Font Wars

Replicating the superimposition

One of the principal arguments for the forgery side is that by simply typing the text of a memo into Word using the default settings, you will get an exact letter for letter match. I tried this and found it to be more difficult than claimed, but feasible. In the interests of journalistic integrity, here's my latest try:

I exported the CBS CYA memo as a Tiff. I did not try to fix any rotation problems.

I went back to the Word memo text. I made sure to put double spaces in the same places this time. This caused the text to format completely differently from the CYA memo. My default margins turn out to be 1.18 in., maybe because I am working in the Spanish-language version of Office, which is based on centimeters, or because I changed the defaults to picas and when they converted to inches the came out 1.18.

I reset the margins to 1.25 and 1 in. The text formatted correctly. I don't know if these are default Word settings. They are standard secretarial correspondence settings, though.

Then I made a PDF of the same thing in Word. I didn't print it out and scan it because I am getting tired of it all. Given the quality of the original scan, it's good enough. I rasterized it in Photoshop at 144 dpi, the same as the CYA memo export. The difference in vertical spacing is a faxing effect, I'm sure, as is the rotation. The two images were drastically different in size. I reduced the Word image to fit the CYA memo measure.

So this is a point for the forgery side.

The only way to solve the puzzle is to see the original documents. Even high-resolution scans are not going to resolve all the questions. A physical test will easily determine if they were printed with a carbon film ribbon or a laser printer. CBS is going to stonewall on that. Therefore the argument falls into the category of phenomenology. There is no physical evidence, just digital images whose provenance can never be determined by visual inspection.

Do not fall in love with any given hypothesis. This is a swamp. Nothing is real. At the moment, CBS holds all the cards. If they have decided to crush Bush, it's Nineteen Eighty-Four. It's not like they don't have the resources to create anything at all that they want or need to do it.

In 1977, Mike Salisbury showed me the originals for Marlon Brando's dossier in Apocalypse Now. They were on screen for seconds. They were absolutely perfect as far as I could tell. They didn't look like stats. They looked like actual documents -- an article from Life magazine really stood out. I think he may have had it printed on a small offset press, but I will have to ask him to be sure.

Given the poor quality of the scans, I don't see how anyone can definitively show a space in front of a th. It's like the kerning. I don't see it. The rest is maybe. You can't have it both ways on the scans. If the inconsistent apostrophes are generation artifacts, then I will argue that it's not valid to compare them with 100% accurate laser output.

My typography guru is convinced that it wasn't done on a Composer. A lot of people who have worked on Composers are claiming that it would have been entirely feasible. They say that Composer font balls with superscripts were available, both on special balls with many other symbols, and also hybrid balls that had standard typographical refinements, including fractions. The Composer had automatic centering. They also say that IBM routinely made custom font balls.

Now that I've actually matched the Press Roman letter forms to the memos, it will take a lot to convince me that they could not have been created on a Composer.

You can't say that Killian wouldn't have known how to have operated a Composer. It's well established that he never used a typewriter. Neither does Bill Clinton. Does that mean that his correspondence is fake?

Now let me drop a name: Chester Anderson, the founder of The Communication Company. I went to the print shop in Mendocino in 1973 looking for someone to type a book-length manuscript. They sent me up to see Chester. I said, "The Chester Anderson?" He responded, "The Jules Siegel?"

I gave him the manuscript and he said he would give me an estimate on typing it. I came back a few days later and he had set half of it in type. He loved the book and just whaled into it. Why set it in type? "It's all the same to me. This is the machine I use for everything."

I later worked on that very Composer in Mendocino in 1975 after Chester quit. That's how I learned to set type. I used Press Roman because it was the closest to real type. I wrote my Playboy story about Thomas Pynchon on it and also another called "A Modern Romance" that appeared in Chic. The owner used to give me letters to do because he liked the way they looked.

You can look at this in two ways:

[1] Jules/Chester = Killian/Operator

[2] Owner/Jules = Killian/Operator

If you can't shoot down the Composer theory on the basis of the physical evidence, such as it is, you have to argue not just that it's unlikely that he had access to one, but show historical evidence that he didn't. That is not going to be a trivial task.

No arguments based on superimpositions are valid, not even mine, because the scans have no evidentiary value. My images are visual aids. They aren't any kind of proof. The scans I am now using were extracted from the PDF with Acrobat 5 as TIFFs. They are beyond fucked.

When my typography guru got very heavy about the mathematical impossibility of creating the memos on a Composer, I wrote an update agreeing that they are probably fakes. It would be a great thing to send out. It wouldn't embarass me at all. It shows my devotion to truth and integrity, exactly my favorite image.

Once I actually fiddled with the Press Roman specimens, I got the very strong feeling that it was looking very possible to recreate these memos on a Composer. Gerry Kaplan came really close. To fully satisfy myself either way I would have to sit down at a Composer and try it myself. I am not going to go on record as convinced that it couldn't be done on a Composer because I am afraid that someone will do it and make me look like a fool.

A statement from IBM that no Composer font balls ever had superscript "th"s would be the strongest evidence that it could not have been used. Show me that and I will heartily agree that the memos could not have been created on a Composer.

Anything else is conjecture. My position is that it's all conjecture. I'm comfortable with that. I'm very uncomfortable taking a definitive stand one way or the other, because either one could blow up in my face.

I think it will dissolve into a he said she said argument. Kitty Kelley is next. That's going to be a lot more interesting than fonts.

We are on Tropical Storm Warning in Cancun right now. Hurricane Ivan is going to brush us tomorrow. It's already begun to rain in fitful torrents and the front door is leaking. Tomorrow is going to be an absolute mess. I hope it stays on the current forecast track. Otherwise we will be in the soup. I just dread the thought of being evacuated from the Hotel Zone.

Font Wars become quite trivial when you look at the satellite images of Hurricane Ivan bearing down on your home, believe me.

Posted by jules_siegel at September 12, 2004 10:12 PM | TrackBack