Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The Parallel Keyboard

  1. Feb 5, 2005 #1
    All current keyboards are sequential. Even when you press more than one key at a time, you must press one of those keys first--shift-e is not the same as e-shift. It strikes me that this is inefficient. We are capable of pressing 10 keys at a time.

    What if a keyboard were made so that many letters could be pressed at the same time (or within a very short interval) where the order doesn't matter, and the computer could figure out what you intended?

    For example: To type the word "chair," I must currently press the c, then the h, then the a, then the i, then the r. I can do this in a fraction of a second, but it's still one at a time. Instead, on a Parallel Keyboard, I might press all the letters at once (the keyboard would be set up to make it easier for me to do this) and the computer would instantly recognize that the only possible arrangement for those 5 letters is "chair." If I had pressed "ploo" (the keyboard would have duplicates of common letters like o and e) the computer might give me the options "loop" (option 1), "pool" (option 2), or "pool" (option 3). I'd press 3, using 2 keystrokes instead of 4.

    It seems like you might easily get 200+ WPM on a keyboard like that, once you learned how to use it.
  2. jcsd
  3. Feb 6, 2005 #2
    Not if you have to decide which word you mean. My friend has that on his cellphone and if anything it takes a lot longer and is annoying. Also typiing at 200wpm assumes you have 200wpm of **** to say. Generally you're gonna wanna be thinking slower than that; typing faster will be pointless because either the quality of your writing will drop (literary problems) or the data you are trying to communicate may be more jumbled as you don't have time to think about what you are saying. In my opinion its just one of those things where the juice isn't worth the squeeze.
  4. Feb 6, 2005 #3


    User Avatar
    Science Advisor
    Gold Member

    Yup, at the moment I can type far faster than I can think. This probably results in 95% of what I post being rubbish.
  5. Feb 6, 2005 #4
    The Twiddler, and other chording keyboards

    If that were the case, then it might be called a chording keyboard.

    Many wearable-computing pioneers use the Twiddler, a one-handed chording keyboard that has only three buttons for each of four fingers (plus several thumb buttons).

    Max speed on the Twiddler (and Twiddler2) is about 60 WPM, though that takes a lot of practice. And though it is a bit slow, the Twiddler makes up for that by the fact that it can be typed on while the typist is doing just about anything else at the same time. I have found that I can walk, run, and even ride my bike while typing on a Twiddler. And wearable-computing researcher Steve Mann reported on the wear-hard list that he can type into his wearable computer with his Twiddler while running down stairs, skipping three at a time.

    The wear-hard archives are stored here, by the way:
  6. Feb 6, 2005 #5
    At first perusal, the chording keyboard has the important difference that while more than 1 KEY may be pressed at a time, only 1 LETTER may be entered at a time. Simultaneous key press notwithstanding, a chording keyboard is still fundamentally a sequential keyboard.

    I think that people commonly talk 200+ WPM.
  7. Feb 6, 2005 #6
  8. Feb 7, 2005 #7
    People who are talking quickly often go 200+. I can type 120 and it is a lot slower than talking. I have tried typing songs before as they are playing and I can only keep up with a fairly slow song.
  9. Feb 7, 2005 #8
    Even if it was possible to press the keys at the same time, I wouldn't be able to use
    that keyboard. Because if the word 'chair' comes up in my mind, I can't possibly
    think of all 5 letters simultaneously.
    Try it at home: Think of the word 'GOLD' and now try pressing all keys simultanously.
    You will put your fingers one after another.
  10. Feb 7, 2005 #9
    It's a matter of learning it. If you learn to put your fingers in a certain manner when you think of the word "gold," that's the way you'll do it, even if simultaneous. I know that I don't think about the letters as I type them anymore; my hands just sort of move. They are "programmed" with how to type most common words and I don't have to think about individual letters unless it's a word I don't often use. When I think "gold" when typing, I don't think 'g,' 'o,' 'l,' 'd.' I just think of a particular way that my hands must move.

    The human brain can work in parallel; the limit is the keyboard.
  11. Feb 7, 2005 #10
    More importantly because you can type at 120wpm under ideal conditions does not mean you can reproduce such when listening to music. There is a difference in the amount of time required between conception of a thought and and its being put to type and the amount of time needed for interpretation and comprehension of lyrics prior to typing them. You should always consider your premises before making an assertion. And never measure indirectly what you can measure directly (count the wpm for a song). I honestly feel that a keyboard of the kind you are describing would be pretty much useless for most people.
    Last edited: Feb 7, 2005
  12. Feb 8, 2005 #11


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Wow great idea! But why limit it to 10 keys? Why should the number of fingers we have limit it. I could come up with a device that hits all 101 keys at the same time... Surely the computer could figure out what I meant. :uhh:
  13. Feb 8, 2005 #12


    User Avatar
    Science Advisor

    I think the limit would be how the english, or any other, language is constructed. No matter how fast I am typing, I am still aware of the sequential nature of the words I am writing.

    How could there possibly be a way for a computer program to extrapolate what a person's intent is when writing? Is there anything even remotely close to doing that out today?
  14. Feb 8, 2005 #13
    You probably mean that if one, for example, sees the word 'GOLD' he knows
    how to set his fingers in a certain position, more like he knows the pattern.
    If would be like playing a C chord with the guitar or the piano.
    Maybe it could be possible for words with few letters. But I doubt that you can
    memorize all the patterns, because there are just too many combinations.

    Another problem would be the following: Suppose you want to type
    the word "choose". How does the computer know if you didn't mean "chose"?

    Another example: MISSISSIPPI, where you have double letters.
    How does the computer know that you mean that word if you only
    type the letters MISP?


    P.S. I know we are criticizing your idea Bartholomew, but please don't
    get discouraged. Always keep up having ideas :approve:
  15. Feb 8, 2005 #14
    Omagadon: I am talking about typing songs that I already know, as they are playing. Even when I type during the instrumental parts I cannot keep up with most songs; and perhaps I do usually type a bit slower than my max, maybe 100 wpm on average (I don't have to try too hard to do that), which is still in the same general range. The speaking speeds for the presidents that hitssquad linked to are for speeches, where it is imperative to speak slowly and clearly. Ordinary conversation is much faster than 100 wpm.

    Integral and Garvin, what makes you think the computer couldn't figure out what you meant? With proper indexing it should be very quick to find the word from the letters that make it up. A Google search takes a fraction of a second, and when you consider how many pages that has to search... a word in a simple English dictionary, where each word is indexed by the letters it contains, could be accessed in virtually zero time. A program to find polo, loop, or pool from oolp would be rather simple.

    Edgardo: Well, if you type the word "choose" on my proposed system, the computer would know it was different from "chose" because it has two "o"'s. But in general if there is more than one possibility for a word, the computer could present multiple options which you would choose from by pressing a key--so some words would require 2 keypresses, one for the letters and one for the choice.

    As I said, there would be duplicates of common letters on the keyboard such as O and E. If the word had duplicates of uncommon letters then you would have to enter it in two parts--the computer guessing the first part of the word, then you enter the second part and it guesses that too.

    I think that you understimate the power of human memory. I _do_ have the patterns for all the common words memorized on the qwerty keyboard, and without consciously trying either. In ancient times people had to remember nearly everything because paper was expensive, and they would remember things like Homer's Odyssey, verse for verse. If we can speak with a 10,000 word vocabulary--each sound learned independently from how it is spelled (often before learning how to spell it)--we can surely do the same thing for a somewhat smaller vocabulary, on a keyboard.
  16. Feb 8, 2005 #15


    User Avatar
    Science Advisor

    The problem would not be finding the words that would contain all of the letters you hit at once. The problem lies in the context of the letters and the sequencing. Combine that with all of the differing dialects/slangs, etc... in the english language alone and I see a lot of headaches and hurdles.

    Suppose you want to type a word that has 20 permutations. According to your theory, the list of applicable words that use all of the letters hit would come up. Then one would have to search through the list and find what was meant and then select it. Does that sound very efficient to you? Saving keystrokes is one thing, but time is a lot more important. The time it takes you to press "oopl" and select "pool" from the list will be a lot longer than me typing "p-o-o-l." I simply don't see any savings or benefit from using this method, especially when more effort and better returns could be gained from technologies such as speech recognition.
  17. Feb 8, 2005 #16
    Well, I don't think there are any words that have 20 English permutations. Most words have only one valid permutation.

    Also, you'd select from the list by typing a 1, 2, 3, etc., not using the mouse. And the most common option from the list would be automatically displayed as a default so that if you just kept typing it would take the default; so for probably 95% of the words, you would not need to do any selecting.

    Anyway, you wouldn't have to read the options for common words because you'd just know that option 1, 2, 3, etc. is the one you want; you'd learn to press "oolp" followed by "2" to say "pool," you'd learn to press "oolp" followed by "3" for "polo," and you'd learn just to press "oolp" for "loop."

    Voice recognition: wonderful. But still a developing technology.
  18. Feb 8, 2005 #17
    I still think you're pretty much just making up numbers unless you're a court reporter. More importantly you don't seem to understand what I am saying is while you are listening to music you are still not focused on typing it doesn't amtter if the song is familiar there is still a lag between thought and action. There has never been one scientist or author who has lamented if only I could have typed twice as fast I'd have been able to complete my work or cure cancer or what have you.
  19. Feb 8, 2005 #18


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    This is what immediately came into my mind as I read this. While on the cellphone it's still entering one letter at a time and trying to guess for me which of the three letters on the key I mean, it's irritating to get to the end of a word and find it's not at all the word I meant and then have to keep going back and scrolling through choices.

    The example given doesn't quite work anyway. In "ploo" how would you type both Os simultaneously? I'd have to think ahead to what letters repeat and which don't before I could type a word? One letter at a time is how I spell, and has nothing to do with the keyboard. I see no reason to learn a new way of spelling. And how does this keyboard handle the mispellers of the world? Multiple permutations of the wrong combination of letters would be even more fun, huh? Plus, nobody has even come up with a spell check that has a comprehensive enough dictionary to handle all the words I use (technical terms are really amusing to run through spell check, BTW), so I don't envision much more rapid success in figuring out all the possible permutations of the characters I choose and it managing to get the right one. Even more to choose from the longer the word is too.

    Though, it'd be fun to watch the two-finger typers try it. :devil:
  20. Feb 9, 2005 #19
    As for the accuracy of my figures, you typed 92 words there. If you can find 5 words from those 92 which have valid permutations and for which you did _not_ use the most common permutation, then you can prove me wrong. I don't think you can do that. All I can find at a casual reading are "still"--"tills" (and you used the more common option), "able"--"bale" (and you used the more common option), "has"-"ash" (and you used the more common option), "are"--"era" (and you used the more common option) "and"--"dna" (and you used the more common option) "lag"--"gal" (ok, that's one where it's not positive whether you used the more common option or not).

    So at a casual reading I only found 1 word out of your 92 where you might have had to select from a list--1 keystroke. My estimate (95%) was that there would be 4. Can you find 4 more and prove me wrong?

    Just trust me when I say that I can type 100 wpm without working up a sweat, even while listening to music. When I'm typing from a manuscript I have to read; arguably more of a distraction than typing words I already know. And what do you say to the fact that those presidents who were speaking 100 wpm were reading speeches, and hence talking deliberately slowly? Why are you so quick to criticize? The site said that average talking speed is "125-150" wpm.

    This would obviously only be for people who need or want to type fast. Secretaries, for example, might get a lot of use out of it, and people might use it for internet chat rooms. Scientists might not care; authors might like it. It wouldn't change anyone's life but it could be nice for a lot of people.

    Yes, it would require more learning, but you had to learn to use QWERTY too.

    You would type both O's by pressing 2 different O keys.

    As for technical words--well, it wouldn't be too hard for the keyboard makers to compile additional technical dictionaries that you could add in. And, even a scientist doesn't type THAT many technical words; the keyboard could have a switch to toggle parallel mode, so if you want to type a technical word you toggle parallel mode, type your word, and toggle it back.

    There wouldn't have to be any support for misspelled words, but it probably could be set up to guess at near-permutations and suggest almost-matches if exact matches are not found.
    Last edited: Feb 9, 2005
  21. Feb 9, 2005 #20
    Do you know what finger independence is, Bartholomew? How many of omagdon7's 92 words can you chord on a qwerty keyboard?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook