Zompist Bboard Again

Darren wrote: ↑Sun Feb 04, 2024 2:48 am

bradrn wrote: ↑Sun Feb 04, 2024 2:16 am For that kind of stuff, I highly recommend Yip’s book Tone. It has a lot of great information about tone sandhi, autosegmental analysis, and so on. (Its only problem is that it relies heavily on Optimality Theory, but that’s not a huge issue.)

What actually *is* Optimality Theory?

a template for producing CV line items. apply theory A to language B, publish, hey presto! you're "doing" "research"

Is it something worth learning?

no

Nortaneous wrote: ↑Sun Feb 04, 2024 12:16 pm

Darren wrote: ↑Sun Feb 04, 2024 2:48 am

bradrn wrote: ↑Sun Feb 04, 2024 2:16 am For that kind of stuff, I highly recommend Yip’s book Tone. It has a lot of great information about tone sandhi, autosegmental analysis, and so on. (Its only problem is that it relies heavily on Optimality Theory, but that’s not a huge issue.)

What actually *is* Optimality Theory?

a template for producing CV line items. apply theory A to language B, publish, hey presto! you're "doing" "research"

I may have a healthy skepticism of linguistic theories, but this strikes me as a little too cynical. There’s at least some interesting stuff in OT, and I think the basic idea of studying cross-linguistic phonological constraints is a worthy one.

Nortaneous wrote: ↑Sun Feb 04, 2024 12:16 pm

Is it something worth learning?

no

Thanks! (that's the answer I wanted)

bradrn wrote: ↑Sun Feb 04, 2024 4:25 pm There’s at least some interesting stuff in OT, and I think the basic idea of studying cross-linguistic phonological constraints is a worthy one.

IMO some of the constraints are interesting (conlangers especially underestimate faithfulness constraints - I see some very silly paradigms sometimes), but I'm not aware of an adequate solution to the problem of candidate form generation, and it looks very unlike current hypotheses about the architecture and functioning of the brain. What else is there?

It's useful to have basic literacy in it - learning to read the spreadsheets and the conventions for naming rules - but as a paradigm it seems silly.

Been reading this over again:
https://www.reddit.com/r/conlangs/comme ... all_gonna/

At first I was confused by it becuase, for their conlang, they told you what onsets produce what tone and what codas produce what tone and nothing about what the result is if you have certaon onsets with certain codas. That's when I realized they were talking about "tone classes" and that they had certain interactions that were summarized by this chart (though I'm still a bit confused how I'm supposed to read it, I think it's onset = columns, codas = rows).

Anyways, the things talked in that post along with the chart seem to match with what I've seen of reconstructed Tibetan or Old Chinese > Middle Chinese words. So I hope what I've come up with based off of it is realistic enough.

One thing I really do worry about is that my conlang will end up looking too much like a Sino-Tibetan language. And it's supposed to be more Caucasian/Siberian/Inuit.

Nortaneous wrote: ↑Mon Feb 05, 2024 1:01 am It's useful to have basic literacy in it - learning to read the spreadsheets and the conventions for naming rules - but as a paradigm it seems silly.

Yeah, this is basically my position too.

conlangers especially underestimate faithfulness constraints

Hmm, I’ve never really got this feeling — do you happen to have any examples?

but I'm not aware of an adequate solution to the problem of candidate form generation, and it looks very unlike current hypotheses about the architecture and functioning of the brain.

Indeed, this is the silly bit. Personally, I think very high-ranking constraints have such a strong effect you might as well consider them restrictions on the input, which makes the theory tractable… but at that point the whole edifice starts crumbling apart (at least according to its practitioners).

Ahzoh wrote: ↑Mon Feb 05, 2024 2:26 am Been reading this over again:
https://www.reddit.com/r/conlangs/comme ... all_gonna/

This looks pretty terrible to me: too dogmatic, and ignoring the variation to be found in natlangs. Not every phoneme produces a consistent tone… in fact, most don’t.

One thing I really do worry about is that my conlang will end up looking too much like a Sino-Tibetan language. And it's supposed to be more Caucasian/Siberian/Inuit.

None of which have tones! (Except Nivkh, and debatably Yeniseian.)

bradrn wrote: ↑Mon Feb 05, 2024 3:40 am This looks pretty terrible to me: too dogmatic, and ignoring the variation to be found in natlangs. Not every phoneme produces a consistent tone… in fact, most don’t.

It's the only thing that's given me a sense of direction in this endeavor. Looking at Tibetan (a language that has kept voicing distinctions and coda consonants) has helped me a little, but not much.

Hmm, was not aware that Ket (one of those Yeniseian languages) might possibly maybe have tone.

bradrn wrote: ↑Fri Feb 02, 2024 8:18 pm

Ahzoh wrote: ↑Fri Feb 02, 2024 6:39 pm I've read a reddit guide on tonogenesis but it doesn't tell you how to manifest tones from closed syllables or syllables with dipthongs or syllables with onset clusters or all three.

I was trying to write up something, but in the process I discovered this paper which looks quite comprehensive: https://doi.org/10.1080/07268602.2022.2157675. It looks like it should answer most of your questions. (I can send you a PDF if you can’t access it.)

Say what? Both sylable loss and open vowels (in Ambel) are reported to have triggered tonogenesis? A conlang of mine's case of second syllable vowels' openness inducing tone in the first syllable and then dropping out might not be so hopeless, after all.

Zju wrote: ↑Mon Feb 05, 2024 3:35 pm

bradrn wrote: ↑Fri Feb 02, 2024 8:18 pm

Ahzoh wrote: ↑Fri Feb 02, 2024 6:39 pm I've read a reddit guide on tonogenesis but it doesn't tell you how to manifest tones from closed syllables or syllables with dipthongs or syllables with onset clusters or all three.

I was trying to write up something, but in the process I discovered this paper which looks quite comprehensive: https://doi.org/10.1080/07268602.2022.2157675. It looks like it should answer most of your questions. (I can send you a PDF if you can’t access it.)

Say what? Both sylable loss and open vowels (in Ambel) are reported to have triggered tonogenesis? A conlang of mine's case of second syllable vowels' openness inducing tone in the first syllable and then dropping out might not be so hopeless, after all.

I think you have to look at the individual cases to check what is going on. I think there is some confusion because of recurring behaviours. For example, in Thai, modern onset voicing affects pitch, so the change (not entirely complete) of preglottalisation to voicing may well have changed the pitch of some words, but this hasn't led to any split in the tones, just a rearrangement of allophones.

Still working on Ossmalic's tone. A lot of the hard part has been getting Zompist's sound change applier to apply the changes in the intended way.

But I am also making decisions on how the tone manifests phonemically.

I have come up with a list of minimal pairs in the voicing or aspiration of stops as well as phonemic tones. I am thinking that falling and rising tones should flatten to mid tones before stops while keeping rising tones in syllables with coda sonorants and falling tones in syllables with coda fricatives. What do y'all think?
ta → tá
da → tà
sda → dá
gda → dà
tsa → tʰá
dsa → tʰà
tua → tu̯â
dua → tu̯ǎ
sdua → du̯â
gdua → du̯ǎ
tsua → tʰu̯â
dsua → tʰu̯ǎ

tan → tán
dan → tǎn
sdan → dán
gdan → dǎn
tsan → tʰán
dsan → tʰǎn
tuan → tu̯án
duan → tu̯ǎn
sduan → du̯án
gduan → du̯ǎn
tsuan → tʰu̯án
dsuan → tʰu̯ǎn

tas → tâs
das → tàs
sdas → dâs
gdas → dàs
tsas → tʰâs
dsas → tʰàs
tuas → tu̯âs
duas → tu̯às
sduas → du̯âs
gduas → du̯às
tsuas → tʰu̯âs
dsuas → tʰu̯às

taj → táj
daj → tàj
sdaj → dáj
gdaj → dàj
tsaj → tʰáj
dsaj → tʰàj
tuaj → tu̯áj
duaj → tu̯àj
sduaj → du̯áj
gduaj → du̯àj
tsuaj → tʰu̯áj
dsuaj → tʰu̯àj

Ahzoh wrote: ↑Mon Feb 05, 2024 9:03 pm Still working on Ossmalic's tone. A lot of the hard part has been getting Zompist's sound change applier to apply the changes in the intended way.

Might I use this opportunity to advertise my own sound change applier?

(It even comes with a tonogenesis example! Though probably not a plausible one.)

bradrn wrote: ↑Mon Feb 05, 2024 9:05 pm

Ahzoh wrote: ↑Mon Feb 05, 2024 9:03 pm Still working on Ossmalic's tone. A lot of the hard part has been getting Zompist's sound change applier to apply the changes in the intended way.

Might I use this opportunity to advertise my own sound change applier?

(It even comes with a tonogenesis example! Though probably not a plausible one.)

It looks very flexible. If only I did not spend 2 days learning how SCA works and getting it to do what I wanted. I might as well stick with SCA now.

EDIT: I gave it a try. I can't seem to get the dipthongs to work properly nor can I get categorical combinations to work properly like [V +Unassigned] / [V + High] / TF _

Currently, I'm working on a protolang with the vowels <i u e> [i u ə]...and the suggestion in the very enjoyable Conlang Year - https://www.quothalinguist.com/2024/01/ ... y-29-2024/ - is to let one of the vowels mutate...

I understand that the [ə] probably wouldn't mutate to [e], as that would be too close to both [ə] and to [.i.]...so I was thinking of [ə] mutating in places to [ɛ].

But would that still be too close? Might it distance itself from the other (and unmutated) [ə]s, and become [a] ?

Thank you.

keenir wrote: ↑Wed Feb 07, 2024 6:28 pm Currently, I'm working on a protolang with the vowels <i u e> [i u ə]...and the suggestion in the very enjoyable Conlang Year - https://www.quothalinguist.com/2024/01/ ... y-29-2024/ - is to let one of the vowels mutate...

I understand that the [ə] probably wouldn't mutate to [e], as that would be too close to both [ə] and to [.i.]...so I was thinking of [ə] mutating in places to [ɛ].

But would that still be too close? Might it distance itself from the other (and unmutated) [ə]s, and become [a] ?

Thank you.

Splitting /ə/ info [ə] and [a] sounds like a good idea to me; a system without any low vowels seems unstable IMO (despite the reconstructions of PIE with /e/ and /o/).

Travis B. wrote: ↑Thu Feb 08, 2024 2:11 pm

keenir wrote: ↑Wed Feb 07, 2024 6:28 pm Currently, I'm working on a protolang with the vowels <i u e> [i u ə]...and the suggestion in the very enjoyable Conlang Year - https://www.quothalinguist.com/2024/01/ ... y-29-2024/ - is to let one of the vowels mutate...

I understand that the [ə] probably wouldn't mutate to [e], as that would be too close to both [ə] and to [.i.]...so I was thinking of [ə] mutating in places to [ɛ].

But would that still be too close? Might it distance itself from the other (and unmutated) [ə]s, and become [a] ?

Thank you.

Splitting /ə/ info [ə] and [a] sounds like a good idea to me; a system without any low vowels seems unstable IMO (despite the reconstructions of PIE with /e/ and /o/).

PIE /o/ was probably low.

Are there any languages without or with few phonetic alveolar or dental stops?

I'm playing with a conlang where the single alveolar stop has lenited to [ts] in onset and coda and to [r] intervocalically, leaving [t] only in the geminated [tː], the realisation of /pt tt kt/ sequences.

Knit Tie wrote: ↑Fri Feb 09, 2024 1:12 am Are there any languages without or with few phonetic alveolar or dental stops?

I'm playing with a conlang where the single alveolar stop has lenited to [ts] in onset and coda and to [r] intervocalically, leaving [t] only in the geminated [tː], the realisation of /pt tt kt/ sequences.

Most Mekeo varieties only have [ts] as an allophone of /k/ before /i/, no other coronal stops.

Knit Tie wrote: ↑Fri Feb 09, 2024 1:12 am I'm playing with a conlang where the single alveolar stop has lenited to [ts] in onset and coda and to [r] intervocalically, leaving [t] only in the geminated [tː], the realisation of /pt tt kt/ sequences.

This sounds very similar to my dialect of English, where /t/ is [ɾ] intervocalically and [tˢ] in the onset and coda, except in consonant clusters where it’s [t̚].

bradrn wrote: ↑Fri Feb 09, 2024 3:20 am

Knit Tie wrote: ↑Fri Feb 09, 2024 1:12 am I'm playing with a conlang where the single alveolar stop has lenited to [ts] in onset and coda and to [r] intervocalically, leaving [t] only in the geminated [tː], the realisation of /pt tt kt/ sequences.

This sounds very similar to my dialect of English, where /t/ is [ɾ] intervocalically and [tˢ] in the onset and coda, except in consonant clusters where it’s [t̚].

That sounds fascinating. What dialect is this?

Knit Tie wrote: ↑Fri Feb 09, 2024 4:33 am

bradrn wrote: ↑Fri Feb 09, 2024 3:20 am

Knit Tie wrote: ↑Fri Feb 09, 2024 1:12 am I'm playing with a conlang where the single alveolar stop has lenited to [ts] in onset and coda and to [r] intervocalically, leaving [t] only in the geminated [tː], the realisation of /pt tt kt/ sequences.

This sounds very similar to my dialect of English, where /t/ is [ɾ] intervocalically and [tˢ] in the onset and coda, except in consonant clusters where it’s [t̚].

That sounds fascinating. What dialect is this?

Something between South African and Australian.