Xref: utzoo comp.std.misc:181 comp.periphs:1975 Path: utzoo!attcan!uunet!lll-winken!ames!amdahl!pyramid!ctnews!mitisft!jamie From: jamie@mitisft.Convergent.COM (Jamie Riotto) Newsgroups: comp.std.misc,comp.periphs Subject: Re: Chord Keyboards Message-ID: <812@mitisft.Convergent.COM> Date: 28 Jul 89 19:24:30 GMT Organization: Convergent Technologies, San Jose, CA Lines: 93 I have been reading the previous discussions about chord keyboards with great interest. In a sense, the term chord keyboard has been misused in this group. There have been a number of "one-handed keyboards" proposed and built. Only two that I know of qualify as "chord" keyboards. A chord is defined as pressing multiple keys simultaneously which produce a STRING of characters. This is very different from the MicroWriter keyboard or the ASCII Ball keyboard which require pressing multiple keys to get ONE character. The stenographer keyboard mentioned in an earlier article has the operater press multiple keys simultaneously to produce shorthand characters which represent a phonetic symbol. This meant that a word takes as many strokes as there are phonetic syllables. Since this number is less than the number of letters in a word, this keyboard has been called a chord keyboard. In the mid seventies, work went on at IBM to develop a one-handed keyboard they called the "Chord Keyboard" (patented and trademarked!). This keyboard was a remarkable step forward in one-handed typing. The keyboard was made up of a five-by-two array of square keys operated by the fingers and a row of four rectangular keys operated by the thumb. The finger keys had rounded depressions called dimples. Each square key had a dimple in the center, half a dimple on each side, and a quarter of a dimple on each corner. Thus a dimple was formed at the intersection of two side by side keys, as well as the intersection of four keys arranged in a square. Each type of dimple produced different tactile feedback as pressing a center dimple involved pressing one key, pressing a side dimple involved pressing two keys and pressing a corner dimple involved pressing four keys. The innovation of overlaying dimples at the intersection of keys meant that far fewer keys are necessary and the "travel distance" of the finger is much shorter. The thumb keys had "troughs" instead of dimples, and one can press either one thumb key (center trough) or two adjacent thumb keys (side troughs). There is a finger dimple for every letter and number. The thumb troughs controlled UPPERCASE, UPPER-LOW, SPECIAL (!#$%* etc.), SPACE, REVERSE and others. Since finger keys are used for different meanings depending on which dimple is pressed, some possible combinations of letters are not available in a single chord. IBM did a great deal of research into digrams and trigrams (two and three character combinations) encountered in legal, technical and general literature. The keyboard layout was arrived at by running massive amounts of text through layout simulators which measured average number of characters per chord and finger distance travelled, amoung other things. They also did studies on which combination of finger motion is most reliable in humans (ie which "chords" on a five key keyboard could be pressed the fastest with the most accuracy by untrained individuals). This was also fed into the simulator so only "favorable" chords were allowed. For example, most people find that pressing thier middle finger and pinky at the same time is difficult if the fourth finger isn't pressed as well. The final layout consisted of common constenants on the left and vowels on the right dimples. For example the characters "the", "tho", "ing", and "er", are all one chord. This is what makes the chord keyboard so efficient. One particularly innovative feature was in the use of the thumb keys to modify finger chords. The most common character in the english language is the space character. This was assigned to a thumb key so that it could be typed in conjunction with any finger chord. Thus, the characters " the", is also one chord. Pressing UPPER with "the" produced "THE". Pressing UPPER-LOW with "the" produced "The". And pressing the trough between SPACE and UPPER-LOW along with "the" produced " The"! REVERSE would change "le" to "el". A typical sentenced would be typed: /In/ thi/s/ se/g/me/nt/ of/ te/xt/ the/ cho/rd/ bo/und/ar/i/es/ ha/ve/ be/en/ s/ho/wn/ by/ li/ne/s/. I trained on the chord keyboard for about three months in 1978 and was able to type between 45 and 50 words a minute. I used the chord keyboard for the rest of the year to do all my memos, reports and software input. I found that parsing sentences into chords became a very natural exercise. Interestingly, most people learned to type with their "weak" hand (ie right-handers used their left hand to type). This meant your normal writing hand was free to jot notes or turn pages! Since many of the chords that could be pressed made no sense, (pressing REVERSE and a single character for example), many chords were available for special things. A list of dozens of common english words were available through special chords like "this that them for have had could would" etc. In fact, the entire kanji charater set can be mapped to the chord keyboard! References: "The Chord Keyboard" - Nathaniel Rochester and Frank Bequaert December 1978 IEEE COMPUTER "Chord Keyboard Architecture" - Nathaniel Rochester and Frank Bequaert 1978 IBM Poughkeepsie Laboratory Tech. Report "One-handed Chord Keyboard And Its Control Means" - Frank Bequaert and Nathaniel Rochester, US PATENT 4,042,777 issued Aug 16, 1977. Cheers, Jamie Riotto UNISYS (408) 435-3785 convergent.com!risky!jamie