The prehistory of QWERTY and that myth debunked
March 26, 2025 2:23 PM Subscribe
Most of us have heard how the QWERTY keyboard layout was devised to slow down typists so that early typewriters didn't jam if the typist was too fast. A Japanese paper exploring the evolution of the modern keyboard argues that this was not the case. The original use was in telegraphy and is nonsensical that a Morse receiver should be slowed down so as to be less able to keep up with the sender.
(Sections 1, 3 and 4 give you the substance of the paper if you don't want to read the detailed history in section 2.)
The layout evolved over time incorporating multiple compromises between inventors, producers and patent holders. Initially, the intended users were telegraphers and shorthanders who had different habits to modern typists.
The "keyboard" of the first Type-Writer did in fact resemble a piano keyboard having two rows of keys, the top row containing A - N running left to right, the bottom containing O - Z running right to left; this resembled a device already in use at the telegraph college which purchased the Type-Writer.
(Sections 1, 3 and 4 give you the substance of the paper if you don't want to read the detailed history in section 2.)
The layout evolved over time incorporating multiple compromises between inventors, producers and patent holders. Initially, the intended users were telegraphers and shorthanders who had different habits to modern typists.
The "keyboard" of the first Type-Writer did in fact resemble a piano keyboard having two rows of keys, the top row containing A - N running left to right, the bottom containing O - Z running right to left; this resembled a device already in use at the telegraph college which purchased the Type-Writer.
I always thought it was so that people who don't care about password efficacy could more easily type 'asdf'.
posted by BigHeartedGuy at 2:54 PM on March 26 [2 favorites]
posted by BigHeartedGuy at 2:54 PM on March 26 [2 favorites]
A continuing reminder that the better the story, the more likely the Liberty Valence effect is in play.
posted by drewbage1847 at 3:30 PM on March 26
posted by drewbage1847 at 3:30 PM on March 26
Damnit, and here I've thought all these years that Samuel Clemons suffered to give us all that he did. (A lot.)
Well, I'm going to keep on loving him anyways -- he's worth loving.
posted by dancestoblue at 4:51 PM on March 26
Well, I'm going to keep on loving him anyways -- he's worth loving.
posted by dancestoblue at 4:51 PM on March 26
nonsensical that a Morse receiver should be slowed down so as to be less able to keep up with the sender.
I'm not here to argue with the substance of an article which I haven't read but this analogydoesn't make sense for several reasons. Given that we're talking about a keybed to Morse code mechano-electric coupling, it's super possible that the electronics of the 19th century couldn't spit out morse at the same rate as a person could type. Further, unless we also have completely automated receivers, it would be reasonable to slow down the typewriter-speed channel input if the decoding was a more laborious process - which it appears to be.
The analogy seems to be treating the sender-reciever as simply the operator-morse device interaction when obviously it's not. Maybe they address this in the article.
posted by DeepSeaHaggis at 5:59 PM on March 26 [2 favorites]
I'm not here to argue with the substance of an article which I haven't read but this analogydoesn't make sense for several reasons. Given that we're talking about a keybed to Morse code mechano-electric coupling, it's super possible that the electronics of the 19th century couldn't spit out morse at the same rate as a person could type. Further, unless we also have completely automated receivers, it would be reasonable to slow down the typewriter-speed channel input if the decoding was a more laborious process - which it appears to be.
The analogy seems to be treating the sender-reciever as simply the operator-morse device interaction when obviously it's not. Maybe they address this in the article.
posted by DeepSeaHaggis at 5:59 PM on March 26 [2 favorites]
Given that we're talking about a keybed to Morse code mechano-electric coupling, it's super possible that the electronics of the 19th century couldn't spit out morse at the same rate as a person could type
The paper talks about human telegraph operators using proto-QWERTY typewriters to transcribe Morse code on the receiving end; it doesn’t discuss any devices that use a keyboard for transmitting Morse code. In its description of the earliest such case, it even specifies that a telegraph key is used on the sending end.
posted by mbrubeck at 6:23 PM on March 26 [5 favorites]
The paper talks about human telegraph operators using proto-QWERTY typewriters to transcribe Morse code on the receiving end; it doesn’t discuss any devices that use a keyboard for transmitting Morse code. In its description of the earliest such case, it even specifies that a telegraph key is used on the sending end.
posted by mbrubeck at 6:23 PM on March 26 [5 favorites]
INTERESTING. THANK YOU FOR COMPENSATING FOR MY LAZINESS.
posted by DeepSeaHaggis at 7:11 PM on March 26 [3 favorites]
posted by DeepSeaHaggis at 7:11 PM on March 26 [3 favorites]
Apologies, I was distracted while preparing this post and missed something out.
I saw this paper mentioned on BlueSky in a post which hat-tipped this ycombinator thread, which discusses it as only ycombinator can; this is the paper's third appearance on ycombinator but only the first time it has attracted attention.
posted by epo at 3:17 AM on March 27
I saw this paper mentioned on BlueSky in a post which hat-tipped this ycombinator thread, which discusses it as only ycombinator can; this is the paper's third appearance on ycombinator but only the first time it has attracted attention.
posted by epo at 3:17 AM on March 27
Among the various design considerations, it's probably not an accident that the word "typewriter" can be typed using only the top letter row, making it easier for salespeople to demonstrate.
posted by Brachinus at 5:18 AM on March 27 [3 favorites]
posted by Brachinus at 5:18 AM on March 27 [3 favorites]
Clearly it was the result of early gamers who ensured WASD was arranged JUST SO as early as 1873
posted by caution live frogs at 11:27 AM on March 27 [1 favorite]
posted by caution live frogs at 11:27 AM on March 27 [1 favorite]
And the nineteenth-century vi users, who had a vested interest in placing HJKL on the right-hand side.
posted by mbrubeck at 12:09 PM on March 27
posted by mbrubeck at 12:09 PM on March 27
close - QWERTY arrangement was not meant to slow down the typist, it was meant to minimize jamming by arranging keys so that typewriters would jam less frequently.
at least that is the way I interpreted the history as it was presented to me. (about 4 decades ago when I was learning to type on an Eagle II and had experience purposely hitting typewriter keys as fast as possible so they would jam spectacularly)
posted by gkr at 12:10 PM on March 27 [2 favorites]
at least that is the way I interpreted the history as it was presented to me. (about 4 decades ago when I was learning to type on an Eagle II and had experience purposely hitting typewriter keys as fast as possible so they would jam spectacularly)
posted by gkr at 12:10 PM on March 27 [2 favorites]
QWERTY arrangement was not meant to slow down the typist, it was meant to minimize jamming by arranging keys so that typewriters would jam less frequently
Yeah that’s the “corrected” version of the story that I’ve most often heard.
posted by atoxyl at 12:27 PM on March 27 [1 favorite]
Yeah that’s the “corrected” version of the story that I’ve most often heard.
posted by atoxyl at 12:27 PM on March 27 [1 favorite]
I'm too lazy to RTFA, so can someone tell me - is the point not to spread frequently used letters out such that you're typing about equally with both hands and all fingers? Seems like that would actually help typing speed, since fingers on the one hand can be moving into position while fingers on the other are still returning to the home row, or whatever. Also seems more ergonomic, and less likely to cause jams on old typewriters.
Imagine if A, E, and I were all on the left side and the only letters in reach of your right pinky were X, Y, and Z. Your left hand would be doing all the work.
posted by OnceUponATime at 1:56 PM on March 27
Imagine if A, E, and I were all on the left side and the only letters in reach of your right pinky were X, Y, and Z. Your left hand would be doing all the work.
posted by OnceUponATime at 1:56 PM on March 27
« Older Legends, Lattes, and Lament | Well sometimes I go out by myself Newer »
But seriously this is super interesting, for a long time I've known that the slowing down typists thing was a urban legend, but seeing all the different layouts together is cool. Reminds me of the "letter A through history" starting with a bull's head.
I accurately typed 'etaoin shrdlu' into google to get the link
posted by AzraelBrown at 2:33 PM on March 26 [2 favorites]