A machine, in 1868, that kept jamming
In the late 1860s, a Milwaukee printer named Christopher Latham Sholes put together the first commercially viable typewriter. At first, his keys were arranged in alphabetical order, A, B, C, D, the way you’d expect. It was elegant, it was logical, and it didn’t work.
The trouble came from the mechanics. Each key flicked a small metal arm forward, which struck the paper through an inked ribbon. When two adjacent keys were pressed in quick succession, the two arms crossed in mid-flight and jammed. The typist would have to stop, untangle them by hand, and resume. On a machine where alphabetically neighbouring letters (therefore often neighbouring in words too) were also mechanically next to each other, the problem was constant.
QWERTY is a compromise, not an optimum
Sholes spent several years moving keys around. The QWERTY layout came out of this slow tinkering and was locked in around 1873–1874, when his machine was bought by E. Remington & Sons. Researchers Koichi and Motoko Yasuoka, who reconstructed the story from the patents, have shown that the logic wasn’t only to space out problematic letter pairs. It was also to gather certain combinations useful to Morse-code telegraphists, who were among the first heavy users of typewriters in newsrooms, transcribing wire dispatches in real time.
The romantic story that QWERTY was designed to slow typists down is therefore only half true. The layout wasn’t built against speed: it was built against a very specific mechanical collision, in a very specific machine. And it was built with as much arbitrariness as logic.
Why nothing has changed since
The mechanics of crossing arms vanished more than a century ago, replaced by spinning balls, then electric keyboards, then virtual ones. None of those technologies has any reason to prefer QWERTY. And yet today’s keyboard is identical to the one frozen in 1873.
What hardens isn’t the machine — it’s the fingers.
In 1936, the psychologist August Dvorak proposed a more efficient layout, putting vowels under the strongest fingers and minimising hand travel. Measured in lab conditions, the Dvorak layout roughly halves the distance fingers cover. It never broke through: not because it’s bad (it’s better), but because learning a new keyboard costs weeks of productivity, all the keys have to be reprinted, and any individual user would only switch if every other user switched too.
The French AZERTY is just a local adaptation of QWERTY from the early 20th century, motivated by French letter frequencies and a few typographic habits. It isn’t optimal either. It has no reason to move either.
The keyboard under your fingers is, in this respect, one of the finest examples of an object frozen by its users. The machine it was built for is dead. It has survived, almost untouched, because we can no longer agree on what the next version would look like.
