Alternative Orthographies for GSF

Table of Contents

(skip table of contents)

• 1 Introduction
• 2 Comparison
• 3 Cyrillic
  • 3.1 Varieties showing palatalisation on the consonant
    • 3.1.1 "100%" Cyrillic
    • 3.1.2 "50%" Cyrillic
    • 3.1.3 "25%" Cyrillic
  • 3.2 Varieties showing palatalisation on the vowel
    • 3.2.1 "100%" Cyrillic
    • 3.2.2 "50%" Cyrillic
    • 3.2.3 "25%" Cyrillic
• 4 Armenian
• 5 Georgian
• 6 Arabic
  • 6.1 Full orthography
  • 6.2 Basic orthography
• 7 Hebrew
• 8 Thaana
• 9 Devanagari
• 10 Thai
• 11 Yi
• 12 Kanji + Kana
  • 12.1 scriptio defecta
  • 12.2 scriptio plena
  • 12.3 scriptio mixta
• 13 Braille
• 14 In the pipeline

Introduction

The main orthographies for GSF are the Greek-based orthography, followed by the Latin-based and ASCII-only orthographies.

However, other orthographies are sometimes used, either as standard transliterations or as orthographies in their own right (for example, by speakers of GSF who are more accustomed to another alphabet). While some of those orthographies are ad hoc, a number have been standardised, and these are demonstrated on this page.

Comparison

Cyrillic

This was the first alternative orthography I developed. Since the official orthography already incorporated a number of Cyrillic letters, in particular for palatal and voiced stops and alveolar and postalveolar fricatives and affricates, it was a small step from the Greek orthography to a Cyrillic orthography by replacing the Greek letters with equivalent -- as far as possibly -- Cyrillic letters.

Varieties showing palatalisation on the consonant

"100%" Cyrillic

This was the first version of the Cyrillic orthography. I took advantage of the fact that there were a number of letters in the Cyrillic alphabet specifically to represent certain Greek letters; in particular, θ ψ ξ.

For additional sounds that were in Greek but not in Cyrillic, I tried to pick something somehow suitable. I studied the alphabet of a number of languages using the Cyrillic alphabet in order to determine how they represented specific sounds; in particular, с γ δ ŋ.

Of those, γ ŋ were the easiest, as it seemed that "ghe with stroke" and "en with descender", respectively, were fairly standard for those sounds. For с, as the palatal version of χ, I picked a similar letter; namely, "ha with descender". But δ had me stumped for quite a bit. There didn't seem to be a good precedent in any of the languages I looked at (I also seem to recall reading somewhere that it's a comparatively rare sound world-wide). I couldn't use the etymological equivalent, д, since I was already using that for the voiced alveolar/dental stop and didn't want to use it for the fricative.

I finally settled on ђ, as a sort-of d-like sound.

Another issue to determine was how to represent ι: the obvious candidates were и and і. I went with the latter in the end as being the etymological equivalent (и corresponding more to η; compare the numerical values and the alphabetical order).

In a departure from the Greek orthography, I used г and not ґ for the voiced velar stop, since there was no possible confusion with the uppercase form of the letter for γ.

То ворьо анемо ќе то іљо ҳі лоғомаҳісі пьос апо афтос іне то пьо ісҳіро, отан ена таѯіђьоті ҳі перасі пу ҳі іне перітіљіѯі мено се ена зесто епанофорі.

То ђіо ҳі сімфоњісі оті афто пу прото ҳі боресі на анаңгасі то таѯіђьоті на вғаљі то апанофорі ту ѳа ѳеорісі ѳі пьо ісҳіро апо то ало.

Тоте то ворьо анемо ҳі фісіѯі тосо ђінато осо ҳі боресі, ала осо пьо пољі афто ҳі фісіѯі тосо пьо пољі то таѯіђьоті ҳі тіљіѯі то епанофорі ту јіро апо афто, ќе тељіка то ворьо анемо ҳі еңгатаљіѱі то проспаѳіа.

Тоте то іљо ҳі ламѱі зесто, ќе амесос то таѯіђьоті ҳі вғаљі то епанофорі ту.

Еці то ворьо анемо ҳі анаңгасі ѳі на омолојісі оті то іљо ҳі іне то пьо ісҳіро апо тос ђіо.

"50%" Cyrillic

Since a number of characters I had picked for my first script weren't in common fonts, I tried to create a version that only used characters from the Russian, Ukrainian, Serbian, and Macedonian alphabets, since support for those characters seems to be the best.

My initial draft of this version used digraphs for ψ ξ (viz., пс and кс), but I wanted to make unambiguous conversion from this script to Greek possible, so I decided on пъ and къ instead, using ъ as a kind of all-purpose modifier -- since I also used to make нъ for ŋ. С turned into хь, since it's basically a palatal χ, though with the simplification (based on the ASCII-only orthography) that the soft sign is not written before е or і. And the voiceless counterpart of ђ became ћ, naturally enough. (It helped that, in particular in its capital form Ћ, it looks a bit like a "Th" digraph.)

I decided to make г stand for γ here rather than ґ and used ґ for ґ, as in the Greek orthography.

То ворьо анемо ќе то іљо хі логомахісі пьос апо афтос іне то пьо ісхіро, отан ена такъіђьоті хі перасі пу хі іне перітіљікъі мено се ена зесто епанофорі.

То ђіо хі сімфоњісі оті афто пу прото хі боресі на ананъґасі то такъіђьоті на вгаљі то епанофорі ту ћа ћеорісі ћі пьо ісхіро апо то ало.

Тоте то ворьо анемо хі фісікъі тосо ђінато осо хі боресі, ала осо пьо пољі афто хі фісікъі тосо пьо пољі то такъіђьоті хі тіљікъі то епанофорі ту јіро апо афто, ќе тељіка то ворьо анемо хі енъґатаљіпъі то проспаћіа.

Тоте то іљо хі лампъі зесто, ќе амесос то такъіђьоті хі вгаљі то епанофорі ту.

Еці то ворьо анемо хі ананъґасі ћі на омолојісі оті то іљо хі іне то пьо ісхіро апо тос ђіо.

"25%" Cyrillic

This is a version of the above, using only the characters in the Russian alphabet.

Ъ gets a bit more to do here, since it now also appears in дъ (for д; the letter д itself standing for δ in this orthography -- since that sound is more common, I wanted it to be less marked orthographically), тъ (for θ), and гъ (for ґ); this last led to the lovely (*cough*) trigraph гъь for ѓ. Ѕ and џ were "spelled out" as дз and дж, respectively; this means that the conversion is not unambiguous, but I had already used дъ and nothing better came to mind. Also, I figured that the sequence дζ would be extremely unlikely (and дж even more so, since neither ж nor џ are a part of the core alphabet), so the ambiguity would be small in practice.

This orthography also uses и instead of і, and the soft sign ь for гъь кь ль нь (again, with simplification before и and, for velars, е).

То ворьо анемо ке то ильо хи логомахиси пьос апо афтос ине то пьо исхиро, отан ена такъидьоти хи пераси пу хи ине перитиликъи мено се ена зесто епанофори.

То дио хи симфониси оти афто пу прото хи бореси на ананъгъаси то такъидьоти на вгали то епанофори ту тъа тъеориси тъи пьо исхиро апо то ало.

Тоте то ворьо анемо хи фисикъи тосо динато осо хи бореси, ала осо пьо поли афто хи фисикъи тосо пьо поли то такъидьоти хи тиликъи то епанофори ту гьиро апо афто, ке телика то ворьо анемо хи енъгъаталипъи то проспатъиа.

Тоте то ильо хи лампъи зесто, ке амесос то такъидьоти хи вгали то епанофори ту.

Еци то ворьо анемо хи ананъгъаси тъи на омологиси оти то ильо хи ине то пьо исхиро апо тос дио.

Varieties showing palatalisation on the vowel

At one point, I remembered that in Russian orthography, the vowel signs can actually modify the consonant -- there are palatalised and unpalatalised consonants, and the difference is not marked on the consonant but on the following vowel.

That seemed like a fairly natural choice for GSF, too, since it also uses palatalisation to a fair extent. So by using five extra vowel symbols, I could save a couple of consonant symbols.

I decided to use а е и о у to indicate lack of palatalisation and я є і ё ю to denote its presence -- loosely based on Ukrainian (where и also does not palatalise, and which uses є and і similarly, as I understand it -- but not ё, for which it uses ьо instead).

"100%" Cyrillic

То ворё анемо кє то илё хі логомахіси пёс апо афтос ине то пё исхіро, отан ена таѯиђёти хі пераси пу хі ине перитиліѯи мено се ена зесто епанофори.

То ђио хі симфоніси оти афто пу прото хі бореси на анаңґаси то таѯиђёти на вгалі то епанофори ту ѳа ѳеориси ѳи пё исхіро апо то ало.

Тоте то ворё анемо хі фисиѯи тосо ђинато осо хі бореси, ала осо пё полі афто хі фисиѯи тосо пё полі то таѯиђёти хі тиліѯи то епанофори ту гіро апо афто, кє теліка то ворё анемо хі еңґаталіѱи то проспаѳиа.

Тоте то илё хі ламѱи зесто, кє амесос то таѯиђёти хі вгалі то епанофори ту.

Еци то ворё анемо хі анаңґаси ѳи на омологіси оти то илё хі ине то пё исхіро апо тос ђио.

"50%" Cyrillic

(This was actually the first version of the palatalised variety; the "100%" variety came afterwards.)

То ворё анемо кє то илё хі логомахіси пёс апо афтос ине то пё исхіро, отан ена такъиђёти хі пераси пу хі ине перитилікъи мено се ена зесто епанофори.

То ђио хі симфоніси оти афто пу прото хі бореси на ананъґаси то такъиђёти на вгалі то епанофори ту ћа ћеориси ћи пё исхіро апо то ало.

Тоте то ворё анемо хі фисикъи тосо ђинато осо хі бореси, ала осо пё полі афто хі фисикъи тосо пё полі то такъиђёти хі тилікъи то епанофори ту гіро апо афто, кє теліка то ворё анемо хі енъґаталіпъи то проспаћиа.

Тоте то илё хі лампъи зесто, кє амесос то такъиђёти хі вгалі то епанофори ту.

Еци то ворё анемо хі ананъґаси ћи на омологіси оти то илё хі ине то пё исхіро апо тос ђио.

"25%" Cyrillic

I had originally not planned to make a "25%" version of the palatalised orthography, since I thought I needed more vowel signs than I had available.

However, a look at a number of other languages using the Cyrillic alphabet showed the possibility of using а э ы о у versus я е и ьо ю; while э looks a bit marked and ы a bit odd from the point of pronunciation, one could do worse. (I considered using ё for о + palatalisation but went with ьо in the end, mostly because several natural languages appear to have gone this route.)

The consonants themselves are basically the same as for the non-palatalised varieties of the Cyrillic orthography, though palatal consonants are omitted.

То ворьо анэмо ке то ыльо хи логомахисы пьос апо афтос ынэ то пьо ысхиро, отан эна такъыдьоты хи пэрасы пу хи ынэ пэрытыликъы мэно сэ эна зэсто эпанофоры.

То дыо хи сымфонисы оты афто пу прото хи борэсы на ананъгъасы то такъыдьоты на вгали то эпанофоры ту тъа тъэорысы тъы пьо ысхиро апо то ало.

Тотэ то ворьо анэмо хи фысыкъы тосо дынато осо хи борэсы, ала осо пьо поли афто хи фысыкъы тосо пьо поли то такъыдьоты хи тыликъы то эпанофоры ту гиро апо афто, ке тэлика то ворьо анэмо хи энъгъаталипъы то проспатъыа.

Тотэ то ыльо хи лампъы зэсто, ке амэсос то такъыдьоты хи вгали то эпанофоры ту.

Эцы то ворьо анэмо хи ананъгъасы тъы на омологисы оты то ыльо хи ынэ то пьо ысхиро апо тос дыо.

Armenian

Armenian and Georgian came next.

It helped that the alphabets seemed to be related, historically, to Greek, as is apparent in the alphabetical order, among other things. So I relied quite a bit on etymological relationships, even if the current pronunciations are not that close -- in particular, I used aspirate voiceless stops to represent Greek voiceless fricatives.

I decided to use p-s and k-s digraphs for ψ ξ, and digraphs with y for palatalised κ ґ λ ν. I had originally planned to have a separate letter for с and to use y for ј, but decided to use digraphs with y (khy, ghy) by analogy instead. (With the "y" being dropped before "e" and "i", again.)

Difficult sounds to find letters for were δ (again), ŋ, and υ. For the first, I went with chh (more or less arbitrarily); for the second, c'h (since the shape was a bit similar, with a "tail" curling down like with ŋ). For the third, I briefly debated using "ow" (as the sound is written in current Armenian, as in modern Greek) but decided I wanted a digraph, so I went with simply "w". If "w" can be used as a vowel letter in the Latin alphabet used for Welsh, I thought, it can also be one in the Armenian alphabet used for GSF!

Տո վոռյո անէմո կէ տո իլյո քի լողոմաքիսի պյոս ապո ափտոս ինէ տո պյո իսքիռո, ոտան էնա տակսիչյոտի քի պէռասի պւ քի ինէ պէռիտիլիկսի մէնո սէ էնա զէստո էպանոփոռի.

Տո չիո քի սիմփոնիսի ոտի ափտո պւ պռոտո քի բոռէսի նա անացգասի տո տակսիչյոտի նա վղալի տո էպանոփոռի տւ թա թէոռիսի թի պյո իսքիռո ապո տո ալո.

Տոտէ տո վոռյո անէմո քի փիսիկսի տոսո չինատո ոսի քի բոռէսի, ալա ոսո պյո պոլի ափտո քի փիսիկսի տոսո պյո պոլի տո տակսիչյոտի քի տիլիկսի տո էպանոփոռի տւ ղիռո ապո ափտո, կէ տէլիկա տո վոռյո անէմո քի էցգատալիպսի տո պռոսպաթիա.

Տոտէ տո իլյո քի լամպսի զէստո, կէ ամէսոս տո տակսիչյոտի քի վղալի տո էպանոփոռի տւ.

Էծի տո վոռյո անէմո քի անացգասի թի նա ոմոլողիսի ոտի տո իլյո քի ինէ տո պյո իսքիռո ապո տոս չիո.

Georgian

Since I had done the Armenian orthography first, I based the Georgian one on it, especially for the choice of c' for δ and of q' (with a curly descender) for ŋ. I was also initially unsure whether I could use y for the soft sign, since it isn't used in current Georgian any more, but decided to go with it in the end since it was in the (few) fonts I looked at.

ტო ვორჲო ანემო კე ტო ილჲო ქი ლოღომაქისი პჲოს აპო აფტოს ინე ტო პჲო ისქირო, ოტან ენა ტაკსიწჲოტი ქი პერასი პუ ქი ინე პერიტილიკსი მენო სე ენა ზესტო ეპანოფორი.

ტო წიო ქი სიმფონისი ოტი აფტო პუ პროტო ქი ბორესი ნა ანაყგასი ტო ტაკსიწჲოტი ნა ვღალი ტო ეპანოფორი ტუ თა თეორისი თი პჲო ისქირო აპო ტო ალო.

ტოტე ტო ვორჲო ანემო ქი ფისიკსი ტოსო წინატო ოსო ქი ბორესი, ალა ოსო პჲო პოლი აფტო ქი ფისიკსი ტოსო პჲო პოლი ტო ტაკსიწჲოტი ქი ტილიკსი ტო ეპანოფორი ტუ ღირო აპო აფტო, კე ტელიკა ტო ვორჲო ანემო ქი ეყგატალიპსი ტო პროსპათია.

ტოტე ტო ილჲო ქი ლამპსი ზესტო, კე ამესოს ტო ტაკსიწჲოტი ქი ვღალი ტო ეპანოფორი ტუ.

ეცი ტო ვორჲო ანემო ქი ანაყგასი თი ნა ომოლოღისი ოტი ტო ილჲო ქი ინე ტო პჲო ისქირო აპო ტოს წიო.

Arabic

Full orthography

This was the first Arabic version.

I had a look at Uyghur and a couple of other languages that display vowels as separate letters, rather than as diacritics or just leaving them out altogether or hinting at them with matres lectionis, to see what they picked for various sounds. Based on that, I decided to go with alif for α, "e" for ε, ya' for ι, waw for ο, and waw-with-damma for υ. (Other candidates for ε included yeh baree and final-form heh; another candidate for ο was waw with inverted v.)

The consonants were a little tough in some cases, and I'm still not quite satisfied with the assignments. In particular, given that extended Arabic has so many letters, it's a bit frustrating that I can't seem to find "nice" assignments for с ґ ј ќ љ њ ψ ξ. (Subjective, I know.)

I was pleased to find out, though, that there were already reasonably choices for ѕ and ц, since Pashto (IIRC) has letters for these sounds. (I had originally planned to use dad and toe, respectively, and was pleased to find that Hausa uses toe for /ts/. But the Pashto assignments seemed "cleaner" or more authentic, somehow.)

The β sound was a toss-up between waw, waw with three dots ("ve"), and feh with three dots; the last one won out, partly because it has four forms and not just two, so it makes for a nicer initial and medial character. For ŋ, I went with kaf-with-three-dots ("ng"), after considering gaf-with-two-dots ("ngoeh"; as in Pashto or Sindhi, IIRC) and `ayn-with-three-dots (as in old Malay).

For the soft sign, I used hamza-on-yeh. Not sure whether something better might show up.

FWIW, current vague candidates for palatal and palatalised consonants are: hah for ç; nyeh for ј (since it has two dots below like yeh); kaf with ring or kaf with dot above for ќ; gaf with ring or gaf with two dots below (or possibly even gueh) for ѓ); lam with dot above for љ; noon with ring and noon with dot below for њ. Ideally, I'd like a similar "motif" for all of ќ ѓ љ њ, e.g. a ring or dot above or two dots below for each, but I can't seem to find good characters for those. In particular, the number of letters based on lam is rather low. And while two dots below would be cute for similarity to yeh, there's no appropriate character based on kaf or lam, and the one for noon might be confusing (though the medial shape with one dot above and two below shouldn't be more confusing than the shape with one dot above and one below.)

تو ڤورئو انېمو كې تو يلئو خي لوغوماخيسي پئوس اپو افتوس ينې تو پئو يسخيرو، وتان ېنا تاكسيذئوتي خي پېراسي پۇ خي ينې پېريتيليكسي مېنو سې ېنا زېستو ېپانوفوري.

تو ذيو خي سيمفونيسي وتي افتو پۇ پروتو خي بورېسي نا اناڭگاسي تو تاكسيذئوتي نا ڤغالي تو ېپانوفوري تۇ ثا ثېوريسي ثي پئو يسخيرو اپو تو الو.

توتې تو ڤورئو انېمو خي فيسيكسي توسو ذيناتو وسو خي بورېسي، الا وسو پئو پولي افتو خي فيسيكسي توسو پئو پولي تو تاكسيذئوتي خي تيليكسي تو ېپانوفوري تۇ غيرو اپو افتو، كې تېليكا تو ڤورئو انېمو خي ېڭگاتاليپسي تو پروسپاثيا.

توتې تو يلئو خي لامپسي زېستو، كې امېسوس تو تاكسيذئوتي خي ڤغالي تو ېپانوفوري تۇ.

ېڅي تو ڤورئو انېمو خي اناڭگاسي ثي نا ومولوغيسي وتي تو يلئو خي ينې تو پئو يسخيرو اپو توس ذيو.

Basic orthography

I thought I'd try another variant, limiting myself to "common" letters: the Arabic alphabet and the four Persian additions. After a bit of looking, I added feh-with-three dots for /v/, since that also seemed pretty well-supported.

This restriction actually made my life easier; since there weren't so many characters to play with in the first place, I didn't have to dither so much about find the best possible character and simply went with something which could work.

For vowels, I kept alif for α, yeh for ι, and waw for ο, and went with heh for ε (a bit odd, but could work; looks especially good in final form) and waw-with-hamza for υ (similar visually to waw-with-damma, aka "Arabic letter U").

Here, I even came up with letters for ψ and ξ: sad and qaf, respectively. (I like qaf for /ks/, since it's kind of k-ish, anyway.) I went with the old dad-for-dz and toe-for-ts shtick here, too. `Ayn stood in for /ŋ/, and I kept yeh-with-hamza for soft sign.

تو ڤورئو انهمو كه تو يلئو خي لوغوماخيسي پئوس اپو افتوس ينه تو پئو يسخيرو، وتان هنا تاقيذئوتي خي پهراسي پؤ خي ينه پهريتيليقي مهنو سه هنا زهستو هپانوفوري.

تو ذيو خي سيمفونيسي وتي افتو پؤ پروتو خي بورهسي نا اناعگاسي تو تاقيذئوتي نا ڤغالي تو هپانوفوري تؤ ثا ثهوريسي ثي پئو يسخيرو اپو تو الو.

توته تو ڤورئو انهمو خي فيسيقي توسو ذيناتو وسو خي بورهسي، الا وسو پئو پولي افتو خي فيسيقي توسو پئو پولي تو تاقيذئوتي خي تيليقي تو هپانوفوري تؤ غيرو اپو افتو، كه تهليكا تو ڤورئو انهمو خي هعگاتاليصي تو پروسپاثيا.

توته تو يلئو خي لامصي زهستو، كه امهسوس تو تاقيذئوتي خي ڤغالي تو هپانوفوري تؤ.

هطي تو ڤورئو انهمو خي اناعگاسي ثي نا ومولوغيسي وتي تو يلئو خي ينه تو پئو يسخيرو اپو توس ذيو.

Hebrew

Since Hebrew stops (not just voiced but also voiceless; the "begad kefat" set) were fricatives in certain circumstances, I decided to use that to a systematic advantage: I used pe-tav-kaf-bet-dalet-gimel for /p t k b d g/ and fe-sav-khaf-vet-dhalet-ghimel for /f T x v D G/, marking the stops with a dagesh and letting the dagesh-less form stand for the fricative. I felt that was rather clever and nicely symmetric.

I also decided to go with hiriq as a palatalisation mark; this meant that I could dispense with letters for с ј ѓ ќ љњ, though it did mean that some letters had double diacritics (specifically, gimel-with-hiriq-and-dagesh for ѓ, and kaf-with-hiriq-and-dagesh for ќ, plus the other stops when force-palatalised).

Finding characters for some sounds not present in the basic alphabet wasn't that hard, either; I used z' g' ts' for ж џ ч, respectively, as is fairly well-established practice AIUI, and extended the use of geresh to get d' n' p' k' for ѕ ŋ ψ ξ. (I suppose, for symmetry, I could have used s' for ш, but shin was already there....)

For vowels, I used alef - `ayin - yod - vav+holam - vav+shuruq; the result should not be too surprising. (For `ayin for ε, I followed the lead of Yiddish.)

תּוֹ בוֹרִוֹ אנעמוֹ כִּע תּוֹ ילִוֹ כִי לוֹגוֹמאכִיסי פִּוֹס אפּוֹ אפתּוֹס ינע תּוֹ פִּוֹ יסכִירוֹ، וֹתּאן ענא תּאכּ׳ידִוֹתּי כִי פּעראסי פּוּ כִי ינע פּעריתּילִיכּ׳י מענוֹ סע ענא זעסתּוֹ עפּאנוֹפוֹרי.

תּוֹ דיוֹ כִי סימפוֹנִיסי וֹתּי אפתּוֹ פּוּ פּרוֹתּוֹ כִי בּוֹרעסי נא אנאנ׳גּאסי תּוֹ תּאכּ׳ידִוֹתּי נא בגאלִי תּוֹ עפּאנוֹפוֹרי תּוּ תא תעוֹריסי תי פִּוֹ יסכִירוֹ אפּוֹ תּוֹ אלוֹ.

תּוֹתּע תּוֹ בוֹרִוֹ אנעמוֹ כִי פיסיכּ׳י תּוֹדוֹ פינאתּוֹ וֹסוֹ כִי בּוֹרעסי، אלא וֹסוֹ פִּוֹ פּוֹלִי אפתּוֹ כִי פיסיכּ׳י תּוֹסוֹ פִּוֹ פּוֹלִי תּוֹ תּאכּ׳ידִוֹתּי כִי תּילִיכּ׳י תּוֹ עפּאנוֹפוֹרי תּוּ גִירוֹ אפּוֹ אפתּוֹ، כִּע תּעלִיכּא תּוֹ בוֹרִוֹ אנעמוֹ כִי ענ׳גּאתּאלִיפּ׳י תּוֹ פּרוֹספּאתיא.

תּוֹתּע תּוֹ ילִוֹ כִי לאמפּ׳י זעסתּוֹ، כִּע אמעסוֹס תּוֹ תּאכּ׳ידִוֹתּי כִי בגאלִי תּוֹ עפּאנוֹפוֹרי תּוּ.

עצי תּוֹ בוֹרִוֹ אנעמוֹ כִי אנאנ׳גּאסי תי נא וֹמוֹלוֹגִיסי וֹתּי תּוֹ ילִוֹ כִי ינע תּוֹ פִּוֹ יסכִירוֹ אפּוֹ תּוֹס דיוֹ.

Thaana

Thaana is used to write the Divehi language of the Maldives. And, if you include the modified letters used to transcribe Arabic words, it has just the right number of letters and phonemes for GSF!

Well, nearly; it would have been nice to have a separate letter for ј and ь (and preferably also ϊ), but this'll do, since I don't feel like manhandling one of the remaining letters (e.g. /h/ or /s`/) into either role.

ތޮ ވޮރްޔޮ އަނެމޮ ޗެ ތޮ އިޅޮ ޙި ލޮޣޮމަޙިސި ޕްޔޮސ އަޕޮ އަފްތޮސ އިނެ ތޮ ޕްޔޮ އިސްޙިރޮ، އޮތަނ އެނަ ތަޤިޛްޔޮތި ޙި ޕެރަސި ޕު ޙި އިނެ ޕެރިތިޅިޤި މެނޮ ސެ އެނަ ޒެސްތޮ އެޕަނޮފޮރި.

ތޮ ޛިއޮ ޙި ސިމްފޮޏިސި އޮތި އަފްތޮ ޕު ޕްރޮތޮ ޙި ބޮރެސި ނަ އަނަޢްގަސި ތޮ ތަޤިޛްޔޮތި ނަ ވްޣަޅި ތޮ އެޕަނޮފޮރި ތު ޘަ ޘެއޮރިސި ޘި ޕްޔޮ އިސްޙިރޮ އަޕޮ ތޮ އަލޮ.

ތޮތެ ތޮ ވޮރްޔޮ އަނެމޮ ޙި ފިސިޤި ތޮސޮ ޛިނަތޮ އޮސޮ ޙި ބޮރެސި، އަލަ އޮސޮ ޕްޔޮ ޕޮޅި އަފްތޮ ޙި ފިސިޤި ތޮސޮ ޕްޔޮ ޕޮޅި ތޮ ތަޤިޛްޔޮތި ޙި ތިޅިޤި ތޮ އެޕަނޮފޮރި ތު ޔިރޮ އަޕޮ އަފްތޮ، ޗެ ތެޅިކަ ތޮ ވޮރްޔޮ އަނެމޮ ޙި އެޢްގަތަޅިޞި ތޮ ޕްރޮސްޕަޘިއަ.

ތޮތެ ތޮ އިޅޮ ޙި ލަމްޞި ޒެސްތޮ، ޗެ އަމެސޮސ ތޮ ތަޤިޛްޔޮތި ޙި ވްޣަޅި ތޮ އެޕަނޮފޮރި ތު.

އެޠި ތޮ ވޮރްޔޮ އަނެމޮ ޙި އަންއަޢްގަސި ޘި ނަ އޮމޮލޮޔިސި އޮތި ތޮ އިޅޮ ޙި އިނެ ތޮ ޕްޔޮ އިސްޙިރޮ އަޕޮ ތޮސ ޛިއޮ.

Devanagari

I thought it would be nifty if GSF could be written in Devanagari, too.

The more I looked at it, the more I liked this idea. Also, the writing system is nearly perfect for GSF the way it's currently arranged; the only thing I was missing was a "proper" character for ζ and a basic character for љ, and I wasn't sure which of the sibilants to assign to ш and which to σ.

In the end, though, I settled on a plan which went something like this. Unaspirated velar, palatal, dental, and labial consonants stand for stops; aspirated consonants stand for fricatives. (It's handy that digraphs with h are commonly used for both aspiratation and fricativisation, e.g. "ph" for /p^h/ or /f/.) The retroflex row I used for alveolar and postalveolar fricatives. I went with j* for ζ, s. for ш, and s for σ. Palatal s' and retroflex n. were left over.

I also used the existing "extended Devanagari" characters with nukta for a couple of extra sounds: y* for ж (apparently the typical sound value of this sign anyway), j* for ζ (as mentioned before; also the standard value of this letter), k* for ξ (standard /q/), and p* for ψ (standard /f/). Finally, I used retroflex l. to represent palatal /L/.

The assignment as it stands works better for Sanskrit pronunciations than for Hindi, where apparently the palatal consonants tend towards fricatives (e.g. nominal [c] turning into [tS] or similar), but fit GSF better.

This way, I could also work completely without digraphs for the basic alphabet, which left me well chuffed.

Now to decide on vowels. I went with aa e i o u for α ε ι ο υ, because I wanted a separate vowel letter for α rather than letting it be the inherent vowel. (Partly because I thought I'd use the consonant with its inherent vowel at the end of words, where it'd be shwa or silent in Hindi AIUI.) For symmetry, I also went with aa for the stand-alone vowel.

For accented vowels, I used the "long" versions ai ii au uu, and with "short o" for accented a, since it seemed kinda-sorta appropriate in shape. For diphthong-forming vowels, I went with post-vocalic y and v (the latter not otherwise occurring).

For forced palatalisation, I went with conjunct consonants with y; in general, I did quite a bit of conjunct consonants, turning a consonant not followed by a vowel into consonant + virama. The visual effect of those conjunct consonants is sometimes rather nice (especially when preceded graphically by /i/, which comes before the entire complex!).

All in all, I find it a rather nifty adaptation, even if the keymap was quite a bit more work due to having to distinguish between, say, τ on the one hand (rendered with virama) and τα τε τι το τυ (with vowels, no virama), and I'm glad I did it.

तो भोर्यो आनेमो चे तो इळो छि लोघोमाछिसि प्योस आपो आफ्तोस इने तो प्यो इस्छिरो, ओतान् एना ताक़िध्योति छि पेरासि पु छि इने पेरितिळिक़ि मेनो से एना ज़ेस्तो एपानोफोरि.

तो धिओ छि सिम्फोञिसि ओति आफ्तो पु प्रोतो छि बोरेसि ना आनाङ्गासि तो ताक़िध्योति ना भ्घाळि तो एपानोफोरि तु था थेओरिसि थि प्यो इस्छिरो आपो तो आलो.

तोते तो भोर्यो आनेमो छि फिसिक़ि तोसो धिनातो ओसो छि बोरेसि, आला ओसो प्यो पोळि आफ्तो छि फिसिक़ि तोसो प्यो पोळि तो ताक़िध्योति छि तिळिक़ि तो एपानोफोरि तु झिरो आपो आफ्तो, चे तेळिका तो भोर्यो आनेमो छि एङ्गाताळिफ़ि तो प्रोस्पाथिआ.

तोते तो इळो छि लाम्फ़ि ज़ेस्तो, चे आमेसोस तो ताक़िध्योति छि भ्घाळि तो एपानोफोरि तु.

एटि तो भोर्यो आनेमो छि आनाङ्गासि थि ना ओमोलोझिसि ओति तो इळो छि इने तो प्यो इस्छिरो आपो तोस धिओ.

---

१ चे ओलोक्ळिरो तो झि छि एछि एना घ्लोसा चे एना फोञि. २ चे ओतान आफ्तोस छि चिञिसि आपो तो आनातोळि, आफ्तोस छि भ्रि एना पेधिआधा स्तो झि सेनाआर्; चे छि कातिचिसि एचि. ३ चे तो एना छि पि स्तो आलो· एर्थि, मास आस काञि प्ळिन्थोस, चे मास आस फ़िसि आफ्तोस से फोत्या; चे तो मेन् प्ळिन्थोस छि ख्रिसिमेफ्सि झा आफ्तोस आन्दि झा पेत्रा, तो धे आस्फाल्तो छि ख्रिसिमेफ्सि झा आफ्तोस आन्दि झा पिलो. ४ चे आफ्तोस छि पि· एर्थि, मास आस इकोधोमिसि झा मास एना पोळि चे एना पिर्घो, पु तो कोरिफि तु ना फ्तासि औस तो उरानो; चे मास आस आपोक्तिसि झा मास ओनोमा, मिपोस मास था धिआस्पिरि थि पानो स्तो प्रोसोपो आप्तो झि. ५ चे तो चिरिओ छि कातेभि झा ना धि तो पोळि चे तो पिर्घो, पु तोस झोस आप्तोस आन्थ्रोपोस छि इकोधोमिसि. ६ चे तो चिरिओ छि पि· इधु एना लाओ, चे ओलोस एछि ऐना घ्लोसा, चे आफ्तोस छि आर्छिसि ना काञि तुतो; चे तोरा धेन था एम्बोधिसि थि स्' आफ्तोस काथे ति पु आफ्तोस स्कोपेफ्सि ना काञि. ७ एर्थि, मास आस कातेभि, चे मास आस सिङ्छिसि तो घ्लोसा तास, झा ना मि कातालाभि तो एना तो घ्लोसा आप्तो आलो. ८ चे तो चिरिओ छि धिआस्पिरि आफ्तोस आपो एचि पानो स्तो प्रोसोपो आपो ओलोक्ळिरो तो झि; चे आफ्तोस छि पाफ़ि ना इकोधोमिसि तो पोळि. ९ झाफ्तो, तो ओनोमा तु छि ओनोमासि थि भाभेल् (धिलाधि, सिङ्छिसि), एपिधि एचि तो चिरिओ छि सिङ्छिसि तो घ्लोसा आपो ओलोक्ळिरो तो झि; चे आपो एचि तो चिरिओ छि धिआस्पिरि आफ्तोस पानो स्तो प्रोसोपो आपो ओलोक्ळिरो तो झि.

Thai

Thai has traditionally been used to write Pali and Sanskrit, and so there's an established mapping from Devanagari to Thai. So I simply transplanted by Devanagari mapping to Thai.

The fact that there are no independent vowel letters, as well as the fact that vowels are typed in visual rather than logical order, was a bit annoying, so the keymap may not be perfect (in particular, it's hard to type an "e" or "o" vowel after the fact, since those have parts which go both before and after the consonant -- but if you type an entire syllable, it should be fine).

Another slightly annoying moment came when I saw that there didn't seem to be a good corresponding sign for Devanagari nukta, which I had used to make the equivalent of ξ ζ ψ ж. However, since Thai has a couple of letters more than Devanagari, I picked some of those. Two handy extra ones in the velar and labial series served for ξ and ψ, respectively (even if the velar letter is considered obsolete); an s sound in the palatal row turned into ζ; and (the equivalent of) palatal ś got pressed into service for ж.

I considered leaving out spaces between words, as in Thai, but am not sure how well that will work for GSF (though it appears to have worked well enough for Ancient Greek).

โตะ โภะรฺโยะ อะเนะโมะ เจะ โตะ อิโฬะ ฉิ โละโฆะมะฉิสิ ปฺโยะส อะโปะ อะผฺโตะส อิเนะ โตะ ปฺโยะ อิสฺฉิโระ, โอะตะน เอะนะ ตะฃิธฺโยะติ ฉิ เปะระสิ ปุ ฉิ อิเนะ เปะริติฬิฃิ เมะโนะ เสะ เอะนะ เซะสฺโตะ เอะปะโนะโผะริ.

โตะ ธิโอะ ฉิ สิมฺโผะญิสิ โอะติ อะผฺโตะ ปุ ปฺโระโตะ ฉิ โพะเระสิ นะ อะนะงฺคะสิ โตะ ตะฃิธฺโยะติ นะ ภฺฆะฬิ โตะ เอะปะโนะโผะริ ตุ ถะ เถะโอะริสิ ถิ ปฺโยะ อิสฺฉิโระ อะโปะ โตะ อะโละ.

โตะเตะ โตะ โภะรฺเยะ อะเนะโมะ ฉิ ผิสิฃิ โตะโสะ ธินะโตะ โอะโสะ ฉิ โพะเระสิ, อะละ โอะโสะ ปฺโยะ โปะฬิ อะผฺติ ฉิ ผิสิฃิ โตะโสะ ปฺโยะ โปะฬิ โตะ ตะฃิธฺโยะติ ฉิ ติฬิฃิ โตะ เอะปะโนะโผะริ ตุ ฌิโระ อะโปะ อะผฺโตะ, เจะ เตะฬิกะ โตะ โภะรฺโยะ อะเนะโมะ ฉิ เอะงฺคะตะฬิฝิ โตะ ปฺโระสฺปะถิอะ.

โตะเตะ โตะ อิโฬะ ฉิ ละมฺฝิ เซะสฺโตะ, เจะ อะเมะโสะส โตะ ตะฃิธฺโยะติ ฉิ ภฺฆะฬิ โตะ เอะปะโนะโผะติ ตุ.

เอะฏิ โตะ โภะรฺโยะ อะเนะโมะ ฉิ อะนะงฺคะสิ ถิ นะ โอะโมะโละฌิสิ โอะติ โตะ อิโฬะ ฉิ อิเนะ โตะ ปฺโยะ อิสฺฉิโระ อะโปะ โตะส ธิโอะ.

Yi

The Yi script is very interesting; it's possibly the largest standardised syllabary, with 1165 syllables encoded in Unicode -- stemming for a fairly large number of consonants, seven vowels, and four tones.

The consonant inventory maps rather nicely to that of GSF; the major odd-one-out was, again, the voiced dental fricative δ, but also ρ was a little tricky to slot in. Ψ and ξ again sneaked in as voiceless aspirates.

One fun part was that Yi has prenasalised voiced stops, so I decided to take advantage of them and create single characters for sequences of μб, νд, ŋґ, νѕ, and ŋѓ plus vowel. (And νџ, though I have no immediate use for it, but the code points were there.)

A not-so-fun part, though, was that Unicode only encodes those Yi syllables that are attested (sort of like Unicode 1.0 Hangeul), rather than all possible syllables. This leaves various holes in a two-dimensional (vowel+tone vs. consonant) diagram, some smaller (e.g. no NDIEP), some larger (e.g. no retroflex followed by I or IE, in any tone). And while some of those were for combinations I didn't intend to use, others were more annoying, so I got creative and tried to find other, sort-of similar combinations (typically either with a different vowel -- UO being a common choice, since I wasn't using it for anything else -- or a different consonant).

I also used the various tones to my advantage. I used the mid level tone for the basic CV syllable; the low falling tone (-P) for force-palatalised consonants (e.g. FO for φο but FOP for φьο); the rising tone (-X) for accented vowels (e.g. BU for πυ but BUX for πύ; using this tone has the pleasant effect that the shape is the same as the shape for the basic syllable but with a kind of inverted breve above, which looks a bit like an accent mark); and the high tone (-T) for the combination of forced palatalisation and accent (currently used only for πьό, spelled BOT). Note that since Yi has a palatal series, the -P syllables not used together with κ ґ χ γ ν λ; rather, a combination of the appropriate palatal consonant and the regular vowel is used. (For example, ќα is JA and not GAP.) The -P series is also used with independent vowels to represent the second component of diphthongs; hence, IP for ϊ and UOP for ϋ (since there is no *UP).

For consonants not followed by a vowel, I mapped those to syllables in -Y (for labials and palatals) or in -E (for other consonants). (I couldn't use the same vowel for all consonants due to the aforementioned holes.) I decided to make a special syllable for final sigma, mostly Because I Could, using SY (while using SE for other cases of sigma not followed by vowel).

GSF consonants map to Yi consonants like this:

And vowels map like this:

Here are the syllables I had to re-map because of holes:

Note: if you are looking for a font for Yi, I used SIL Yi (the Keyman keymap that comes with the font also influenced my own keymap for punctuation -- especially the choice to use fullwidth punctuation and the ideographic space character).

I'm told that Code2000 also includes Yi support and that there's a download at Microsoft for GB18030 that includes a font SimSun18030.ttc that also does Yi (as well as Chinese, Tibetan, Mongolian, and Thai), but I haven't verified either claim.

ꄏ　ꃰꉙ　ꀊꆂꂿ　ꏦ　ꄏ　ꀂꆦ　ꑟ　ꇉꊈꂷꑟꌋ　ꀨꌦ　ꀊꀧ　ꀊꃠꄏꌦ　ꀂꆂ　ꄏ　ꀨ　ꀂꌜꑟꉘ，　ꀑꄉꆏ　ꀆꆅ　ꄉꇺꅸꄂ　ꑟ　ꀜꉐꌋ　ꀮ　ꑟ　ꀂꆂ　ꀜꉈꄂꆙꇺ　ꂳꆌ　ꌎ　ꀆꆅ　ꌯꌜꄏ　ꀆꀠꆌꃖꉈ。

ꄏ　ꅪꀑ　ꑟ　ꌋꃋꃖꑌꌋ　ꀑꄂ　ꀊꃠꄏ　ꀮ　ꀴꉘꄏ　ꑟ　ꁧꉌꌋ　ꆅ　ꀊꆅꈴꌋ　ꄏ　ꄉꇺꅸꄂ　ꆅ　ꃼꊂꆙ　ꄏ　ꀆꀠꆌꃖꉈ　ꄖ　ꄣ　ꄟꀑꉈꌋ　ꄜ　ꀨ　ꀂꌜꑟꉘ　ꀊꀧ　ꄏ　ꀊꇉ。

ꄏꄅ　ꄏ　ꃰꉙ　ꀊꆂꂿ　ꑟ　ꃏꌋꇺ　ꄏꌙ　ꅪꆅꄏ　ꀑꌙ　ꑟ　ꁧꉌꌋ，　ꀊꇁ　ꀑꌙ　ꀨ　ꀧꆙ　ꀊꃠꄏ　ꑟ　ꃏꌋꇺ　ꄏꌙ　ꀨ　ꀧꆙ　ꄏ　ꄉꇺꅸꄂ　ꑟ　ꄂꆙꇺ　ꄏ　ꀆꀠꆌꃖꉈ　ꄖ　ꑳꉘ　ꀊꀧ　ꀊꃠꄏ，　ꏦ　ꄅꆙꇤ　ꄏ　ꃰꉙ　ꀊꆂꂿ　ꑟ　ꀆꈴꄉꆙꀺ　ꄏ　ꀴꉘꌜꀠꄜꀊ。

ꄏꄅ　ꄏ　ꀂꆦ　ꑟ　ꇁꃋꀺ　ꌯꌜꄏ，　ꏦ　ꀊꂳꌙꌦ　ꄏ　ꄉꇺꅸꄂ　ꑟ　ꃼꊂꆙ　ꄏ　ꀆꀠꆌꃖꉈ　ꄖ。

ꀆꊏ　ꄏ　ꃰꉙ　ꀊꆂꂿ　ꑟ　ꀊꆅꈴꌋ　ꄜ　ꆅ　ꀑꂿꇉꑳꌋ　ꀑꄂ　ꄏ　ꀂꆦ　ꑟ　ꀂꆂ　ꄏ　ꀨ　ꀂꌜꑟꉘ　ꀊꀧ　ꄏꌦ　ꅪꀑ。

Kanji + Kana

A weird idea: use Japanese to write GSF.

At first I thought this wouldn't work, since Japanese doesn't distinguish enough consonants and because its syllabary has no virama or equivalent.

However, both of those problems can be "overcome" by letting the writing be defective: that is, it doesn't completely and unambiguously represent speech. Instead, some distinctions will not be reflected in the language (such as between some stops and their homorganic fricatives), and consonant clusters will either be simplified by not writing some of the sounds (scriptio defecta) or by writing each some or all of the consonants with dummy or filler vowels (scriptio plena). These dummy vowels are not standardised, though u is a common one, as is the vowel of the preceding or following syllable.

Using Japanese also provides the opportunity of using kanji to represent words (scriptio mixta), a bit as in real Japanese or a bit like the Sumerograms of Akkadian or Hittite. (Though I think I'll use them nearly exclusively for meaning, and not or nearly not for sound, since I've got a syllabary for that.) Note that in this case, too, it is not standardised how much of the word is written as a phonetic complement (to provide an inflected ending or to disambiguate between several words with the same sign but different readings), and that this writing is often ambiguous as to which reading of a japonogram is intended. Conversely, the same word may be rendered by several different japanograms to express various shades of meaning, or simply for stylistic purposes or by scribal preference.

Scriptio mixta is frequently written without spaces, since the mixture of japanograms and syllabic writing serves a similar purpose in delimiting words: many content words will start with a japanogram.

scriptio defecta

と　ぼりょ　あねも　け　と　いりょ　ひ　ろごまひせぃ　ぴょ　あぽ　あと　いね　と　ぴょ　いひろ、　おたん　えな　たきでょてぃ　ひ　ぺらせぃ　ぷ　ひ　いね　ぺりてぃりき　めの　せ　えな　ぜと　えぱのふぉり。

と　でぃお　ひ　せぃんふぉにせぃ　おてぃ　あと　ぷ　ろと　ひ　ぼれせぃ　な　あなんがせぃ　と　たきでょてぃ　な　がり　と　えぱのふぉり　とぅ　た　ておりせぃ　てぃ　ぴょ　いひろ　あぽ　と　あろ。

とて　と　ぼりょ　あねも　ひ　ふぃせぃき　とそ　でぃなと　おそ　ひ　ぼれせぃ、　あら　おそ　ぴょ　ぽり　あと　ひ　ふぃせぃき　とそ　ぴょ　ぽり　と　たきでょてぃ　ひ　てぃりき　と　えぱのふぉり　とぅ　ぎろ　あぽ　あと、　け　てりか　と　ぼりょ　あねも　ひ　えんがたりぴ　と　ぽぱてぃあ。

とて　と　いりょ　ひ　らんぴ　ぜと、　け　あめそ　と　たきでょてぃ　ひ　がり　と　えぱのふぉり　とぅ。

えつぃ　と　ぼりょ　あねも　ひ　あなんがせぃ　てぃ　な　おもろぎせぃ　おてぃ　と　いりょ　ひ　いね　と　ぴょ　いひろ　あぽ　と　でぃお。

scriptio plena

と　ぼりょ　あねも　け　と　いりょ　ひ　ろごまひせぃ　ぴょす　あぽ　あふとす　いね　と　ぴょ　いすひろ、　おたん　えな　たきせぃでょてぃ　ひ　ぺらせぃ　ぷ　ひ　いね　ぺりてぃりきせぃ　めの　せ　えな　ぜすと　えぱのふぉり。

と　でぃお　ひ　せぃんふぉにせぃ　おてぃ　あふと　ぷ　ぷろと　ひ　ぼれせぃ　な　あなんがせぃ　と　たきせぃでょてぃ　な　ぶがり　と　えぱのふぉり　とぅ　た　ておりせぃ　てぃ　ぴょ　いすひろ　あぽ　と　あろ。

とて　と　ぼりょ　あねも　ひ　ふぃせぃきせぃ　とそ　でぃなと　おそ　ひ　ぼれせぃ、　あら　おそ　ぴょ　ぽり　あふと　ひ　ふぃせぃきせぃ　とそ　ぴょ　ぽり　と　たきせぃでょてぃ　ひ　てぃりきせぃ　と　えぱのふぉり　とぅ　ぎろ　あぽ　あふと、　け　てりか　と　ぼりょ　あねも　ひ　えんがたりぴせぃ　と　ぷろすぱてぃあ。

とて　と　いりょ　ひ　らんぴせぃ　ぜすと、　け　あめそす　と　たきせぃでょてぃ　ひ　ぶがり　と　えぱのふぉり　とぅ。

えつぃ　と　ぼりょ　あねも　ひ　あなんがせぃ　てぃ　な　おもろぎせぃ　おてぃ　と　いりょ　ひ　いね　と　ぴょ　いすひろ　あぽ　と　でぃお。

scriptio mixta

と北りょ風けと日ひ語戦せぃ誰あぽ彼いねとぴょ強ろ、おたん一旅でょてぃひ通せぃぷひいね回包きめのせ一温と上着。

と二ひ同声せぃおてぃ彼ぷ一とひ会せぃな強がせぃと旅でょてぃな脱りと上着とぅた考りせぃてぃぴょ強ろあぽと他。

とてと北りょ風ひ吹きとそ強とおそひ会せぃ、あらおそぴょ多彼ひ吹きとそぴょ多と旅でょてぃひ包きと上着とぅ回あぽ彼、け終かと北りょ風ひ捨りぴと試。

とてと日ひ照ぴ暖と、け直そと旅でょてぃひ脱りと上着とぅ。

えつぃと北りょ風ひ強がせぃてぃな認ぎせぃおてぃと日ひいねとぴょ強ろあぽと二。

Braille

Hey, even blind people should be able to read GSF, right?

I tried to use, as much as possible, the standard Braille assignments for Cyrillic, Greek, and Latin, and I think I did this reasonably successfully.

The Braille orthography for GSF uses vowels to indicate palatal or palatalised consonants, so I have ten vowel signs; for ya and yu I based it on Cyrillic; ye yi are "longer" versions of e i; and yo is reflected o (aka "ö" in some other Braille alphabets -- which matches Cyrillic Braille's use of "ü" for yu).

I briefly considered using dots-1456 for δ and dots-23456 for θ, as in Albanian Braille (which are "d" and "t", respectively, plus dot-6), but that clashed with Cyrillic Braille's use of dots-23456 for soft sign as well as with Greek Braille's use of dots-1456 for theta. So in the end, I went with dots-1456 for theta and picked dots-2346 for delta on the basis of its use for "the" (with voiced dental fricative) in English Grade 2 Braille.

Γ is dots-126, from its use as "gh" in English Braille, and ŋ is dots-346, from its use as "ing" in English Braille. (In a similar vein, I considered using dots-146 for ш on the basis of English "sh", but then found that Cyrillic Braille uses dots-156 for this, which I went with.) I don't remember where I got dots-2456 for ѕ from any more, or whether dots-12456 for џ is only because that's the dot combination for ĝ in Esperanto Braille.

For accented vowels, I used dot-4 before (which is pretty standard, I think) and for diphthong vowels, I use dots-45 before.

⠞⠕ ⠧⠕⠗⠪ ⠁⠝⠑⠍⠕ ⠅⠡ ⠞⠕ ⠊⠇⠪ ⠓⠌ ⠇⠕⠣⠕⠍⠁⠓⠌⠎⠊ ⠏⠪⠎ ⠁⠏⠕ ⠁⠋⠞⠕⠎ ⠊⠝⠑ ⠞⠕ ⠏⠪ ⠊⠎⠓⠌⠗⠕⠂ ⠕⠞⠁⠝ ⠑⠝⠁ ⠞⠁⠭⠊⠮⠪⠞⠊ ⠓⠌ ⠏⠑⠗⠁⠎⠊ ⠏⠥ ⠓⠌ ⠊⠝⠑ ⠏⠑⠗⠊⠞⠊⠇⠌⠭⠊ ⠍⠑⠝⠕ ⠎⠑ ⠑⠝⠁ ⠵⠑⠎⠞⠕ ⠑⠏⠁⠝⠕⠋⠕⠗⠊⠲

⠞⠕ ⠮⠊⠕ ⠓⠌ ⠎⠊⠍⠋⠕⠝⠌⠎⠊ ⠕⠞⠊ ⠁⠋⠞⠕ ⠏⠥ ⠏⠗⠕⠞⠕ ⠓⠌ ⠃⠕⠗⠑⠎⠊ ⠝⠁ ⠁⠝⠁⠬⠛⠁⠎⠊ ⠞⠕ ⠞⠁⠭⠊⠮⠪⠞⠊ ⠝⠁ ⠧⠣⠁⠇⠌ ⠞⠕ ⠑⠏⠁⠝⠕⠋⠕⠗⠊ ⠞⠥ ⠹⠁ ⠹⠑⠕⠗⠊⠎⠊ ⠹⠊ ⠏⠪ ⠊⠎⠓⠌⠗⠕ ⠁⠏⠕ ⠞⠕ ⠁⠇⠕⠲

⠞⠕⠞⠑ ⠞⠕ ⠧⠕⠗⠪ ⠁⠝⠑⠍⠕ ⠓⠌ ⠋⠊⠎⠊⠭⠊ ⠞⠕⠎⠕ ⠮⠊⠝⠁⠞⠕ ⠕⠎⠕ ⠓⠌ ⠃⠕⠗⠑⠎⠊⠂ ⠁⠇⠁ ⠕⠎⠕ ⠏⠪ ⠏⠕⠇⠌ ⠁⠋⠞⠕ ⠓⠌ ⠋⠊⠎⠊⠭⠊ ⠞⠕⠎⠕ ⠏⠪ ⠏⠕⠇⠌ ⠞⠕ ⠞⠁⠭⠊⠮⠪⠞⠊ ⠓⠌ ⠞⠊⠇⠌⠭⠊ ⠞⠕ ⠑⠏⠁⠝⠕⠋⠕⠗⠊ ⠞⠥ ⠣⠌⠗⠕ ⠁⠏⠕ ⠁⠋⠞⠕⠂ ⠅⠡ ⠞⠑⠇⠌⠅⠁ ⠞⠕ ⠧⠕⠗⠪ ⠁⠝⠑⠍⠕ ⠓⠌ ⠑⠬⠛⠁⠞⠁⠇⠌⠯⠊ ⠞⠕ ⠏⠗⠕⠎⠏⠁⠹⠊⠁⠲

⠞⠕⠞⠑ ⠞⠕ ⠊⠇⠪ ⠓⠌ ⠇⠁⠍⠯⠊ ⠵⠑⠎⠞⠕⠂ ⠅⠡ ⠁⠍⠑⠎⠕⠎ ⠞⠕ ⠞⠁⠭⠊⠮⠪⠞⠊ ⠓⠌ ⠧⠣⠁⠇⠌ ⠞⠕ ⠑⠏⠁⠝⠕⠋⠕⠗⠊ ⠞⠥⠲

⠑⠉⠊ ⠞⠕ ⠧⠕⠗⠪ ⠁⠝⠑⠍⠕ ⠓⠌ ⠁⠝⠁⠬⠛⠁⠎⠊ ⠹⠊ ⠝⠁ ⠕⠍⠕⠇⠕⠣⠌⠎⠊ ⠕⠞⠊ ⠞⠕ ⠊⠇⠪ ⠓⠌ ⠊⠝⠑ ⠞⠕ ⠏⠪ ⠊⠎⠓⠌⠗⠕ ⠁⠏⠕ ⠞⠕⠎ ⠮⠊⠕⠲

In the pipeline

Some more vague ideas for scripts that are in the pipeline:

• maybe Khmer, on the Thai and/or Devanagari model; the independent vowel letters and "proper" vowel encoding (all vowels are combining characters, as I understand it; none of this "what appears before the consonant is typed before the consonant as a spacing character" business) may make this a better choice than Thai, though I think font support may be worse
• Probably not Lao, on account of having too few characters
• Ogham?
• Runes?
• Tengwar and/or Cirth?
• Ethiopic, though the syllabic thing might not work so well
• Canadian Aboriginal Syllabics (though the four-vowel model might flummox me, and I'm not sure whether even Carrier has enough of the "right" consonants)

$Id: altorthography.html 267 2007-06-15 04:54:53Z PNE $