A Latin Abugida

[quadibloc]

This topic was imported from the Typophile platform

I can not now locate the thread in which I posted the idea that we could add some letters borrowed from Cyrillic to the Latin alphabet to have distinctive letters for the unvoiced b, g, and d sounds in Chinese.

Anyways, my recent studies of the Brahmi-derived abugidas of the East have suggested another idea to me.

Tibetan places a dot between each syllable, and it seems to me that one of the reasons Devanagari uses ligatures extensively is also to make the division between syllables unambiguous.

I have been thinking of how the division between upper- and lower- case in the Latin alphabet could be used for the same purpose.

Instead of having any need for a virama symbol, things would work like this:

Small capitals, having no ascenders or descenders, would stand for consonants plus an inherent -a vowel following.

Small capital A would stand for the "null" consonant.

Accents would be used to select other vowels - Á would be e, À would be i, Â would be o, and Ä would be u. But while those accents would be used on consonants like K, in the case of A, instead of accents, Á would be written as small-capital E, À would be written as small-capital И, Â would be written as small-capital O, and Ä would be written as small capital U.

И is used instead of I because small-capital I following a small-capital letter is used to form a dipthong, indicating a long vowel. To follow existing practice (in Devanagari, to which this would be essentially isomorphic) the accent goes on the I instead of the preceding letter.

Capital letters, which always ascend, not leaving as much room for accents, would indicate pure consonants - thus avoiding the virama. A capital letter would be used to indicate the last consonant of a syllable if there is such a consonant after the vowel.

All other consonants in the syllable, whether consonants before the consonant immediately preceding the vowel (the small capital one) or after the vowel but prior to the last consonant, would be in lower case.

Where the lower-case letter is not sufficiently distinct from the small capital, an alternate is found; i.e., lower-case s could possibly be the long s.

This way, speakers of languages such as Tibetan, Burmese, Thai, Hindi, Gujarati... could all use the same alphabet, without having to change how they spell, and they would be using glyphs with (most of) which many of them would already be familiar.

So, basically, the alphabet would be like this:

Vowels:
A(अ) E(ए) И(इ) O(ओ) U(उ)
AI(आ) EI(ऐ) ИI(ई) OI(औ) UI(ऊ)

Consonants:
- Gutturals
K(क) X(ख) G(ग) Г(घ) ?(ङ)

- Palatals
Ч(च) Ջ(छ) J(ज) Ж(झ) Ն(ञ)

- Cerebrals
Δ(ट) Θ(ठ) Դ(ड) Ð(ढ) ?(ण)

- Dentals
T(त) Þ(थ) D(द) Д(ध) N(ऩ)

- Labials
P(प) П(फ) B(ब) Б(भ) M(म)

- Semivowels
Y(य) R(र) L(ल) Л(ळ) V(व)

- Sibilants
Ш(श) Շ(ष) S(स) H(ह)

To get four kinds of N, though, since the Cyrillic N looks like H, I'm at a loss for what alphabets to raid. Apparently, ण has changed its form, as I expected a slanting stroke at the bottom, starting from the first vertical line, not touching the other two vertical lines (instead of looping into the second)... initially, I had some trouble finding all the Devanagari characters I was trying to match.

[John Hudson]

This looks to me like an elaborate transliteration scheme, in which you use Latin and Cyrillic letters to represent not the phonetics of language (transcription) but the graphical structure of how the language is represented in another writing system. Considering that there are already well established Latin alphabetic transcription systems for Indian languages, I wonder if trying to introduce a Latin+ abugida transliteration system would not simply be confusing.

There are regional variant forms of Devanagari letters, which might explain some of your difficulties identifying some characters. The Unicode charts use modern Hindi forms of the western variety.

[quadibloc]

My goals were more ambitious than a mere transliteration scheme.

Instead, this is being (although not too seriously) proposed as a new script that would replace almost all existing writing systems.

According to the book "Significance of the Alphabet", by Charles Kraitsir, "The Dēva nāgari is the only scheme of writing perfectly adapted to represent to the eye the Indo-European tongues."

Well, English is an Indo-European tongue. So, with this abugida, English can be written, effectively, in Devanagari, without its speakers having to learn a whole new alphabet which would be less legible to them.

And because of the dominance of English, many people in the languages using abugidas derived from the Brahmi are familiar with Latin letters. Right now, Thai, Burmese, Lao, Tibetan, Hindi, Tamil, and so on... while they are written either in Devanagari or systems essentially isomorphic to it, they are written in a wide variety of different alphabets.

If, instead, they all used the same alphabet, then they could all take advantage of the great quantity of type design work done for European alphabets (although throwing in a few Armenian letters somewhat weakens this argument), and one obstacle to communication between those nations would be removed.

[John Hudson]

According to the book "Significance of the Alphabet", by Charles Kraitsir, "The Dēva nāgari is the only scheme of writing perfectly adapted to represent to the eye the Indo-European tongues."

Er, yeah. What does that even mean?

Let me rephrase my original comment: I think what you have devised is a very nice parlour game, and one that I would probably enjoy playing.

[quadibloc]

Well, I'll agree that one person praising Devanagari in that fashion doesn't prove anything. His complaint about our alphabet was that in English and French, because it doesn't have enough letters, combinations like th and sh are needed to make up the deficiency... and in Polish, sometimes the very same letter needs more than one different accent mark! (French has both acute and grave accents on the e, but perhaps following Devanagari, he was only worried about the consonants.)

Devanagari, on the other hand, had enough letters to follow the rule of one letter per sound! Thus, why invent something artificial, like the Shaw Alphabet, when the natural genius of mankind has already created a phonetic alphabet ready for use.

[hrant]

My kind of stuff! Keep it up my friend - probability of adoption is no foundation for cultural progress.

hhp

[quadibloc]

Now I see, though, that the idea posed in the earlier thread is (almost) superfluous. In researching additional data to satisfy the curiosity I would have piqued by adding pictures of old typewriter arrangements to my web site, I have found that one doesn't have to go to Sanskrit and its related abugidas... there is one existing alphabetic script that has three letters instead of two where the Latin alphabet has (b,p) or (g,k) or (d,t), thus representing voiced b, g, and d, plosive p, k, and t, and the unvoiced but not plosive letters found in Chinese.

At least in its Eastern and Classical pronounciations. The Western version has two consonant sounds in each group, not three, and so two of the three letters in each group sound the same.

Unfortunately, unlike Greek and Cyrillic, for example, the letter forms are very unfamiliar to Latin-alphabet users. Thus, while lowercase h looks the same (when it's not written as a sort of zig-zag line) and O actually stands for O, but (uppercase) S stands for t, and too much else doesn't even begin to suggest a sound... and so I don't think it will be possible to persuade speakers of English, French, Spanish, and so on to switch to this alphabet simply so that they can now unambiguously and correctly transliterate Chinese.

Of course, this reminds me of an old Hägar the Horrible cartoon, in which the title character responds to the statement "We will never have peace on Earth until all men learn to understand each other!" - he responds, in one panel, "True..."

and then in the final panel, asks "But how are we going to get everyone to speak Norwegian?".

[Chris_Harvey]

According to the book "Significance of the Alphabet", by Charles Kraitsir, "The Dēva nāgari is the only scheme of writing perfectly adapted to represent to the eye the Indo-European tongues."

That's kinda weird. Devanagari can't even represent Hindi without diacritics: nuqta. The Indo-European languages have such divergent phonologies that what works for Sanskrit isn't going to be good for much else. I'd love to see the Devanagari ligature for words like "sixths" Russian "взгляд". Whenever you see the words "perfectly" and "writing" in the same sentence, beware.

The idea of using Greek letters in Latin script is not without precedent. Bruyas used θ and χ to represent [th] and [kh] in Mohawk, distinguishing them from the other allophones of /t/ and /k/.

speakers of languages such as Tibetan, Burmese, Thai, Hindi, Gujarati... could all use the same alphabet

But why would they want to? And sorry, but replacing the indigenous systems with centuries of literature with essentially a Latin script smacks of...

[ahyangyi]

If you're trying to differentiate sounds of the same phoneme (like /pʰ/ and /p/ in English), you might need more symbols:

Seriously, even IPA, with its plethora of characters, cannot claim to have 1 character for each sound (/pʰ/ is a counterexample). I don't think Devanagari has a chance here.

[quadibloc]

As the Indo-European tongues are a subset of all tongues, their phonemes are a subset of all phonetic entities. So either the Devanagari script or the Armenian alphabet might be adequate for them.

For the world to adopt the Armenian alphabet because it's better suited to transliterating the Wu (Shanghai and related dialects) dialect of Chinese, or for the Brahmi-script users to switch to something Latin-based to suit foreigners... yes, neither is likely to happen. But for at least some purposes, such as new scripts for indigenous people adopting the Latin alphabet, it would be nice for the Latin alphabet to have a few extra standardized letters - so that the same letter would be used to express the same non-English sound in those languages that use it, for example.

[John Hudson]

...it would be nice for the Latin alphabet to have a few extra standardized letters - so that the same letter would be used to express the same non-English sound in those languages that use it, for example.

This is largely what has happened, on a regional basis, in the extension of the Latin alphabet for sub-Saharan African languages and for aboriginal languages of North America. The linguists working in these areas have mostly followed one another in adapting IPA forms for phonemes not found or poorly represented in European orthographies.