Culture
Words by Frank Grießhammer
History of the Hershey fonts
Originally created in 1967, the “Hershey Fonts” are among the earliest digital representations of type. Their creator, Dr. Allen Vincent Hershey, was not an ordinary typeface designer – he was employed as a theoretical physicist at the Naval Weapons Laboratory in Dahlgren, VA. Hershey’s usual work was mostly theoretical: he concerned himself with the analysis of submerged ships and their influence the waves flowing above, he calculated how liquids behave when they flow around objects, and analyzed ship hulls.
Originally created in 1967, the “Hershey Fonts” are among the earliest digital representations of type. Their creator, Dr. Allen Vincent Hershey, was not an ordinary typeface designer – he was employed as a theoretical physicist at the Naval Weapons Laboratory in Dahlgren, VA. Hershey’s usual work was mostly theoretical: he concerned himself with the analysis of submerged ships and their influence the waves flowing above, he calculated how liquids behave when they flow around objects, and analyzed ship hulls.

Hershey’s research involved complex calculations, to complete those, he had sophisticated tools at his disposal, such as the NORC (IBM’s “Naval Ordnance Reference Calculator”) – at the time the fastest computer in the world. Like any computer of the era, the NORC would output, on paper, rows and rows of numbers which had to be interpreted and converted to graphs and drawings by hand. The computer of the day was not much more than a calculating machine.
Consequently, reports in which Hershey summarized his projects were typed up with a Varityper, a fairly sophisticated typewriter which could be described as a pre-digital word processor. The Varityper featured proportional, interchangeable fonts similar to the IBM Selectric, but it was still not easy to type mathematical equations with it. For special characters (such as Ω or π), plastic bits called “Typits” had to be appended to the machinery. Completing a report with many equations was a tedious process, to say the least.
While computer output was very limited at first, technical progress brought the introduction of more sophisticated output methods, such as COM (computer output on microfilm) devices. Those new machines enabled computers to produce visible results at a much faster pace; drawing graphs and curves was now possible without manual interpretation (which in turn made a whole lot of draftsmen lose their jobs). A notable example of a COM device is the Stromberg-Carlson S-C 4020. At the heart of the machine is a cathode ray tube called a “Charactron”, capable of exposing the trace of a modulated electron beam on a photographic negative. The resulting negative is exposed on photo-sensitive paper, and thus generates a “printout”. Today, it is hard to imagine how groundbreaking a device the S-C 4020 really was – nothing short of a paradigm shift. This new technology transformed computer output from a huge, labor-intensive undertaking to a commodity.
“Calligraphy for Computers”
Font: Minotaur Bold
The S-C 4020 was, of course, available at Dahlgren, and Hershey embraced the device, using it to create some of the first maps purely based on digital data. He also employed it to draw charts and curves, but soon realized its scope was not limited to that: In 1967, Hershey released the report “Calligraphy for Computers”, in which he described how computers can be used for more than calculations, but how they also help to reduce labor. He presented a suggestion on how to create (formerly typewritten) reports using digital source data, which could then be output on a COM device. The benefit of writing digital reports would be the ability for correcting errors on the fly (something we can gladly take for granted today), and lifting character set limitations implied by the typewriter, or any other fixed method. Since fonts did not exist at the time, Hershey also included examples of letters as they could be drawn with the S-C 4020. His quite extensive paper contains roughly 40 pages of text, which are followed by over 200 pages of letter diagrams, including their computer-readable representations. Hershey’s letters are stroke based, and follow a “connect-the-dots” principle on a fairly coarse grid (up to 21 raster units high). Nevertheless, they are exquisitely designed:
- Different optical sizes serve different purposes: smaller, less detailed letters (9 raster units high) to be used for maps or equations, more detailed letters (13 or 21 raster units) for different roles in text.
- The repertoire includes complete upper- and lowercase alphabets for Roman, Italic, and Script styles, plus Greek and Cyrillic – and even Blackletter (Hershey calls them English, Lombardic and German).
- Mathematical characters as well as cartographic, meteorological, and astrological symbols are part of the repertoire.
- About 2000 Japanese ideographs are provided as a proof of concept.
Despite Hershey’s obvious enthusiasm for character design within the constraints of the S-C 4020 grid, he concludes his report with the sobering realization that the time for page layout via a COM printer has not yet arrived:
“The latest cathode ray printers are almost adequate for the preparation of mathematical reports.” This line probably explains why “Calligraphy for Computers” is still composed on a Varityper.
“The latest cathode ray printers are almost adequate for the preparation of mathematical reports.” This line probably explains why “Calligraphy for Computers” is still composed on a Varityper.
Official portrait of Dr. Allen V. Hershey in NAVY uniform, 1961. Courtesy estate of Dr. A.V. Hershey/Dorothy Donhauser Cogitating in the garden, with daughter Dolly. Courtesy estate of Dr. A.V. Hershey/Dorothy Donhauser “The Laboratory Log”, a Dahlgren publication, mentions Dr. A.V. Hershey’s achievements in the field of mapping. Draft of Japanese Kanji. Draft of swash capitals for Duplex Principal Script. Example output of a map based on Hershey’s digital data. Hershey’s notes. Various of the styles as digitized by Hershey, as well as digital modifications. The effect of stroke thickness on a skeleton font. Viewer application for Hershey’s font data by Earl F. Glynn II, written in Delphi 7. http://www.efg2.com/Lab/OtherProjects/Hershey.htm Hershey Fonts used in a CAD environment. Hershey-designed steam engine. © Dolly Donhauser
As disappointing as this this sentence is, the paper rightfully assumed it would be only a matter of time before technical progress could make digital typesetting a reality: Only two years later, in 1969 (an improved S-C 4060 had been released in the meantime), Hershey writes a follow-up: “Fortran IV Programming for Cartography and Typography”. This report successfully introduces his concept for creating mathematical (and other) papers using a digital system, and comes with many examples for the application of Hershey’s typesetting method and letters. The best part: The whole report is set using the system and letters it describes – there is no better way to prove a point. Hence the conclusion: “The present system can be used for the preparation of mathematical material.” Success!

Dr. A. V. Hershey receiving the Ed Rosse Award.
Allen V. Hershey is an unsung pioneer of digital typeface design. It is likely he did not see himself as a typeface designer, and just did what was necessary to do a good job. Still, he can serve as an example for any type designer today. Why? Hershey was thinking out of the box, creating a product that would be useful for more people than himself. He would not shy away from breaking new ground. It is unlikely, for example, that he was a specialist in blackletter, which he decided to include because of its use in mathematics. Also, Hershey definitely did not speak Japanese, and still decided to learn the system of ideographs because it served the project. Hershey was not discouraged by impediments. Only patience and perseverance would lead him to successfully implement his typesetting method years after the original idea. Last but not least, Hershey would make smart choices in reference material. Most of his letters are informed by existing techniques (such as the Leroy lettering set), or analysis of good source material (Jan Tschichold, Frederic Goudy, or ATF Specimen books).
“Hershey’s letters are stroke based, and follow a “connect-the-dots” principle on a fairly coarse grid”
Font: Minotaur Sans Regular
Computers and computer output have come a long way since Hershey. Nevertheless, his work is significant because of the decisions he made. Of course, his designs may look quaint and naïve today, but the straight-lined construction of Hershey’s letters was the pinnacle of possibilities at the time. An uninformed viewer may think the letters represent only an edgy, stylistic design aesthetic, but it is worth noting how well-planned and well-designed Hershey’s letters are, even in the given constraints.
Without being a revival, Minotaur is a nod to Hershey’s achievements, and the Lombardic caps are a gem for those who know his work a little better. Of course, Minotaur goes beyond Hershey’s original character set and range of weights. Still, Minotaur demonstrates that Hershey’s aesthetic is not limited to the styles he envisioned; “Beef” is a refreshing departure from the text styles, still bearing Hershey’s traits. A good lesson to learn from Minotaur, is how historical inspiration (as weird as it may be) can lead to entirely new and interesting design solutions.
Initially published in 10/2017 for Minotaur type specimen, ISBN 9791093578057.