"Calligraphy for Computers"

The primary criterion for a choice between character designs is based on what looks best. Attempts to apply mathematical rules have not been entirely adequate. The ultimate criterion certainly is subjective and is an aspect of gestalt psychology. (27)

Contents

1 - Introduction

"Calligraphy for Computers" [Hershey 1967] describes two distinct efforts at character digitization. One, the earlier, involved digitization using plotted dots. Dr. Hershey's bibliographic note indicates that this system was described in "Subroutines for the NORC CRT Printer" [Hershey 1960]. "Calligraphy for Computers" presents "an improved version" of this digitization, with both plots and coordinate data. The other effort, "currently under preparation" (2), is the vector digitization that has become known as the Hershey repertory. The report presents vector plots of both the occidental and the oriental vector repertories, but no coordinate data. It does not indicate the identity of special characters for the occidental repertory, but does present a lexicon of Japanese which uses the characters of the oriental repertory.

These digitizations were designed for a type of output device with which I am unfamilar: the cathode-ray (CRT) printer.1 Dr. Hershey describes these printers in this way: "characters are displayed on the face of a cathode ray tue and are photographed by a camera." (3) He notes four devices used: Stromberg-Carlson Corporation SC-4010 and SC-4020 CRT Printers, described as "dot plotters," an SC-4060 CRT Printer, described as a "vector plotter," and a Linotron (Mergenthaler Linotype Corporation) of unspecified model and, it would seem from the brief notes in this report, different technology. The report investigates physical issues of printing on these devices, including speed, the effects of visual acuity, diffraction, and so forth. It is interesting to note that the smallest dot size achievable on the CRT printers then available had just reached one raster unit for the vector printer (SC-4060). For the earlier dot printers (SC-4010, SC-4020) it was 2.9 and 2.3 times the raster unit, respectively. This means that the vector fonts were designed with a relatively thick pen. Dr. Hershey also notes that the "diameter [of the vector plotting dot] must be no less than one raster unit in order that solid areas may be swept out." (9)

Beyond such physical/hardware issues, Dr. Hershey investigates the issues involved in occidental and oriental character design by his method of polygonalization at finite resolutions.

2 - Computer Hardware

Dr. Hershey indicates that his work was conducted on two computers, NORC and STRETCH. He does not describe these machines further2, but these are well-known machines in the early history of computing.

NORC was the Naval Ordnance Research Calculator, a one-off computer built by IBM that was, from its delivery in 1954 through "about 1963" [da Cruz] the most powerful computer in the world. It was decommissioned in 1968, only a year after the publication of Dr. Hershey's "Calligraphy for Computers" [Hershey 1967]. See:

da Cruz, Frank "The IBM Naval Ordnance Research Calculator." http://www.columbia.edu/acis/history/norc.html

STRETCH was the IBM 7030 Data Processing System, built in quite limited numbers. It was also said to be the world's fastest computer at the time of its delivery. (The first system was delivered, to the Los Alamos Scientific Laboratory, in 1961. I do not know when the Dahlgren system was delivered or when it was decommissioned. Other STRETCH machines were decomissioned in the 1970s.) See:

Smith, Eric. "IBM Stretch (aka IBM 7030 Data Processing System)." http://www.brouhaha.com/~eric/retrocomputing/ibm/stretch/

These were both extremely fast, extremely important machines at a time when computers themselves were relatively rare. While these computers were not new in 1967 when Dr. Hershey published "Calligraphy for Computers," it would appear that he was doing typographic work on them since at least 1960 (when he published "Subroutines for the NORC CRT Printer" [Hershey 1960], and NORC was the most powerful computer in the world.) That Dr. Hershey was able to obtain machine time on such computers to do typesetting of any sort, let alone to explore "the virtuosity of the cathode ray printer ... with calligraphic digitalizations" (1), is altogether remarkable.

Notes

1 Curiously, I found recently a reference to this technology in Oscar Ogg's study of the history of the alphabet, The 26 Letters [Ogg 1971].

2 He refers to NORC as "obsolete" in "Terrestrial and Celestial Cartography" [Hershey 1979], 25 years after it was built.

Exploring Dr. Hershey's Typography
CircuitousRoot