From: Eric Fischer Newsgroups: alt.folklore.computers Subject: Early versions of the QWERTY keyboard Date: 13 Oct 2000 04:01:21 GMT Organization: EnterAct Corp Turbo-Elite News Server Lines: 191 Message-ID: <8s61eh$2uvj$1@news.enteract.com> NNTP-Posting-Host: shell-3.enteract.com X-Trace: news.enteract.com 971409681 97267 207.229.143.42 (13 Oct 2000 04:01:21 GMT) X-Complaints-To: abuse@enteract.com NNTP-Posting-Date: 13 Oct 2000 04:01:21 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Originator: enf@enteract.com (Eric Fischer) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.nextra.ch!news1.sunrise.ch!news.imp.ch!uni-erlangen.de!newsfeed.germany.net!newsfeed01.sul.t-online.de!newsfeed00.sul.t-online.de!t-online.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.enteract.com!news.enteract.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:66027 Several days ago in this newsgroup I posted about Richard Dickerson's theory that the QWERTY keyboard was optimized to put the most frequently used characters as far away as possible from one another in the circular type basket used on the early Sholes and Glidden typewriters. I have since repeated his calculations for several early versions of the QWERTY keyboard. The first Sholes/Glidden/Soule patent features a two-row keyboard in alphabetical order: - 3 5 7 9 N O P Q R S T U V W X Y Z 2 4 6 8 . A B C D E F G H I J K L M Plotting the frequency of use of various digraphs in a sample piece of English text against the number of typebars separating each of the two corresponding typebars in a circular type basket shows a fairly random distribution, as would be expected (though with some common digraphs conveniently far apart): 18000 | | 16000 | th | 14000 | | in | he er 12000 | at | te es 10000 | on an re | st is 8000 | ro ou me en | or to nd se 6000 | ed nt it ma co ss ge il ha al si ar ng | de ec be ve ne io of 4000 | ce fi ea ag el ai as ct ll hi ac ur ch ul na ri sg et fr pe 2000 ee ef tr ad ie id ci un ld iv yo ex am lo ay wa ho sc bb ab eg ps ei fa mu ib bj cl bl mb bn ap hr do cu av au 0 aa--bc--bd--cf-ae--af--bh--ah--dl--aj-ak--bm--ao--az--aq--ax-aw--bs--bt 0 2 4 6 8 10 12 14 16 18 The earliest known version of the QWERTY keyboard appears on the cover of the August 10, 1872 issue of Scientific American: ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ | | | | | | | | | | | | | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | _ | , | ' | |___|___|___|___|___|___|___|___|___|___|___| | | | | | | | | | | | | Q | W | E | . | T | Y | I | U | O | - | |___|___|___|___|___|___|___|___|___|___| | | | | | | | | | | | | A | S | D | F | G | H | J | K | L | M | _|___|___|___|___|___|___|___|___|___|___|_ | | | | | | | | | | | | | & | Z | C | X | V | B | N | ? | ; | R | P | |___|___|___|___|___|___|___|___|___|___|___| Repeating the frequency analysis shows for this keyboard a fairly linear relationship between the most frequently paired digraphs and their separation in the type basket, supporting Dickerson's theory that it was designed that way: 18000 | | 16000 | th | 14000 | | in | he er 12000 | at | te es 10000 | an on re | st is 8000 | ou ro me en | ornd to se 6000 | it om ed ma co ss ng hail ge al ar le | io ecde ve be of ne 4000 | as ea ag ce ctai ms em icel fi ll prac et sgch ot gr rifr no rs pe hi di 2000 ee mp wepl ad mo tulo ie im wi ex ay owsy la amid iv if bb rm bnbj eq nf eyav ei ab irck pt cl bodt bu au ibap by du 0 aa-az-bvaq-cv-aw-afae-br-bm-ahfw-aj-dr-akcm-cp-dy-biao-bt-ek----------- 0 5 10 15 20 25 There is a photo that supposedly represents the 1873 state of the machine. It does seem possible that this represents an intermediate stage between the 1872 version and the later keyboard: ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ | | | | | | | | | | | | | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | - | , | _ | |___|___|___|___|___|___|___|___|___|___|___|_ | | | | | | | | | | | | (S key is missing | Q | W | E | . | T | Y | I | U | O | P |???| from photo; key |___|___|___|___|___|___|___|___|___|___|___| marked ??? has no | | | | | | | | | | | label) | Z | S | D | F | G | H | J | K | L | M | _|___|___|___|___|___|___|___|___|___|___|_ | | | | | | | | | | | | | A | X | & | C | V | B | N | ? | ; | R | ' | |___|___|___|___|___|___|___|___|___|___|___| In any case, its frequency distribution shows only minor differences from the above. The next documented stage in the evolution of QWERTY is the version from the 1878 Sholes patent (applied for in 1875): ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ | | | | | | | | | | | | | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | - | , | _ | |___|___|___|___|___|___|___|___|___|___|___|_ | | | | | | | | | | | | ("|" represents the | Q | W | E | R | T | Y | U | I | O | P | : | triple-dot symbol, __|___|___|___|___|___|___|___|___|___|___|___| its meaning unknown | | | | | | | | | | | | to me, found on | | | A | S | D | F | G | H | J | K | L | M | monocase Sholes |___|___|___|___|___|___|___|___|___|___|___|_ keyboards) | | | | | | | | | | | | | & | Z | C | X | V | B | N | ? | ; | . | ' | |___|___|___|___|___|___|___|___|___|___|___| This keyboard seems to have been the one used on all Sholes and Glidden machines manufactured before Remington took over (and perhaps also on the early Remington versions), and seems actually to be a bit of a step backward from the above. Perhaps the move of R to its current position was on aesthetic rather than functional grounds, because while it improved the gap between O and R, it made the more important E and R too close. 18000 | | 16000 | th | 14000 | | in | er he 12000 | at | te es 10000 | re an on | st is 8000 | ou ro meen | nd to or 6000 | it ed ma co ss ng ar al us le | io liec de ve be ofne 4000 | as ag ce ct ms ai el ic ll ac et sgch ot ns rana pr no pe fr gr di 2000 ee sc op ad mo im ie ld ex ayni la un amiv if ci id bb bnbj eq nf eyav ir ab eick jo cl epbo au ib gibu do by ap 0 aa-az-bvbf-aq-ax-afbs-ae-bm-ahjz-aj-bp-akcm-fp-bi-dpbr-ao-bt-ek-------- 0 5 10 15 20 25 The modern QWERTY arrangment, which switched C and X and moved M to the bottom row, made no significant changes to the distribution. All of the above have assumed a circular type basket. But upstrike typewriters went out of favor after the "visible" typewriter was invented, and these front-strike machines have the typebars arranged along an arc instead of a circle. The frequency distribution for the QWERTY keyboard and this typebar arrangement is distressing: 18000 | | 16000 | th | 14000 | | in | er he 12000 | at | es te 10000 | re on an | st is 8000 | ou en mero | se nd to or 6000 |ed om nt it co ma ss ilg ge ar usha si le al |de ecioli ve be ne of 4000 | ceasct ea fiagc em ms el ai llfr gr hietnacrasg taot nsdi na pr pe 2000 eeex plopefc admp honiv if ci ayiee ld amosw la pa bbby mubnbbjdteqegffueboavduireab ei clauu - iep ps ap 0 aaaqzawbvdaxaebcbbdcqafhbshpbbpdmgpahkjzfajcp--daklw-aowpzp------------ 0 5 10 15 20 25 30 35 40 All of the optimizations have turned to pessimizations, and the most frequently used digraphs are closer together than the rarer ones! On a front strike machine, QWERTY doesn't discourage typebar clashing; if anything, it encourages it. eric ###### From: "Philo" Newsgroups: alt.folklore.computers References: <8s61eh$2uvj$1@news.enteract.com> Subject: Re: Early versions of the QWERTY keyboard Date: Fri, 13 Oct 2000 23:07:43 -0500 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Lines: 33 Message-ID: <39e7dd3b$0$174$3936f1fe@news-core.twtelecom.net> NNTP-Posting-Host: 207.250.193.79 X-Trace: 971496764 news-core.twtelecom.net 174 207.250.193.79 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!news-stu1.dfn.de!news-koe1.dfn.de!RRZ.Uni-Koeln.DE!news.netcologne.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.twtelecom.net!news.twtelecom.net!news.twtelecom.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:66104 since i never did put away my 1939 l.c.smith corona...i decided to try to get the machine to jam. it will never jam typing the most common words such as "and" and "the" etc... however it's pretty easy to get jams by typing odd combinations of letters (such as "gyd") anyway...back when the machines were built they really had to rely quite a bit on emprical results... and the qwerty machines just worked! i know that as a kid i always wondered why the keyboards were laid out that way...and was told...it allowed for the typists to maximize speed...but the qwerty keyboards actually reduced the speed..and therefore reduced the jam-ups. btw: i have one of IBM's first electric's (the non-selectric type) It has a dial (1-10) for carbon copies. with a single sheet of paper in it...and set on 10..the zero or "o" will punch a hole in the paper! (good thing i've never punched a hole in my crt screen while printing numerous copies) -- Philo website: www.plazaearth.com/philo ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Early versions of the QWERTY keyboard Date: Sat, 14 Oct 00 07:39:11 GMT Organization: UltraNet Communications, Inc. Lines: 41 Message-ID: <8s9dok$jkd$3@bob.news.rcn.net> References: <8s61eh$2uvj$1@news.enteract.com> <39e7dd3b$0$174$3936f1fe@news-core.twtelecom.net> X-Trace: b4+cdFQvGrS3Nu0NCDJ87vIk6DWCg+V/EYVJ+tBDo+o= X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 14 Oct 2000 10:49:56 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeeds.belnet.be!news.belnet.be!feed2.onemain.com!feed1.onemain.com!xfer13.netnews.com!netnews.com!feed1.news.rcn.net!rcn!207-172-245-242 Xref: chonsp.franklin.ch alt.folklore.computers:66135 In article <39e7dd3b$0$174$3936f1fe@news-core.twtelecom.net>, "Philo" wrote: >since i never did put away my 1939 l.c.smith corona...i decided to try to >get the machine to jam. > >it will never jam typing the most common words such as "and" and "the" >etc... >however it's pretty easy to get jams by typing odd combinations of letters >(such as "gyd") >anyway...back when the machines were built they really had to rely quite a >bit on emprical results... >and the qwerty machines just worked! > >i know that as a kid i always wondered why the keyboards >were laid out that >way...and was told...it allowed for the typists to >maximize speed...but the >qwerty keyboards actually reduced the speed..and therefore reduced the >jam-ups. Sigh! What is it with you guys that you can't understand about production rates. The typing "speed" was defined as a rate-- number of correct words/minute. And I mean _CORRECT WORDS_. The way one's typing speed was calculated was to subtract each typo from the number of words. Typing speed measurements were harsh. If the typist had to spend 20 seconds of that minute unjamming the typewriter, the rate went down. If the typist could type without jamming the keys, that meant that the typist had 20 seconds more/minute to put characters on the paper. Remember, it's not the number of Hertz in your computer, it's what you get done that counts. /BAH Subtract a hundred and four for e-mail. ###### Message-ID: <39E887B2.D872A8E8@plazaearth.com> Date: Sat, 14 Oct 2000 11:20:02 -0500 From: philo X-Mailer: Mozilla 4.7 [en] (X11; I; Linux 2.2.14-15mdk i586) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: Early versions of the QWERTY keyboard References: <8s61eh$2uvj$1@news.enteract.com> <39e7dd3b$0$174$3936f1fe@news-core.twtelecom.net> <8s9dok$jkd$3@bob.news.rcn.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 13 NNTP-Posting-Host: 207.250.193.107 X-Trace: 971540710 news-core.twtelecom.net 173 207.250.193.107 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.nextra.ch!news1.sunrise.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.twtelecom.net!news.twtelecom.net!news.twtelecom.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:66123 a good observation... but in my case i never gained enough skill as a typist to even have much of a chance of jamming a keyboard no matter how poor the mechanical layout. and now that i've switched over somewhat to the computer...there are other problems of course... such as hitting a fuction key and while i think i'm typing...i'm activating drop-down windows etc :) ! Philo