Two letters on your keyboard moved in 1882 to dodge a patent

You have almost certainly been told why the letters on your keyboard are in that order. The story is short enough to survive any retelling: early typewriters jammed when two nearby typebars were struck in quick succession, so the inventor deliberately scattered the common letters to slow typists down, and we have been stuck with the sabotage ever since.

It is a good story. It has a villain, a mechanism, and an irony — progress frozen by a fix for a problem that stopped existing. What it does not have is a source.

The thing that is missing

The 1870s were not a quiet decade for paper. Christopher Latham Sholes and his collaborators filed patents. Their backer wrote letters. E. Remington and Sons kept manufacturing records, and the machine was advertised, demonstrated, reviewed and sold. A claim about why a commercial product was designed the way it was is exactly the kind of claim that leaves a trail — somebody explains it to an investor, or a court, or a customer.

Nobody has produced that trail. (Corrected after publication — see the note at the foot of this page. I originally wrote that Koichi and Motoko Yasuoka had searched and found no such document. They did not say that. They call the story “nonsense” and argue against it from the machine’s development history, which is a different and weaker thing, and the difference was the whole title.) What is documented is the pedigree of the story itself: it can be traced back through Gould, David and Hoffer in the 1980s, and no further.

Absence of evidence is a weak weapon in general, but it is a strong one here, because of what the evidence we do have looks like.

What the paper actually shows

The first Sholes machine did not have QWERTY. It had a keyboard like a small piano — two rows, the letters running in something close to alphabetical order. That is the design of somebody who has not yet thought hard about which letters go where.

What followed was not a redesign. It was a series of them. Between that alphabetical piano and the machine Remington shipped in 1874, the arrangement moved through several intermediate states, none of which is the final one. The layout you are typing on is a sediment, not a decision — and the last layers were put down inside a rifle factory, by mechanics adapting somebody else’s prototype for manufacture, years after the original inventor had stopped being the one making the choices.

You can see the seams from your chair. E and R sit next to each other on the top row. “ER” is one of the most common letter pairs in written English. If the governing rule had been keep frequent neighbours apart so the bars do not clash, that pair is the first thing the rule would have forbidden, and it is sitting right there in the middle of the row, where it has been for a hundred and fifty years.

Why the wrong story is the one that survived

I think the slowdown story wins because it is the only one shaped like an explanation. It names an agent, a motive and an outcome. The true history — a committee, a decade, a factory, some arguments about telegraphy, and a commercial accident hardening into a standard — is a process, and processes are much harder to carry in one sentence.

That asymmetry is not specific to keyboards. It is the same reason “they lost the recipe” beats “the demand went away”, and the same reason a phrase said about one man’s instrument can slide onto another man’s and harden into a number. A story with an author beats a story with a cause, every time, in the retelling — and retelling is the only mechanism by which most of what we know reaches us.

The honest position on your keyboard is smaller and stranger than the myth: nobody in particular chose this, several people each chose part of it for reasons that were not about you, and the thing settled. Then, once enough hands had learned it, the reason stopped mattering at all. That is how most standards work. It is just unusual to have one sitting under your fingers all day, so you can check.

Correction, appended about twenty minutes after publishing

I published this with the title “Nobody has found the document that says QWERTY was built to slow you down”, and a fact-checking agent I had running against the primary sources came back and took the title out.

The absence claim is not one my source makes. Koichi Yasuoka and Motoko Yasuoka (ZINBUN 42, 2011) call the slow-the-typist story “nonsense” and argue against it — from who the first buyers were (Morse operators, who cannot be slowed below the sender), from the typebars of E and R sitting 16° apart in the basket, from published 1870s fingerings. That is inference from the machine’s development history, not a search of the archive coming back empty. I turned an argument into an absence, put the absence in the headline, and it was the most striking sentence in the piece. That is the tenth time in fifteen essays the title has been the wrong part.

Three more corrections, all against me:

1. The window is 1868–1882, not 1868–1873. QWERTY as you type it arrived when Wyckoff, Seamans & Benedict altered the Remington Standard No. 2 to evade patents assigned to the Type Writer Company — M moved next to N, C exchanged with X. The last edit to your keyboard was a legal manoeuvre. That is the new title.
2. “E and R are adjacent” is the wrong adjacency. Jamming is a typebar phenomenon. Yasuoka’s actual measurement is that the ER/RE typebars sat 16° apart in the basket while TH/HT sat opposite each other at 180°. Key adjacency on the top row is a side effect of the same fact, not the evidence. Every use of “adjacent” here needed a subject and did not have one.
3. A 2025 paper argues the opposite of my frame and I had not read it. Craig Galbraith and Neil Kay (Business History, 2025), working from the Densmore correspondence in Milwaukee, hold that anti-jamming was a real, deliberate and successful design goal — via an “infrequency principle” (where two typebars must be neighbours, pick a pair that is a rare digram in English) kept as a trade secret. Not slowing the typist; engineering around the clash. They also report that in hundreds of letters about the machine, Sholes and Densmore barely mention the keyboard at all.

What survives: the layout is sediment, it took fourteen years, several hands made it, and the popular one-sentence explanation with a villain is not in any period document I or my checker could reach. What does not survive is my claiming somebody proved that.