- Notifications
You must be signed in to change notification settings - Fork1
cluesurf/tone
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
A Cross-Cultural Writing System
Checkout theDiscord server with anythoughts/comments/questions you have, or start aGitHub discussion or addsomething to theReddit group ifyou'd like as well.
Tone Text (esoteric name Hanákana) is a Universal Writing System (UWS),capable of representing all human voice sounds in a concise and readablescript. We like to think of them asmodern runes. This means you canwrite any human language on Earth in tone text (constructed language ornatural language) and read it without having to learn that language'sspecific quirks. Here is an example of the modern runes using the newToneEtchfont. You can try it in theinteractive online Editor.
Tone text is a sound-based script like theAncient Runes, which means whenyou read see a symbol/letter, you know exactly how it is pronounced.There are no weird edge cases or specific words for which you have tomemorize the sound. You simply learn the script and can pronounce thewords and texts.
You may note that Linguistics also has the International PhoneticAlphabet(IPA),which seems to accomplish a similar thing. But the IPA is not forwriting anything but the most detailed pronunciation guides, not forwriting sentences or long texts. Tone Text is meant for the later, forwriting like you would in a normal language.
The ASCII encoding is mapped to IPA for your knowledge in theChat Text. Ideally, Tone Text wouldbe included in the Unicode standard, butthat is unlikely. Wehave suggested it be used as a writing systemon Mars 😂.
It doesn't necessarily play well with dyslexia, which might find thehighly symmetric patterns difficult. We think there are far bettermethods of learning outside of writing systems than trying to invent theperfect script for all cases. To most natural languages, dyslexia is anunsolved problem. We instead strive to make a minimal seed from whichknowledge can be modelled from primitive parts, memorized, and then usedto derive further knowledge.
Tone Text is also a monospaced font, to be used eventually forprogramming. All sounds are derived from 3 base shapes, that fori
,a
, andu
. Rotating and extending tails on them gives you theremaining base letters. Then adding dots to them extends the sound ofthe base letter slightly. From this it should be straightforward tomemorize how to pronounce words in a day or so, and from there you canread and write any language pretty well.
importtone,{symbols,consonants,vowels,numerals,}from'@cluesurf/tone'// Makes TalkText ASCII format into Tone-compatible font format.tone('aa+xa+a-ita-')// => aa3xa3a4ita4
These are the main sounds. Everything else are variations of thesesounds. These "variations" are accomplished in a structured way, see theguide. Indic retroflex consonants are added with adot on the tailless side. Arabic sounds are added with a dot on thetailed side usually. Clicks are mapped to the closestmouth-position-oriented sound. Ejectives and implosives are simply addedwith accent marks. Etc.
Alphabetical order is this (1-36):
ieaouAEIUOu$o$mnqgdbptkhsfvzjxcCywQ'lr
Everything else is a modification on this base. Here are many moresounds. They are given the numerical value of the corresponding basesymbol, and are ordered correspondingly, with retroflex, thenpharyngealized, then ejectives, implosives, stops, tenses, clicks. Thestops and pharyngealizations don't get any extra points.
N nL lX xD dJ jT tR rH hK kV vF fd~ d (pharyngealization, 1 dot on tailed side)t~ ts~ sC~ Cl~ lp! p (ejective)t! tk! ks! sx! xT! tK! kg? g (implosive)b? bp. p (stop, 2 dots on tailless side)t. tk. kp@ p (tense, 3 dots on tailless side)t@ tk@ ks@ sd* d (click, arc on tailless side)p* pt* tk* kl* l
Here is the full spectrum of sounds basically (about 175 sounds):
ieaouAEIUOu$o$mnqgdbptkhsfvzjxcCywG'lri$a$e$NDTKHFVJXZRLYQp!t!k!s!x!T!K!F!c!X!H!S!g?d?b?b.d.g.p.t.k.p@t@k@s@d*p*t*k*l*mh!nh!qh!gh~dh~bh~ph~th~kh~hh~jh~xh~lh~rh~Xh~Dh~Jh~Th~Lh~Rh~Kh~gQ~dQ~bQ~pQ~tQ~kQ~sQ~cQ~CQ~lQ~xQ~KQ~mG~nG~gG~dG~bG~pG~tG~kG~sG~fG~vG~zG~jG~xG~lG~rG~my~ny~gy~by~py~ky~sy~fy~vy~zy~jy~xy~cy~Cy~ly~ry~Ty~Ky~Xy~mw~qw~gw~dw~bw~pw~tw~kw~Dw~Tw~Kw~gh~y~dh~y~bh~y~ph~y~th~y~kh~y~jh~y~xh~y~lh~y~rh~y~Xh~y~Th~y~Kh~y~
Here is the consonant IPA phonology chart mapping.
There are at least 10 classes of numerological encodings:
- 36 sound encoding. This is where letters are in order based ontheir sound, numbered 1-36.
- 29 sound encoding. This counts the 14 voiced and unvoiced consonantsas 7 instead of 14, as described below.
- 22 shape encoding. These are the unique shapes ignoring mirrorsymmetry. qjfvQxschwkbaoeiE2uAI1.
- 16 shape encoding. These are the unique shapes ignoringrotational symmetry. qnfrQyszhmkpaeIO.
- 9 shape encoding. These are the unique shapes ignoring rotationaland mirror symmetry. They are also numbered similarly. qfQshkaeI.
- 12 shape encoding. These are the tailless shapes, including allsymmetries. They are numbered i, a, o, u.
- 10 shape encoding. These are the tailless shapes, ignoring mirrorsymmetry. They are numbered i, a, u.
- 4 shape encoding. These are the tailless shapes, ignoring rotationalsymmetry. They are numbered i, a, o, u.
- 3 shape encoding. These are the tailless shapes, ignoring rotationaland mirror symmetry. They are numbered i, a, u.
- 2 shape encoding. These are comparing tailed vs. tailless characters,ignoring rotational and mirror symmetry.
The most common are the 36 sound encoding, the 16 shape encoding, andthe 9 shape encoding. The 16 shape encoding is how you would play a gamewith one-sided tiles. The 9 shape encoding is how you would play a gamewith two-sided tiles. The 16 shape encoding can be encoded inhexadecimal. The 9 in digits 1-9.
Here is the 29 sound encoding.
ieaouAEIUOu$o$mnqg kd tb phs zf vj xc CywQ'lr
To get access to the font for your own system, just click and downloadtheToneEtch.otfdirectly.
- Keyman Developmentfor the Keyman keyboard.
pnpm make:keyboard
MIT
This is being developed by the folks atClueSurf, aCalifornia-based project for helping humanity master information andcomputation. Find us onTwitter,LinkedIn, andFacebook. Check out our otherGitHub projects as well!
About
A Cross-Cultural Writing System