Right. My wife did secretarial work for an ad agency and, later, a real estate developer in the late 1950s and early 1960s. When she saw the ABC News snippet about the criticism she snorted with scorn; the IBM machines she worked with could easily have done all that.
I will not say that the Bushies are liars, but....

Posted by: Al Revzin at September 10, 2004 12:03 AM

Did the IBM Selectric I have the capability to do superscript.

Posted by: John Flood at September 10, 2004 08:30 AM

Some comments so far from the Newsroom-l mailing list.

Walt Cunningham wrote (quoting Jon-Erik Prichard>:

"Another aspect of the type on [the August 18, 1973 memo] suggests, perhaps proves, forgery. The type in the document is kerned. Kerning is the typsetter's art of spacing various letters in such a manner that they are 'grouped' for better readability. Word processors do this automatically. No typewriter can physically do this.

"Two good kerning examples in the alleged memo are the word 'my' in the second line where 'm' and 'y' are neatly kerned and also the word 'not' in the fourth line where the 'o' and 't' overlap empty space."

There is no kerning in the Y and A in CYA. If there were kerning it would be global and apply to all the characters. The effects might also be explained by the way the letters distort as a result of scanning in the course of faxing, copying and scanning. I can see similar effects in the IBM ad.

Thomas Phinney wrote:

It sure does look to be Times New Roman actually. Though the quality of the repro is pretty stinky. However, the character widths match exactly. However, the superscripted "th" in smaller size is indeed a major red flag that suggests it's a forgery, and even that it was done with Word.

A typographer commenting to Fark.com claims that he worked on IBM typewriters that had a "th" key. I don't remember that. If he's wrong, this would be the most significant evidence that the document is forged.

I'm also dubious that a National Guard office would have an IBM Composer at the time. It seems highly unlikely, from what I know of the audience and usage of the composer. Even if they did, it still seems unlikely that they would use it for routine memos and correspondence.

The most likely typewriter is the IBM Executive, not the Selectric (which did not have proportional spacing). It was introduced in 1941, but did not become widely available before 1950, when IBM began promoting a new model. By the 1970s, it had been superseded by the Selectric and the Composer (a Selectric that had a memory and could justify text.) See http://www.wordiq.com/definition/IBM_Executive_series_typewriter

Eli Dickinson wrote:

Also, I would like to point out an error in the anti-forgery evidence posted on the Newsroom-L blog. Microsoft Word displays documents slightly differently on the screen than it does when printed (and, in fact, there may even be slight differences depending on your printer and your version of Word). The image on the blog uses MS Word text taken from a screenshot, not printed on a page and scanned -- you're comparing apples and oranges. The most obvious difference is that the superscripts are displayed slightly lower on the screen than on paper.

I've replaced the image with one scanned from a Word print-out. As Thomas Phinney observed, the original memo is a very bad scan, possibly from a copying machine, but I think the differences in the "g" and the "3" are substantial and not easily explained by the scanning problems.

Roger Helbig wrote:

I used an IBM Executive Selectric typewriter extensively in the early-mid 70s while in the Air Force at three different bases, NY, ME, AZ, so that type typewriter was extensively employed in Government applications.

As it turns out, the Selectric apparently did have proportional type, but didn't have variable spacing, so it couldn't do proper justification. The Selectric Composer did. Like the Varityper, it was widely used in letter shops, and might very well have been used at an Air National Guard base to prepare manuals and the like. If so, all of the anomalies can easily be explained, including the superscript "th."

The Washington Post has an example of a "suspect" memo that uses the regular "th" and the superscript "th" in the same document. This wouldn't happen in Word, which automatically changes the "th" to superscript. Note "111th Fighter Interceptor Squadron" in first line.

It would be possible on a typewriter with the "th" key or, especially, a Selectric or a Selectric Composer, which could have the character on a separate "golf ball" or accessible through some other key combination.

You'll also notice their comment about the lack of the base on the "4." It's not significant. If the document were produced on a well-worn impact typewriter, the base of "4" might have worn off. Or, more likely, it was never there in the first place because the typewriter font was modelled on Times Roman.

Posted by: Jules Siegel at September 10, 2004 08:41 AM

I think you guys are reaching here.

1. The leading (i.e., the vertical line spacing) matches the default settings in Microsoft Word. By the way, it looks like 14 point leading (i.e., the distance from baseline to baseline is 14 points) to me. Typewriters were capable of 12 point (single space) leading, 18 point (1 1/2 space) and 24 point leading. No typewriter was capable of 14 points of leading.

2. Before desktop publishing arrived in the mid-1980s, typesetting systems and high end typewriters would *not* have supercripted the "th" and "st". They would have been "superior" not "superscript". The difference is that "superior" characters are lined up with the top of the capital letters, while "superscript" shifts the character *above* the top of the character. When I was doing typesetting in the late 1980s, only Quark XPress did "superior" characters. And I don't think that Word does them today. You can see a *superior* "th" on one of the typewritten documents that are not in dispute. But no "superscript". The only way to do superscript on a typewriter would be to manually shift the roller up 1/2 space, then type the "th". The superscript "th" in the memo is *exactly* like Word's superscript automatic format. By the way, the "th" didn't superscript whenever the author put a space between the number and the "th". This, too, is how Word handles it.

3. The differences in character shape that you identify seem minor to me and could be the result of multiple generations of photocopies and scans. Indeed, similar "clogging" can be seen on the other letters (which do appear to be identical to the modern Times New Roman font).

4. If you start Word and type the paragraphs using the default settings, the line breaks occur *exactly* where they do in these 1970s memos. So do the letter spaces. Seems an amazing cooincidence, doesn't it?

5. The IBM Executive font that you reference is a very different font than Times Roman. So it's clear that the memos weren't typed on what would have been a *very* expensive machine.

6. "Times New Roman" -- the default font in Word and the font apparently used in these memos -- was designed by Monotype for the Times newspaper. I do not think that it was licensed to any typewriter manufacturer.

For the memos to be genuine, they would have had to have been typed on a typewriter that had the Times New Roman font (in 12 point size), been capable of 14 point leading, capable of doing superscript (rather than superior) "th". The writer would also have used the exact same margins as Word's defaults -- .75" top and bottom, 1" right and left.

As someone who has considerable experience with typography both before and after desktop publishing, the memos appear to be particularly inept fakes. I wouldn't hug this one too hard.

Posted by: Ty at September 10, 2004 10:32 AM

Wouldn't it be really cool if you could find the and example of the superscript "th" being used on previously released documents. Like on this one.

Line 2 clearly has the small "th" in question. This can again be seen here on page 3 (pdf) of the same document by increasing the magnification.

Posted by: Robert McClelland at September 10, 2004 10:37 AM

Forget the line breaks. That's where anybody would break the lines. People break em when the dinger dings!

Instead look at where the last letter of each line hits, first on the original then on the ms word document.

They don't line up. Not even close.

On the original, they line up. It's a TYPEWRITER.

On the second one, they DON'T line up. It's a computer.

Posted by: Bruce at September 10, 2004 12:26 PM

Ty wrote:

6. "Times New Roman" -- the default font in Word and the font apparently used in these memos -- was designed by Monotype for the Times newspaper. I do not think that it was licensed to any typewriter manufacturer.

IBM worked with Monotype on the development of fonts for its output devices. Press Roman, used on the Selectric, is the analog of Times Roman. See Bitstream Font Analogue

Typeface
Identification

Table of Font Analogs

Times Roman®

TmsRmn; TR; Varitimes; Claritas; Dutch 801; English; English 49; English Times; Euro Times; London Roman; Pegasus; Press Roman; Sonoran Serif; Tempora; Tiempo; Timeless; Times New Roman

For the memos to be genuine, they would have had to have been typed on a typewriter that had the Times New Roman font (in 12 point size), been capable of 14 point leading, capable of doing superscript (rather than superior) "th". The writer would also have used the exact same margins as Word's defaults -- .75" top and bottom, 1" right and left.

The Selectric Composer was capable of all of that and more. To my eye, the type in the IBM Executive ad looks indistinguishable from the Killian Memo.

As someone who has considerable experience with typography both before and after desktop publishing, the memos appear to be particularly inept fakes. I wouldn't hug this one too hard.

Could be. We'll see how it works out. It is possible that the documents were forged. If so, maybe some Democrats learned a lot from the Swift Boat Veterans for Truth. I don't approve, but it does make me smile.

Posted by: Jules Siegel at September 10, 2004 03:07 PM

I'm sorry, but the lines *do* line up. Every single line in the 8-18-72 memo lines up exactly the same as its 2004 equivalent. Try it if you don't believe me.

For this to happen on every line of a document, you'd have to be using the same font. Font "metrics" -- the width of individual characters, their height, etc., -- are unique to each font.

The IBM Executive typewriter that is referenced on this site uses a completely different font than Times New Roman. Indeed, a real simple test would be to type any two lines from the memos in Word with Times New Roman. Note the relative positions of several letters on the lines. Then see if they line up the same way on the memos.

For instance, the first 2 lines of the text in the 8-18-73 memo is:

1. Strandt has obviously pressured Hodges more about Bush. I’m having trouble running

interference and doing my job. Harris gave me a message today from Grp regarding

On the memo, the H in "Hodges" vertically lines up with the stem of the first "r" in "Harris" on the second line.

And whaddaya know, they line up *identically* when typed in Word as well. Every single letter in the Word document lines up identically to its memo counterpart.

In other words, the memo's font *has* to be Times New Roman.

And Times New Roman did not exist on any typewriters in the 1970s. Indeed, Times New Roman has only existed as a font since the early 1990s when Monotype created it. There was an earlier font called simply "Times", but it has different metrics than Times New Roman, and, I believe some differences.

So it seems clear to me that these documents were not typed on a typewriter in the 1970s. However...let's see if CBS has the guts to submit the *originals* to any experts to be sure. Anyone care to place a bet on that?

Anyone with typesetting experience knows that these documents were typed on a modern computer and not a 1970s typewriter. In my humble opinion, the anti-Bush folks are making a bad mistake by defending such an inept forgery.

Posted by: Ty at September 10, 2004 03:08 PM

We're probably past the point where it matters much what any individual thinks about these documents. The media is all over this, and I predict that the general consensus will quickly be that these were forged. The match between MS Word and the memos is so close that any disinterested party who holds one up to the other is going to believe the memo was forged.

It may be that this is a PHENOMENAL coincidence, or it may be that Karl Rove is an evil genius... but I think the man in the street is going to start asking, "What did Kerry know, and when did he know it"?

Given his sagging poll numbers, he doesn't need this blowing up in his face for the next week or two. Any momentum he had is gone. Splat!

Posted by: Scott W. Somerville at September 10, 2004 03:46 PM

I have typesetting experience as well. I used to use old Linotype machines, as well as photomechanical machines.

Look at the ends of the lines in the original.

The letter g at the end of regarding, and the letter g at the end of rating both line up. Those both line up right under the letter n in running, and right OVER the letter n in Austin.

Now look at those same lines in the ms word version:

The two letter g's don't line up. They fall between the two n's in running. They fall between the n and the space in Austin.

It's as plain as day.

The spacing is far more variable in the MS word copy.

Times New Roman was invented in 1931 by Stanley Morison. The name was later changed to Times Roman by Linotype.

And it was available on IBM typewriters.

"Anyone with typesetting experience..." Really Ty, I hope your typesetting experience predates Quark. I used lynotype hot lead machines. I know a little bit, too.

Posted by: Bruce at September 10, 2004 03:55 PM

These documents were done on a typewriter. Just look at the vertical character drift. As anyone who has ever used a typewriter can tell you, inaccuracies always occur where some characters do not line up vertically with the rest. Note in this case how the 8 in the date is slightly above the 1 and the A. Also note how the first e in "Memo to file" is again shifted slightly upward. Only a typewriter makes these kinds of inaccuracies. In the computer generated document, the characters are all perfectly aligned on the vertical plane. You can draw a straight line underneath them and not a single one would be out of place. Not so with the original document. The same can be seen on every one of the other documents.

Posted by: Robert McClelland at September 10, 2004 05:00 PM

I would like Bruce to be correct about the position of "regarding," "rating" and "Austin." They are the result of missing spaces after periods. I fixed "regarding" in the scan but didn't do a good enough job. I'll print the corrected memo out again and rescan it.

Sorry for the error. My position remains the same. The main issues are the inconsistency of the "th"s and the apostrophes (some curly, some straight) in the various known-authentic documents and the difference in some of the letter forms. Everything else has been discounted by now as not provable or erroneous.

We'll see what CBS has to say about it this evening. If it's a fake, Bush certainly owes Kerry at least one after the Swift Boat hoax.

Politically speaking:

[1] How about the rest of the revelations released by the military?

[2] Does it really help Bush to have this issue hammered in the media when all it does is call attention to his National Guard lapses?

Posted by: Jules Siegel at September 10, 2004 05:15 PM

Look at the last letter in the second line. It does not line up in the same place with the first line in both versions.

Posted by: Alexander at September 10, 2004 05:20 PM

IBM consumer/business typewriters were able to to superscript by the late 60s -early 70s. Witness the IBM "Selectric II," circa 1970.

http://en.wikipedia.org/wiki/IBM_Selectric_typewriter

"The Selectric II had a lever (above the right platen knob) that would allow the platen to be turned freely but return to the same vertical line whereas the Selectric I did not. This feature permitted the insertion of subscripts and superscripts."

More reading:

http://www.etypewriters.com/history.htm

Posted by: Dr. Winston O'Boogie at September 10, 2004 05:41 PM

Let's blow the B.S. from the document experts out of the water. This is from ABC News. Let's debunk the so-called expert, line by line.

Posted by: Dr. Winston O'Boogie at September 10, 2004 05:57 PM