If you enjoyed a computer interface in your day, you likely have enjoyed or at least tolerated the work of Matthew Carter. He is a typographic legend, whose work has directly appeared in many places throughout modern life? In a world where we only had 10 or so “web-safe fonts,” he was responsible for three of them—Georgia, Verdana, and Tahoma.

But years before he became known for developing key parts of your default computing experience, he took on one of the most important tasks ever presented to a font-forger: He redesigned the font for the phone book.

A sample of typography from a 1959 phone book from Illinois’ Calumet region. The font is Bell Gothic, which was the standard for small copy in phone books for decades. I am pulling from the Yellow Pages, because of the off chance that the five-digit numbers in the White Pages still work. Don’t want to dox someone with a 66-year-old screenshot! (Internet Archive)

AT&T had been using its own custom fonts for decades by the time that Carter got to it. In 1938, the company had commissioned typographer Chauncey H. Griffith to develop a new standardized typeface for its phone books, which it called Bell Gothic. Griffith, earlier in his career, had created a font called Excelsior, well-suited to run at very small sizes in newspapers, a class of typography known as “agate.” Bell Gothic was the phone-book version of that font, and it helped set the vibe for the Bell brand in general.

The problem was, by the 1970s, AT&T had moved to cathode ray typesetting (CRT), a process that allowed more flexibility in layout. Previously, a bold character in a hot-metal Linotype machine had to be the same width as a regular-width character, limiting flexibility. But CRT just did not do any favors to Bell Gothic, which was built to work around the quirks of the metal-type process.

As writer and typographer Nick Sherman noted on his website, CRT did not work particularly well with the agate-sized typography of Bell Gothic:

Bell Gothic worked fine when the directories were still being composed in hot metal on a Linotype machine and printed on a letterpress, but because it was designed for those production methods, it didn’t hold up under the set of limitations presented by newer technologies. Typographic composition was being done photographically with Cathode Ray Typesetting (CRT), and the printing done on high-speed offset lithography presses. These production methods greatly affected the typeface; letterforms (especially in the Light face) broke apart: its strokes became lighter, sometimes eroding completely at the intersections of straight and curved strokes.

For a time, printers tried to compensate for this erosion by over-inking the printing plates; while this helped to thicken the strokes, it brought up a whole new set of problems. Legibility suffered as the already condensed letterforms closed in on themselves. The strokes of different characters ran into each other, making c and l become d; r and n became m; 3 looked like 8; 5 looked like 6. Another problem with over-inking was that the presses had to be stopped frequently for additional cleaning which cost printing time and production money.