memos:

    “These are not real,” she told The Dallas Morning News after examining copies of the disputed memos for the first time. “They’re not what I typed, and I would have typed them for him.”

    She said the typeface on the documents did not match either of the two typewriters that she used during her time at the Guard. She identified those machines as a mechanical Olympia, which was replaced by an IBM Selectric in the early 1970s.

    She said that the culture of the time was that men didn’t type office-related documents, and she expressed doubt that Lt. Col. Killian would have typed the memos. She said she would typically type his memos from his handwritten notes, which she would then destroy.

    Mrs. Knox, now 86, is a Bush-hater, and she claims to remember “yak-yak” about the Lieutenant similar to what is in the memos. She speculates that the memos “may have been reconstructed from memory by someone who had seen Lt. Col. Killian’s private file, but were not transcriptions because the language and terminology did not match what he would have used”:

    For instance, she said, the use of the words “billets” and a reference to the “administrative officer” of Mr. Bush’s squadron reflect Army terminology rather than the Air National Guard. Some news reports attribute the CBS reports to a former Army National Guard officer who has a longstanding dispute with the Guard and has previously maintained that the president’s record was sanitized.

    Mrs. Knox also cited stylistic differences in the form of the notes, such as the signature on the right side of the document, rather than the left, where she would have put it.

  75. It’s proportional, yes. But it doesn’t even come close to matching the memos in question the way Times/Times New Roman does.

    Thomas Phinney’s analysis on Typophile is the best I’ve seen yet.

  76. There was a short but amusing item about this topic this morning on NPR’s Weekend Edition featuring an interview with Jonathan Hoefler. I particularly enjoyed Scott Simon’s line: “Font making… doesn’t usually get much press–figuratively at least.”

  77. cgrepairman says:

    The documents are TOO POOR QUALITY to do any comparison as to length, kerning or font style. In any court the pontificating that so called experts from CMU do as to “this wasn’t possible, or that wasn’t possible” would be laughed at by any reasonable juror. A competent attorney would just show a similar font and it would be closer in shape and size to New Times Roman than the font in the documents is. Nothing like an “EXPERT” who tries to make a point by shading the facts. Fontboy gave a link to someone who is credible in his background but not credible in his conclusions. The point of view is initially from an electronic printing background rather than an analog printing background. Everything is interpreted as if the Electronic type was the standard and not the other way around. The Digitized font is the copy of the Original design. The Digitized font attempts to follow the older Analog version and therefore will look like the older version. It is a natural progression. Nothing Strange that they look the same.
    I remember when fonts were initially beginning to be digitized and the vectors of the electronic fonts were not acceptable and did not look like the originals. Anyone with a loupe could see it. The Electronic fonts had to be brought up to speed and it took several years to do it. The original crappy fonts from Adobe and Apple improved only after Compugraphic ironed out their technology for digitizing the fonts and settled the legal issues concerning the patents and trade secrets issues surrounding the digitizing of those fonts.

  78. cgrepairman says:

    Dr. Newcomer states that computerized typesetting started in 1969. False, Photon corporation started in the late 50’s and Compugraphic had by the mid 60’s sold to the AP the 2900 series and 4900 series computerized phototypesetters. These machines had no integrated circuits as we now know them. They had digital circuits made of discrete components, diodes, resistors, transistors. These were finished products and with the photon corp. systems before them managed to kill the linotype monopoly on typesetting. By the early 70’s a large number of small systems were in use throughout the printing industry including Small scale integrated circuit driven devices (non-CPU)which had all the abilities Dr. Newcomer says were not available to the military. These devices were in mom and pop printshops across the country. A variety of mini-computer systems followed and well before 1975 the first microprocessor based computerized typesetting system was developed. These systems did NOT HAVE PIXELS. The pixels Dr. Newcomer refers to are artifacts of the scanning process and the original documents were apparently made by strike on type. Thus the high quality of image Dr. Newcomer cannot explain.
    The fact is that it was not only Compugraphic and IBM and Addressograph-Multigraph developing these technologies but a host of office equipment sellers such as Wang.
    To flatly state that a capabilities which existed as early as 1920-1930 (Varityper)could not have been developed and marketed some fifty years later when we had already managed to put a man on the moon is ludicrous.

  79. I think you CAN say that the documents have proportional spacing and any number of differences compared to real memos produced by the commanding officer at the time. This page at The Washington Post (may require subscription. Use if you don’t have one) is the clearest comparison I’ve seen.
    I suppose, in THEORY, these documents could have been produced in 1970, courtesy of the technology described by cgrepairman. But does anyone really believe these are real, when every other aspect points out that they are fake? The only argument in favor of them is this “theoretically technologically feasible” angle.

  80. Matthew says:

    I noticed that the postal code (77027) given on the memo does not match the street address (5000 Longmont). They are near each other, so it is possible that the postal codes were revised. One more thing to investigate.

  81. Jp C says:

    There is no doubt in the world that this was made after the 1980s. Besides the typographic doubts, there is the fact that you know a presidential candidate or anyone else isn’t going to risk making a disputable forgery. It’s just to easy to get a 1970s typewriter.

    It were the Bush campaigners themselves that produced this. Bush2004’s little helpers intended to badly duplicate this document from a real 1970s original.

    The reasons for this are obvious. A real document would never have been questioned, and could jeopardise the Bush campaign. A false document –if found out– makes Bush a military hero and Kerry a very bad falsifier.

  82. John Butler says:

    Excuse me, is this tinfoil hat yours?

  83. Jp C says:

    In fact, I must say on many an American forum I have been called a schizo or a raging paranoid f***k-up. I cannot blame people, but I must mention that half-millimeter-thick tin-foil, radiation above about 20 kHz (i.e., including both AM and FM bands) is actually partially blocked.

    ++ Both Bush & Kerry are pro-war, pro-corporate skulls’n’bones goons, so it really makes no difference who wins, this planet is doomed anyway. Typographers, especially experts, shouldn’t be wasting time on them, as there still is enough ComicSans and look-I-made-it-myself-Word-processing around.

  84. Raven says:

    Martin Archer links to “This page at The Washington Post” — which restates the chestnuts about proportional type and superscripts (“No superscript in official documents” — oh no? See Bush’s “chronological listing of service“, second entry)…

    .. but then that graphic goes further, to actually falsify the shape of the letter “f” in order to say it’s kerned (in “identified”, halfway down on the left).

    Blow up the PDF to 500% and look for yourself at the letter “f”, both in that word and in others (like “before” on the line below).

    The original letter-shape has a short top curve.   The WaPo graphic has elongated it to extend over the following “i”.

    So here we see fakery — proven fakery — not by CBS, but by the Washington Post.

  85. Raven says:

    For instance, [Marian Carr Knox] said, the use of the words “billets” and a reference to the “administrative officer” of Mr. Bush’s squadron reflect Army terminology rather than the Air National Guard.

    Here’s a Google search for billet or billets at (official Air Force sites).

    This gets “about 939” hits, for example:

    “The remaining percentages include billets at the wing level and below, from wing staff on down.” (Officer Career Path Guide, ch. 3-1)

    “Staff billets above the wing level are prevalent in every major Air Force command and numerous joint service agencies (ie, Military Traffic Management Command ….” (Officer Career Path Guide, ch. 4.6)

    “All PD updates, and PDs accompanying requests for core billet realignment or for new core billets, will be submitted for certification.” (AFMC Instruction 38-202)

    So much for the claim that “billet” wasn’t an Air Force term.

    (As for “Air National Guard”, specifically, thirteen of those hits were from webpages.)

  86. John Butler says:

    Fancy a bet, Mr/Ms Raven?

    All this idle speculation and still no takers. The deluded persist, but always with enough lingering doubt not to risk a wager.

  87. Hrant says:

    And of course this juvenile, deluded pouting doesn’t get censored.


  88. John Butler says:

    No, that would be laughing. Even you won’t risk the wager, Hrant, because you know I’m right.

    And as much as you claim to want to talk about “larger issues” and what not, you’re still here in this thread. And gullible people are still desperately insisting that the memos could have been produced in 1972 by the IBM Selectric Executive Jizzmatic Composer with rich Corinthian leather.

  89. Raven is clutching at straws. Rather (heh!) than blow up the pdf to 500 percent, why not look at it at something like 100 percent or even 50 percent. Anyone who regularly works with type can see that this is Times (or TNR). When the fakes have been deliberately degraded so much then giant close-ups are hardly informative at all. Seen from a normal reading distance the typeface is obviously Times.

    Counter claims of fakery done by WaPo are ludicrous and shows you’re in denial about having lost the argument Raven.
    Maybe you’re trying to make the “fake but accurate” point…

  90. Hrant says:

    I’m still reading this thread (because once in a while somebody will say something smart, and because I’m a curious person who can’t help it), but my participation is limited to simply pointing out the retardation of some people. Because, yes, discussing this too much makes the overall situation worse, not better.

    And the main thing I’ve learned about you (after having thought highly of you in Leipzig) is not that you’re right or wrong about a petty, idiotic wager (I’m decent enough not to spend enough time figuring out this underhanded trickery), it’s that you need to become a well-rounded person if you wish to be a part of the solution and not the problem. Your general political attitude not only totally clouds your otherwise sound analytic abilities, it’s a revolting insult to all the suffering, miserable people on our earth. Great shame on you.

    Pay attention: the best way you can rise above your current decrepit stance is to travel; and listen to the other side of the fence. And I don’t mean former East Germany.


  91. I think I speak for almost everyone here when I say, “Oh shutup Hrant.”

  92. Hrant says:

    While I know that each of us speaks only for himself, and pretending otherwise is both misguided and misleading.


  93. Raven says:

    Martin Archer wrote:

    “Raven is clutching at straws. Rather (heh!) than blow up the pdf to 500 percent, why not look at it at something like 100 percent or even 50 percent.”

    So that you can examine the letter-shapes more closely — since at smaller magnifications the screen pixellation can contribute to the “blur” (already bad enough due to multiple-generation scans).

    In particular, so that you can see how the top curve of the “f” is short rather than elongated, and does not extend over the following “i” in “identified”, as the WaPo graphic misrepresents it.

    Comparing the short-curved “f” in the PDF to the the long-curved “f” in the WaPo graphic is enough to prove the WaPo falsified this detail.

    “Anyone who regularly works with type can see that this is Times (or TNR).”

    Or something in that general family.   (Since Times was available on typewriters then, this is not the crushing proof you seem to think it is.)   Some have pointed to a similar typeface, Delegate, as an even better fit.   The resolution is poor enough to leave the argument open, though some noticeable differences in the letter and number shapes from MSWord’s Times New Roman have been detailed, e.g. unequal loop-widths in the numeral 8.

    “Counter claims of fakery done by WaPo are ludicrous ….”

    How so, when that “fakery done by WaPo” has just been shown?

    You and anyone else can look at the PDF here (at whatever magnification shows details best for you), and the WaPo graphic here, and see the difference for yourselves.

  94. Raven says:

    Martin, here’s JuliusBlog’s comparison of the CBS memos to MSWord’s Times New Roman, again at high magnification to show details.

    As Hunter at DailyKos pointed out regarding the supposed “identicality” of MSWord documents to the CBS memos:

    If you shrink each document to be approximately 400-500 pixels across, they do indeed look strikingly similar. But that is because you are compressing the information they contain to 400-500 pixels across. At that size, subtle differences in typeface or letter placement simply cannot be detected; the “pixels” are too big. If you compare the two documents at a larger size, the differences between them are much more striking.

  95. Raven says:

    And again thirteen of those hits are on (Air National Guard webpages).

  96. Like I said; clutching at straws.
    Sure the documents are real Raven. SURE they are.
    Perhaps you’ve never played with a photocopier (or a fax). All sorts of weird stuff happens to degrade the document. Particularly in faxing. In my inexpert opinion this accounts for the irregular baseline, although admittedly not the superscript shift.

    And reluctantly turning away from typography and onto other “forensic” analysis…
    It’s laughable that you bother spending time Googling “billet” when Ms. Knox’s interview with CBS (a CBS in desperation damage control mode on Sept. 20) revealed this:

    “…I know that I didn’t type them,” says Knox. “However, the information in those is correct….”


    “…Knox says the information in the four memos that CBS obtained is very familiar, but she doesn’t believe the memos are authentic…”

    Courage, Raven.

  97. Raven says:

    Martin Archer wrote:

    “… the documents are real Raven.”

    I can’t agree with your certainty.   I think the issue isn’t settled either way, and won’t be, until more information comes out — such as originals of these memos, whose ink (or carbon) and paper can be examined; verifiable provenance (“chain of custody”); or even other memos from other officers’ files that can be matched to these.

    “Ms. Knox’s interview with CBS… revealed this:

    “‘..I know that I didn’t type them,’ says Knox. ‘However, the information in those is correct….’


    No, the statement that she didn’t type them is not equivalent to saying they’re fake, since she was not the only person who could possibly have typed them.

    For instance, as Drudge quoted Killian’s son:

    “If my father was going to type a CYA memo, which he didn’t,” Gary Killian responded. “He would have typed it himself because he wouldn’t have wanted anyone to see it.”

    (Looking at “which he didn’t”, I have to ask:   How would Gary Killian know?   He wasn’t there all the time to see everything his father did.)

    But what I’d like to know is — where are the memos Ms. Knox says she did type, with the same or similar information?   What happened to them?

  98. JP Stormcrow says:

    Have never been to this site before, but I ran across an analysis of the Bush memos posted at Utah State Interactive Media Research Lab and was looking for more knowledgeable folks to take a look at it. The author makes some pretty definitive claims as to font and evidence that they were produced on a typewriter. However, the article seems a bit quirky in style – so despite the apparent imprimatur of an academic lab, I am skeptical of the conclusions.

  99. John Butler says:

    Rollin’ rollin’ rollin’
    Keep those heads a-rollin’

  100. Jeff says:

    Oh, come on. I typed papers in the 1970’s. If you wanted a superscript, you rolled the paper up half a line. I had to type math papers and it was slow, but I’m no typist and I could do it.

Post a Comment

Comments at Typographica are moderated and copyedited, just like newspaper “Letters to the Editor”. Abusive or off-topic comments are not published. We appreciate compliments, but don’t publish them unless they add to the dialog. Thank you!