Keenir's scratchpad -

[keenir]

I’m confused. Isn’t the defining attribute of an abjad that it doesn’t have mandatory vowels?
Also, I’m not convinced an abjad would work well for Korean as anything other than an intellectual exercise. Remember, abjads were a reasonable choice for Semitic languages precisely because Semitic languages are based around consonantal skeletons. Korean is quite different in this regard — vowels carry a much higher functional load.

When I replied initially, I was pretty sure I knew what you meant by that; but since then, doubt has crept into my brain and taken over...so, in case I misunderstood your meaning, what do you mean by "carry a much higher functional load"?

In Semitic languages, the vowels are more or less incidental to the meaning of the word: they’re very important grammatically, but in terms of distinguishing the roots themselves, they aren’t all that important. By contrast, in a language without consonantal roots, such as Korean — or English, since I’m unfamiliar with Korean — vowels play a hugely important role in distinguishing words such as pen and pin and pun and pan, or cat and kit and cut and cot, or has and his and he’s and whose.

ah, okay; thank you for clarifying that.

Hm, so as a pure abjad (utterly no vowels shown ever), English and Korean would fall flat utterly. But as an impure abjad like Arabic and Hebrew, it would need a lot of work before its stable?
EDIT: hm...wait, maybe also use radicals? pen-{radical ink vs pin-{radical sharp vs pan-{radical flat or cook...its a line of inquiry to try out, even if it amounts to nothing.

Its still just an intellectual exercise, granted, but this discussion helps clarify things on this subject, which I can use in future projects (always assuming my mind doesn't go blank when i need it, which also happens).

[bradrn]

Hm, so as a pure abjad (utterly no vowels shown ever), English and Korean would fall flat utterly. But as an impure abjad like Arabic and Hebrew, it would need a lot of work before its stable?

Yes, exactly.

EDIT: hm...wait, maybe also use radicals? pen-{radical ink vs pin-{radical sharp vs pan-{radical flat or cook...its a line of inquiry to try out, even if it amounts to nothing.

This is basically how Ancient Egyptian and Sumerian worked. It’s also how many Chinese characters are formed.

[Richard W]

Hm, so as a pure abjad (utterly no vowels shown ever), English and Korean would fall flat utterly. But as an impure abjad like Arabic and Hebrew, it would need a lot of work before its stable?

Not for English. Using Perso-Arabic script and doubling the number of vowel marks from 3 to 6 worked fine for me to keep my diary, even though I wrote non-rhotically.

[Rounin Ryuuji]

This is basically how Ancient Egyptian and Sumerian worked. It’s also how many Chinese characters are formed.

Building on this, in Japanese, it makes it sometimes easy to guess what the on'yomi (borrowed Sinitic reading) of a character is, but not the kun'yomi (native Japanese reading), note:

生 (On: sei, shō~jō, Kun: i(kiru), iki, u(mu/mareru), nama; Gloss: "live, be alive, life, raw"; this one is notorious for its huge number of readings)
姓 (On: sei, shō, Kun: kabane; Gloss: "clan, surname")
星 (On: sei, shō, Kun: hoshi; Gloss: "star, dot", also appears in the names of planets, cf. 水星 Suisei "Mercury", literally "Star of Water")

I believe some stages of Persian also used a spelling that was something like "MLK" (cf. Arabic malik) as an ideograph for their word for "king", from a Semitic source, despite the Persian word being etymologically unconnected and sounding entirely different.

[bradrn]

I believe some stages of Persian also used a spelling that was something like "MLK" (cf. Arabic malik) as an ideograph for their word for "king", from a Semitic source, despite the Persian word being etymologically unconnected and sounding entirely different.

Ah yes, that was Book Pahlavi. Wikipedia lists e.g. ⟨KLB'⟩ for /sag/. It also managed to merge the 22 Aramaic letters to only 14 letters — and 14 ambiguous letters, at that, since some ligatures ended up identical to other letters (source). Eventually they got rid of the whole thing and switched to Avestan.

[Richard W]

I believe some stages of Persian also used a spelling that was something like "MLK" (cf. Arabic malik) as an ideograph for their word for "king", from a Semitic source, despite the Persian word being etymologically unconnected and sounding entirely different.

Hittite would do the same, but not only with Sumerian, but also with Akkadian. I don't think speed-reading was a thing in ancient times.

[Rounin Ryuuji]

I don't think speed-reading was a thing in ancient times.

I'm not sure, but I think ancient orthographies were at least somewhat more readable than they're given credit for being. Japanese has one of the worst modern orthographies of which I'm aware, primarily because all the characters have so many readings. That said, japanese becomes much easier when you introduce Hiragana and Katakana into the equation, and aren't trying to spell everything phonetically in Kanji, and write meaning with Kanji, too.

There is something possibly to be said for logographic writing still, however. Even if I don't know whether to say ōka or sakurabana (according to the one Japanese person I've asked, the second form is uncommon, though bilingual dictionaries, in my experience, tend to list it) for 桜花, I do know immediately what the word (probably) means. If on'yomi were eliminated, a part of me thinks that, for this reason, Kanji would be more tenacious than might be initially imagined — it saves a great deal of space to write 水力 compared with the hypothetical kun'yomi word みずちから (mizuchikara), though the form 水の力 (mizu no chikara) would probably be a more normative modern "calque" into inherited Japonic morphemes. Whether you keep the spelling 水力 or invent something new when mizuchikara collapses into something like micchika in daily speech is another question entirely.

[Richard W]

And unpointed Kufic reduces the Arabic script from 28 letters to about 14!

[keenir]

Wanted to try my hand at making a naming language with only one vowel...and, ideally, as small a consonant inventory as possible.

So far, the closest thing to a name it has, is "pre-Deq" as that was what I jotted down at the top of the page to denote that it was what existed before the Deq Writing Reforms a few centuries down the line.

{and a major inspiration for this, was one of the Chinese counting systems I came across in Writing Systems Of The World, with the script rotating 90 degrees every other position}

Not sure "Noun Accompanyment Class" is a real thing...it was just me trying to make a joke. I am sorry, oh so sorry.




thoughts thus far? I have the Deq Reforms (or rather the number-related section thereof)

...here they are:

[Zju]

Have you figured out any vowel allophony? I find it unlikely speakers of it will just go around repeating [ä ä ä ä ä]

[keenir]

Have you figured out any vowel allophony? I find it unlikely speakers of it will just go around repeating [ä ä ä ä ä]

Thank you.

When you asked, I wasn't sure about allophony...I had been thinking this could be the closest I get to completing a One Vowel Project...but I had no thoughts about how to do allophony or wherein to aim, but I at least had a thought about where to place stress.

...ergo my question in the other thread.

For right now, allophony only shows up in the vowels...though I'd wager that, very soon, the <Vsh.sh>s will see a...de-gemminating? to either <Vs.sh> or just to <V.sh>

For what you're about to read, a quick explanation: as this started as a one-vowel conlang, I figured it would work as a Pure Abjad, meaning no marking of any vowels; now that allophony has raised its head, the only markings would be for either stressed or unstressed syllables...so I picked the latter, underscoring them.

äŋ.ɲa.ɲa , 0 ang.nya.nya
äŋ.sa.sa , 1 ang.sa.sa
äŋ.sa.ʃä , 2 ang.sa.sha
äŋ.na.sa , 3 ang.na.sa
äŋ.ʃa.ʃa , 4 ang.sha.sha
äŋ.'äʃ.ma.ja , 5 ang.ash.ma.ya
äŋ.ʃa.na , 6 ang.sha.na
äŋ.sa.dä , 7 ang.sa.da
äŋ.ʃa.da.da , 8 ang.sha.da.da
äŋ.sa.ta , 9 ang.sa.ta
äŋ.äʃ.ʃa.ma , 10 ang.ash.sha.ma
äŋ.äʃ.ma.ka.pa , 15 ang.ash.ma.ka.pa

Thoughts?

[bradrn]

When you asked, I wasn't sure about allophony...I had been thinking this could be the closest I get to completing a One Vowel Project...but I had no thoughts about how to do allophony or wherein to aim, but I at least had a thought about where to place stress.

Languages with only one or two underlying vowels always have many more at the surface level. For instance, Moloko has an underlying vowel inventory /a (ə)/ (with the latter epenthetic), but a surface vowel inventory of [a ɛ œ ɔ ə ø ɪ ʊ i u]. Similarly, Abkhaz has /a ə/ realised as [a e o ə i u], and Kaytetye Arrernte has /a ə/ realised as [a ə i o u]. Usually the allophony is related to the surrounding consonants: e.g. /əj/ might be realised as .

[keenir]

When you asked, I wasn't sure about allophony...I had been thinking this could be the closest I get to completing a One Vowel Project...but I had no thoughts about how to do allophony or wherein to aim, but I at least had a thought about where to place stress.

Languages with only one or two underlying vowels always have many more at the surface level. For instance, Moloko has an underlying vowel inventory /a (ə)/ (with the latter epenthetic), but a surface vowel inventory of [a ɛ œ ɔ ə ø ɪ ʊ i u]. Similarly, Abkhaz has /a ə/ realised as [a e o ə i u], and Kaytetye Arrernte has /a ə/ realised as [a ə i o u]. Usually the allophony is related to the surrounding consonants: e.g. /əj/ might be realised as

Tempting as it would be to put off making more allophones for when the daughter languages are on the table, I don't think I'm going to get to them {to the daughterlangs}, so I need to do this for this conlang.....even if it was hard enough for me to wrap my head around just using two realizations(??) of <a> - and that was thanks to stress in the vowel.

Took me til the most recent re-readthrough of this post that I noticed you mentioning its "related to the surrounding consonants", so I'll see if that sparks any ideas next time I'm reading through the numbers.

Thank you both, Bradrn and Zju for your replies; they are very much appreciated.

{in honesty, I made the numbers' script first, then when I was thinking up the accompanying signs, I thought of Abkhaz's one (or two, depends who you asked), and a consonant inventory ideally not bigger than Piraha or Hawaiian.....yeahhhh i wasn't asking much of my brain}

[bradrn]

Took me til the most recent re-readthrough of this post that I noticed you mentioning its "related to the surrounding consonants", so I'll see if that sparks any ideas next time I'm reading through the numbers.

Yes, this is the important point here. Usually it’s related to palatalisation and labialisation — a lot of small-vowel-system languages got that way by transferring vowel frontness/backness/rounding onto the surrounding consonants. Thus, say, /ə/ might turn into [i] next to palatalised consonants and [u] next to labialised consonants: e.g. Ubykh /wətɕʼa/ is realised as [utɕʼe], and /azaχʲaʃənan/ as [azaχeʃinan]. This is how Arrente turns two vowels into five, Abkhaz realises two as six, and Marshallese gets twelve vowels from four. A more elaborate system is found in Moloko, which like many Chadic languages has so-called ‘prosodies’ acting at the word level: /a/ and epenthetic /ə/ can each be realised in five different ways, depending on the surrounding consonants and the ‘palatalised’ or ‘labialised’ status of the word’s prosody. (I highly recommend having a look at Friesen’s open-access Moloko grammar for details about how this all works.)

[Travis B.]

I would listen to what bradrn is saying. Practically all languages with very small numbers of vowels have elaborate systems of vowel allophony conditioned by adjacent consonant phonemes and very elaborate consonant inventories, with lots of palatalized and labialized consonant phonemes.

[keenir]

I would listen to what bradrn is saying. Practically all languages with very small numbers of vowels have elaborate systems of vowel allophony conditioned by adjacent consonant phonemes and very elaborate consonant inventories, with lots of palatalized and labialized consonant phonemes.

I am listening, and looking at the article linked {thank you for that} and...now wondering: is there a difference between vowel allophony conditioned by and vowel allophony realized as ?

thank you.

[Zju]

Indeed, I'd expect at least reduction to [ɛ~e~i] in (some) unstressed syllables, variation of [æ~a~ɑ(~ɒ)] in stressed syllables, and maybe one or two more allophony patterns. Infact, or [o] is all but guaranteed to pop up somewhere.

[keenir]

Indeed, I'd expect at least reduction to [ɛ~e~i] in (some) unstressed syllables, variation of [æ~a~ɑ(~ɒ)] in stressed syllables, and maybe one or two more allophony patterns. Infact, [u.] or [o.] is all but guaranteed to pop up somewhere.

{i added .s so the coding wouldn't act up}

This is one of those times where I think "wait...what?" Because to my brain, I'd think that reduction would involve the opposite, going from the [e~i] set and [u.] and [o.] all becoming [a.]

But I know enough to know that I'm wrong.

[bradrn]

I would listen to what bradrn is saying. Practically all languages with very small numbers of vowels have elaborate systems of vowel allophony conditioned by adjacent consonant phonemes and very elaborate consonant inventories, with lots of palatalized and labialized consonant phonemes.

I am listening, and looking at the article linked {thank you for that} and...now wondering: is there a difference between vowel allophony conditioned by and vowel allophony realized as ?

The ‘realisation’ of a vowel is what it’s pronounced as phonetically; the ‘condition’ is what determines that realisation. For instance, English /e/ can be realised as [ə], conditioned by stress.

Indeed, I'd expect at least reduction to [ɛ~e~i] in (some) unstressed syllables, variation of [æ~a~ɑ(~ɒ)] in stressed syllables, and maybe one or two more allophony patterns. Infact, or [o] is all but guaranteed to pop up somewhere.

(Tip: you can use [bk] to make brackets without tagging stuff.)

I agree that [u] will probably turn up, usually as /əw/ or /wə/ (or both); [i] is even more likely. As for reduction, that can go in all kinds of directions, and I wouldn’t necessarily expect [ɛ~e~i]: sometimes it involves vowels moving to the centre of the vowel chart, and sometimes it involves them moving to the periphery. (I recall seeing a good article about it, though I can’t quite seem to find it again.) That being said, I don’t know of any small vowel systems with vowel reduction at all.

All of which means…

This is one of those times where I think "wait...what?" Because to my brain, I'd think that reduction would involve the opposite, going from the [e~i] set and [u.] and [o.] all becoming [a.]

But I know enough to know that I'm wrong.

…you’re not necessarily wrong.

[keenir]

The ‘realisation’ of a vowel is what it’s pronounced as phonetically; the ‘condition’ is what determines that realisation. For instance, English /e/ can be realised as [ə], conditioned by stress.

thank you for the explanation.

But I know enough to know that I'm wrong.

…you’re not necessarily wrong.

Ummmm...I'm sorry...I'm confused. With the conditioning and realizations, I thought I was being encouraged to expand what the vowel is pronounced as in these numbers...and now the reduction can have the vowel increasing in how many pronounciations (for lack of my brain firing at this hour - heater's busted) and also the vowel decreasing its number of pronounciations.