QWERTY Wasn't Designed to Slow You Down

Most people who own a keyboard have heard the story: Christopher Latham Sholes deliberately scattered common letter pairs to slow typists down so the type bars of his early machine wouldn't jam. It's a tidy explanation. It is also probably wrong.

The historian of technology Koichi Yasuoka, working at Kyoto University, went back to Sholes' patents and the Remington archives in a 2011 paper. He pointed out that QWERTY does not actually separate the most common English letter pairs — TH and ER, for instance, are both adjacent. If anti-jamming were the goal, those would be the first to split.

What the early machines did do was sell, in their first years, mostly to a single market: telegraph operators transcribing incoming Morse code in real time. Yasuoka's argument is that the QWERTY layout was iterated to suit that workflow. Common Morse-confusion clusters — the pairs an operator would frequently need to flip between when they realised they had decoded the wrong letter — were placed close together. "Z" and "S", for example, both appear with the dit-dit-dit-dit and dit-dit-dit pattern; QWERTY puts them within easy reach.

The layout was finalised around 1878 on the Remington No. 2. By the time touch typing — invented in the 1880s by Frank McGurrin and others — became dominant, millions of machines used the layout, and the mechanical original constraints had been forgotten.

Dvorak's faster alternative came in 1936. Studies have repeatedly failed to show a decisive speed gap in trained typists. The lock-in is real. The original reason for the lock-in is just not the one everybody quotes.