Jean-Do Bauby wrote 'The Diving Bell And The Butterfly' by blinking at the appropriate letter as they were recited from the most common to the rarest in the French language, because the only voluntary movement he could make after a stroke was to blink one eyelid. I have been told, and i seriously hope this is not true, that blinking in this way, or making some other voluntary movement if another muscle can be moved, is standard practice for communication in this situation. I am also sceptical about blinking being his only voluntary action because it sounds to me that he could also move his eye. If that had not been possible, he would have been blind because vision depends on eye movements.

There is a very clear solution to this problem. I have done a search and can't find a reference to anything similar, so i'm posting it here.

Suppose someone has locked-in syndrome, where they are cognitively intact but can only move one voluntary muscle. Arrange the letters of in this case English in three grids. Combine some letters, namely I and J and U and V, as they were before the eighteenth century. Establish a convention where X is replaced by KS or Z, QU by KW, C is used to spell the sound customarily spelt CH in English and replace C itself by K and S as appropriate. This frees up some space for the most common words of more than two letters. Omit the articles and "is"; they aren't needed in an emergency. Then produce the following grids:

Grid one:

E, T, R
O, A, D
C, U/V, L

Grid two:

I/J, N, M
H, S, F
P, Y, W

Grid three:

B, G, and
Z, K, you
that, for, with

Then, blink (or whatever) as follows:

First blink: which grid. Second blink: X-coordinate. Third blink: Y-coordinate.

There are no spaces and no punctuation, but the former can be produced by pauses and the latter by spelling them out when ambiguity needs to be avoided. It would be possible to communicate each of these letters or words with a maximum of nine blinks, and the speed of communication could be increased further through such techniques as the use of text speech.

Construct similar tables for each language based on frequency. For Austonesian languages, communication could be sufficiently fast to mean that the patient would be motivated to learn, for example, Samoan or Hawaiian simply to communicate faster than in an Indoeuropean language.

If the patient has more movement than just one muscle, they can use the other muscle movements to indicate more than one coordinate at a time. For instance, if they can move one eye freely, they can indicate a letter by moving their eye to the appropriate ninth of their visual field and blinking the requisite number of times.

I considered Morse code and ASCII but this would involve more learning and coordination than such people might be able to manage. This can be acquired much more swiftly, though clearly learning Samoan or Hawaiian would take longer for many people if they spoke, say, Spanish, Russian or English. The same applies to Blissymbols.

This system could be used by people who are disabled in various ways, by people who have been restrained to communicate with each other, by prisoners tapping on pipes and by S&M people who have been restrained for recreational purposes. OK, a couple of frivolous examples there, but basically i am serious about this.

Somebody please tell me this is baked or that something better is, because otherwise there are a load of people out there who are needlessly imprisoned by their disabilities.
-- nineteenthly, Apr 07 2008

About Eye Tracking http://www.technolo...com/Infotech/18254/
[Amos Kito, Apr 07 2008]

Interface Research http://arstechnica..../20060713-7262.html
[Amos Kito, Apr 07 2008]

Other Means http://www.boingboi...tally-paralyze.html
[Amos Kito, Apr 07 2008]

For people like myself who have no idea what an "Etaion shrdlu" is: http://designorati....y-etaoin-shrdlu.php
[Amos Kito, Apr 07 2008]

Guardian on Claude Mendibil http://film.guardia.../0,,2247450,00.html
The ghostwriter who took dictation for "The Diving Bell and the Butterfly" [jutta, Apr 07 2008]

Prisoners of War Tap Code http://www.pbs.org/...feature/sf_tap.html
The 5x5 grid MikeD mentions. [jutta, Apr 07 2008]

Stephen Hawking uses something similarish. He selected words from a long list, by repeatedly bisecting it (A though Z; A through M; A through G; etc until he reaches, say, E; then a similar process to select a specific word from a list of common ones).
-- MaxwellBuchanan, Apr 07 2008


Thanks, [Maxwell]. I've often wondered about Hawking's system but have never bothered to look into it. He's clearly been using it a number of years because i know he's attached to the voice being associated with him and doesn't want it updated for that reason. I remember in the 'seventies the difficulty he used to have communicating, presumably because IT hadn't reached the appropriate stage. I wonder why his method isn't more widely used.

[Amos Kito], thanks for the links. Those are clearly good, particularly the salivary pH thing which is mind-blowing. I did have a rather silly thought about penis use at one point, but couldn't think of something less prurient. However, i seem to remember the problem with using eye-tracking was that people would look in the direction they weren't supposed to simply because they knew they weren't. The other thing is, this works very well in situations where access to that technology is available, but that isn't always so. For instance, a victim of a medical emergency such as a building collapse or RTA would need a less technologically elaborate channel of communication which might need to work very fast. It also doesn't account for the fringe uses i mentioned. Useful though: thanks.
-- nineteenthly, Apr 07 2008


//a building collapse//

Shirley we can’t use any of this for instant communication. “…Now on to Grid 3, the the X-Y coordinates are thus…”
The persons involved would have to already know the code.
-- Amos Kito, Apr 07 2008


Two thoughts:
1. Abbreviations of long words (e.g. writing just the consonants) might be faster.

2. But then, say you just write the consonants, the E-T-A-O-I-N-S-... sequence of frequency will change completely.

3. Three thoughts. There are Three thoughts.

4. There should be choices for "I'm tired, let me sleep please", "call nurse", and other illness related things

5. Four thoughts. There are Four thoughts. No, Five.
-- phundug, Apr 07 2008


Could Morse code be used?
-- phundug, Apr 07 2008


OK, [phundug], it's a question of learning curves. Morse code would probably lead to more efficient communication if you didn't have to learn it, but consider the poor person who's lying there unable to communicate and going quietly mad and suicidal. They are going to want a quick solution, even if it isn't so good in the long term, provided it's good enough. If i tap "Jackdaws love my big sphinx of quartz" out in Morse, i produce ninety-seven dits or dahs. If i do the same with my system, i get hundreds of blinks. However, i also get someone who can communicate while traumatised and at the lowest emotional ebb of their lives, and they get a ray of hope learning an easier system.

Abbreviation could be faster but that doesn't just mean avoiding vowels in English. "An", "in" and "on" all have the same consonants, and although any words which would be abbreviated would be longer than that, the principle remains the same. However, that would work fine in Hebrew or Arabic. You're absolutely right about the frequency though, and that is an issue. Maybe resolve it by constructing frequency tables from text messages?

In practice, i think the person doing the "listening" would end up autocompleting, which gets round both the long-windedness of the system compared to Morse and the problem of changing relative frequency.

The illness-related option would be appropriate in the early stages of locked-in syndrome, but might be less appropriate later. Maybe there should be several options there, or a user-defined sheet. What i like about the several options option is trying to imagine what would appear on the S&M grid.

Then there's T nine of course, but seeing as that drives me up the wall already, in that position i'd prefer not to go there.

Oh, and the other thing is, yes, they would have to know the code for it to work in an emergency, but people do regularly opt to learn the likes of first aid, self-defence and survival skills for similar reasons, so it could happen if it were widely known.
-- nineteenthly, Apr 07 2008


Maybe blinking costs effort. Maybe it's easier to blink at a certain point in time than a certain number of times. If I have that kind of stroke, let's try gaze direction + blink, that sounds more promising.

> Somebody please tell me this is baked or that something better is, because otherwise there are a load of people out there who are needlessly imprisoned by their disabilities.

An improvement in blink code efficiency will hardly change that, much as we wish it would.

> In practice, i think the person doing the "listening" would end up autocompleting,

Yeah, I think that's true regardless of spelling system. I wonder whether it can actually be rather frustrating to a writer. I could see a situation where an author insists, "no, really, just write what I spell, don't insert yourself by trying to `help'."

Looking into the details of the working relationship between Claude Mendibil and Jean-Do Bauby, the Guardian article runs: "Gradually, the process became quicker until Bauby would only have to spell the first couple of letters before Mendibil guessed the rest of the word. This did lead to some misunderstandings - once, when he asked for his glasses (lunettes), his carers were left wondering what he wanted to do with the Moon (lune)." (Doesn't really sound all that frustrated.)

The process gets faster as its participants become more familiar with it: "Just before she stands, I ask her if she still remembers the alphabet. 'Yes. Do you want me to recite it?' I nod. She does so with astonishing speed, each of the 26 letters spoken in a split-second, delivered with shotgun clarity."
-- jutta, Apr 07 2008


Given that a working vocabulary is about 4000 words, a binary system ought to be able to represent any word in about 12 actions (if they are binary actions like look-left/look right), which is not much worse than typing. With four possible gestures (look left/look right/blink/cross eyes), only 6 gestures are needed. Of course, you need some prearranged word list, but in the long term it would work. You'd wind up looking like Marty Feldman, however, during arguments.
-- MaxwellBuchanan, Apr 07 2008


After a quick check on google to make sure I wasn't about to spill the beans on a classified system>

Wherever an American POW is kept, will surely have a 5X5 grid (excluding K) etched on the wall.
-- MikeD, Apr 07 2008


[Jutta], it wouldn't make all the difference to their lives, but there would still be a big difference between the prospect of no or extremely arduous communication and, let's face it, only somewhat less arduous communication. It really should be more than blinking if possible, which would be less arduous, and i just cannot believe that Bauby didn't have any control over his eye movements, because firstly saccades are necessary for vision, and secondly he was able to watch TV, so for him at least, he would have been able to use a grid by means of "pointing" with his direction of gaze.

The autocompletion issue, i suppose, is a little like a stutterer's (which i used to be, incidentally) frustration at having their sentences completed by listeners attempting to be helpful, which might explain why i don't like autocomplete features on cellphones/mobiles.

[Maxwell], that's great and could be a long-term solution, but it'd take a long time to remember the positions of the whole lot. Speakers of languages whose script includes Chinese characters would probably have a head start, i imagine.

Interesting about the tap code. It could be made more efficient, but i see that abbreviations developed anyway, based on the context.

I would also expect there to be some kind of law regarding information and language, in that languages expressed using fewer characters would need more of them to communicate the same amount of information, so the likes of Hawaiian are probably just as long-winded by this method as Cambodian, Caucasian languages or Japanese, but in a different way. Then again, there might be different levels of redundancy.
-- nineteenthly, Apr 07 2008



random, halfbakery