The Bush Guard Memos

Filed Under: Inspiration , Article , Voice , typography

What started over two weeks ago as a scandal over President George W. Bush’s Texas Air National Guard service record apparently ended earlier this week as a scandal over Dan Rather, the reporter on the Wednesday edition of CBS’ 60 Minutes who broke the original “story”. On Monday, September 20, Rather ruefully apologized for having based his report on forged documents. The fate of his future at CBS, as well as that of his producer Mary Mapes, is now being hotly debated.

Rather’s report was based on photocopies of newly discovered memos from the personal files of Lt. Col. Jerry B. Killian, Mr. Bush’s squadron commander. The memos indicated that Mr. Bush had failed to take a required physical examination and that his performance rating was “sugarcoated” by Lt. Col. Killian under pressure from “someone upstairs”. But, almost from the beginning, the documents were attacked as forgeries, principally on the basis of their typography.

Various document experts—many hired by supporters of the President—quickly concluded that the documents bore all the hallmarks of having been composed on a personal computer using Microsoft’s Word program: the typeface (font) was Times Roman, the spacing was proportional, and a superscript th was employed. Just as quickly it was pointed out by others—including those critical of the President—that there were typewriters available during Mr. Bush’s period of service in the early 1970s that were capable of such typographic subtleties.

The machines that attention focused on were the IBM Selectric Typewriter and its successor, the IBM Selectric Composer.

The former, introduced in 1961 with Model 72, was the first typewriter to employ the revolutionary “golf-ball” typehead. The golf-ball was a partial sphere with 88 characters arranged on its surface that rotated as it printed. It obviated the need for the moveable carriage or typebar of a traditional typewriter. Furthermore, it could be removed and replaced with another golf-ball, allowing typists to change font styles and sizes within the same document. And it could be used in other IBM Selectric models. Original fonts were designed by IBM for the golf-ball, among them Courier and Letter Gothic.

The IBM Selectric Composer was, according to the company, “a new kind of printing machine, much like a typewriter.” IBM’s goal was to create a machine that functioned in the area between a traditional typewriter and a composing machine (e.g. a Linotype or Monotype), one that could turn out documents that closely approached the typographic quality of traditional letterpress printing. What that meant to IBM was that the Selectric Composer could offer true typefaces (as opposed to monospaced fonts), a range of type sizes, proportional spacing, vertical line spacing (leading), and justification.

IBM had devised proportional spacing as early as 1944 (though it was not marketed until 1946) and it had introduced automatic justification (as opposed to justification created through retyping) with its IBM Magnetic Tape Selectric Composer in 1964. (When IBM created the latter machine, which had a rudimentary memory, they coined the term “word processor” to explain how it differed from a typewriter.) While the IBM Selectric Typewriter had two type sizes (pica and elite) and line spacing in increments of 1, 1.5 and 2 lines, the IBM Selectric Composer offered fonts in sizes from 7 to 12 pt. with “leading” from solid (0 pt.) to “normal” (2 pts.). The latter also had a variable spacebar.

But most importantly, the IBM Selectric Composer was the first “typewriter” that had “classical type styles”. The first metal typeface to be adapted to the golf-ball was Times Roman, which IBM renamed Press Roman. It was so successful, in the opinion of IBM engineer G.A. Holt, that it was quickly followed by adaptations of Bodoni, Aldine (Bembo), Pyramid (a square serif design) and Univers. All were available in roman, italic, medium, medium italic and bold variants. Furthermore, IBM created custom versions of Press Roman and several of the other fonts for four broad language groups (Domestic [which included American and English], Nordic, Germanic and Latin); and Greek, Technical and Mathematical fonts.

Most of this information comes from the January 1968 issue of the IBM Journal of Research and Development (vol. 12, no. 1) consisting of a series of articles (all written between December 1966 and April 1967) devoted to the creation of the IBM Selectric Composer. I believe that it is more reliable than much of the information that has appeared in newspaper articles or on the many websites (from those specializing in typewriters to political blogs) in the past week. An apparently critical point regarding the authenticity of the Killian memos is the existence of a th superscript. The IBM journal articles do not address the availability of superscripts (other than numerals used in mathematical operations), but several articles and websites (e.g. have quoted former IBM employees as indicating that characters such as the superscript th were available as part of custom fonts. (What no one has mentioned—perhaps because it would have been unlikely for Lt. Col. Killian to have such skill—is the possibility of creating such superscripts using a combination of the machine’s vertical spacing and its range of point sizes.)

Several websites (e.g. contended that the IBM Selectric Composer would not have been used by Lt. Col. Killian because it was too complicated for office use. Yet, IBM had specifically developed the Composer for office use. In the January 1968 issue of the IBM Journal of Research and Development Adrian Frutiger, typographic consultant to IBM (and designer of the Composer version of Univers), likening the IBM Selectric Composer to the medieval scribe, claimed that the machine allowed authors to write their own books “without the assistance of specialists?. Composition, once again, becomes extremely simply and direct.” And IBM engineer G.A. Holt, in the lead article, called the IBM Selectric Composer “an operator-oriented machine.” Clearly IBM was marketing it in the same way as the personal computer was in the early 1980s.

While everyone was investigating the minutiae of typewriter fonts the Killian story took a new twist on September 14 when Marian Carr Knox, Lt. Col. Killian’s secretary, declared in an interview with Rather on the Wednesday edition of 60 Minutes that the memos were fake—but that the views they expressed were accurate. Her conclusion was based on the language and style of the memos—a point that some ex-military critics had already jumped on. Rather used her interview to tenaciously defend his original report, clinging to the truth of its substance even as he began grudgingly to admit the possibility that the memos were suspect.

On September 20, in anticipation of Rather’s public apology, the New York Times reported that CBS was about to acknowledge that the Killian memos were fakes because unnamed experts had asserted they “could have been produced only by a modern-day word processor, not Vietnam War-era typewriters.” The ill-informed, unnamed experts were presumably Linda James and Emily Will who had been separately quoted to the same effect by the New York Times the previous Friday. However, neither James nor Will had provided any details to bolster their claims, nor had they identified the specific word processor that they believed had been used to forge the memos. And all of this more than a week after it had become clear that indeed there were Vietnam War-era typewriters that could produce sophisticated typography using “real” typefaces.

However, Rather’s reluctant acknowledgement that the Killian memos were forgeries made no mention of typographic anomalies. Instead, it came only after he had flown to Texas over the weekend to re-interview Bill Burkett, the former officer in the Texas Air National Guard who had originally provided the photocopied documents to Mapes. In that interview—a small portion of which was broadcast by WCBS following Rather’s public apology—Burkett, a longtime critic of Mr. Bush, admitted the documents were forgeries. Yet, he denied being the forger and, furthermore, refused to identify the person or persons who had given him the documents in the first place. William Safire, the former speechwriter for Richard Nixon and current columnist for the New York Times, has rightly called for an investigation into who created the forged documents, but most media accounts of the affair have preferred to focus on its effect on the careers of Rather and Mapes and the bottom line at CBS and its parent, Viacom.

The Bush scandal may have morphed into the Rather scandal, but there are still questions that have not satisfactorily been answered about the Killian documents. In pursuing their authenticity a number of bloggers, accepting the possibility that they could have been done on an IBM Selectric Composer, turned to arguing over whether or not the Texas Air National Guard would have been able to afford the IBM Selectric Composer, a machine that some claimed was far too expensive. But this line of inquiry is futile. After all, in the 1980s it was revealed by Congress that the Pentagon had purchased $600 toilet seats and other seemingly over-priced items. In her interview Ms. Knox made no mention of the kind of typewriter that she used as secretary to Lt. Col. Killian—and, curiously, Rather did not ask. But, the only way to know for sure if the Texas Air National Guard had IBM Selectric Composers would be to locate procurement records or contemporary news accounts about IBM’s military contracts. No one seems to have done this.

Another way to assess the possibility that Lt. Col. Killian had access to an IBM Selectric Composer would be to examine other documents produced by him or others at the Texas Air National Guard during that period. This should have been an obvious and simple path to pursue, but, once again, no one seems to have tried. (I am basing these conclusions solely on the newspaper reports I have read in the New York Times, various network and cable newscasts and a sampling of weblogs.)

The whole line of inquiry regarding typewriter fonts and typography has been completely speculative. The documents that CBS was basing its story on were not originals but photocopies. (Furthermore, as Emily Will revealed in the September 24 New York Times, the experts who originally vouched opinions on the documents—both for and against their authenticity—did so on the basis of faxed versions of the photocopies!) While some blogs tried to recreate the degradation letterforms undergo when being photocopied, no one specified whether or not they were using an older analog copier or a newer digital one. Photocopies on the latter are crisper and closer to originals and, if sufficiently enlarged, they will show bitmapping. Of course, without having access to CBS’ photocopies, deciding whether or not they were contemporaneous with Mr. Bush’s service or recent was another exercise in speculation. Because CBS did not release photographs of the complete documents it was impossible for anyone, expert or not, to truly analyze them.

Throughout this topsy-turvy fracas it has been sobering to find a typeface at the center of an election dispute. The Killian memos shattered Beatrice Warde’s famed crystal goblet and it is Dan Rather, not George Bush, who has been wounded by the flying glass.

About the Author: Paul Shaw is a calligrapher and typographer working in New York City. In his 20 professional years as a lettering designer, he has created custom lettering and logos for many leading companies, including Avon, Lord & Taylor, Rolex, Clairol and Esté Lauder. Paul has taught calligraphy and typography at New York's Parsons School of Design for over 10 years and conducted workshops in New York and Italy. His work has been exhibited throughout the United States and Europe.