Zompist Bboard Again

I will use this thread to ask a question:

In Vrkhazhian, there are two "classes" of roots, geminate and weak, that possess similar paradigms, but I'm not sure if they should just converge into one set of paradigms since both of them originate from biliteral roots anyways

for roots derived from biliteral/monosyllabic roots by simply reduplicating the last consonant (and then deleting them in some conjugations through haplology): for roots deriving from monosyllables by reinterpreting a long vowel or diphthong as a /w/ or /j/ (which was muddied by sound changes): So, yea, i don't know if i should just coalesce the two paradigms.

For comparison, this is the "full" paradigm:

jal wrote: ↑Thu Sep 05, 2019 1:39 pm

Akangka wrote: ↑Thu Sep 05, 2019 10:44 amRegularization that makes the verb less regular.

Examples? Because like was explained above, most here don't think that's possible with the regular definition of regular.

I'm not a linguist, but the claim that analogy makes Navajo verb system less regular is not mine. It's Leer's. This is an article that also quotes him.

Ahzoh wrote: ↑Thu Sep 05, 2019 8:51 pm In Vrkhazhian, there are two "classes" of roots, geminate and weak, that possess similar paradigms, but I'm not sure if they should just converge into one set of paradigms since both of them originate from biliteral roots anyways

Your examples are interesting: The paradigms seem similar enough that they may coalesce, but different enough that they may not. So I can realistically see it going either way. (Personally, I would say that the differences in the vowels are large enough to keep them separate, but as I said it could go either way.)

Ahzoh: I'm not sure what's going on here. How did *dawal become dul, but *danan become dan? It's hard to say how easy or difficult it would be to merge these if I don't understand what sound changes are operating on them already.

As for "should," that's a matter of taste. The two paradigms appear to be quite different, so there's no reason for them to be confused for one another. My usual answer when it comes to consolidating paradigms is to do it 90% of the way and leave the rest as irregularities.

How realistic is it to have /t͡s d͡z/ but not /t͡ʃ d͡ʒ/ (or any other affricatives)?

That's common.

akam chinjir wrote: ↑Fri Sep 06, 2019 3:37 am That's common.

Could you provide some examples? All the languages with /t͡s d͡z/ I found also had /t͡ʃ d͡ʒ/.

It appears that it's less common than I thought. Either that or I'm lousy at getting information out of PHOIBLE.

I haven't checked much, and the only one I've found that's a perfect match (/ts dz/ and no other affricates) is Chamorro. Malagasy has those two plus prenasalised variants, but no affricates at other places. A bunch of languages (Atayal, Chol, Hopi, Kiowa, Rukai, Wichita) have /ts/ and no affricates at other places, but no /dz/ and maybe, say, an ejective /ts'/. (I guess it's languages of the latter sort that I was thinking of.)

ts without tS is common enough. It's just the voicing contrast that's messing things up: lots of languages don't have voicing, and it tends to be areally common or uncommon which makes it less likely that the two features you want overlap. BUT. Each of these things is common individually, and there is no obvious reason why they can't co-occur. You might as well ask "is it realistic to have both click consonants and a logographic writing system?" You won't find an attestation, but there's no reason it can't happen.

akam chinjir wrote: ↑Fri Sep 06, 2019 4:04 am It appears that it's less common than I thought. Either that or I'm lousy at getting information out of PHOIBLE.

I also find PHOIBLE pretty useless for these sort of queries. I did make one program myself for more advanced queries of PHOIBLE, but I can’t seem to get it to work anymore.

I haven't checked much, and the only one I've found that's a perfect match (/ts dz/ and no other affricates) is Chamorro. Malagasy has those two plus prenasalised variants, but no affricates at other places. A bunch of languages (Atayal, Chol, Hopi, Kiowa, Rukai, Wichita) have /ts/ and no affricates at other places, but no /dz/ and maybe, say, an ejective /ts'/. (I guess it's languages of the latter sort that I was thinking of.)

I must have missed those languages when I checked last time — thank you! As I say below, the voicing isn’t that important, so I see Atayal etc. as also qualifying for my purposes.

Moose-tache wrote: ↑Fri Sep 06, 2019 4:47 am ts without tS is common enough. It's just the voicing contrast that's messing things up: lots of languages don't have voicing, and it tends to be areally common or uncommon which makes it less likely that the two features you want overlap.

That makes sense — I shouldn’t really have insisted on voicing.

Tropylium wrote: ↑Fri Sep 06, 2019 7:57 pm

expris kisuney sibrin pyena mirey u bri akar yedas
nexert pagan trey u kayrimin inuras
xaynis zignasun nix yazar si nix yazard

Could you explain what this is?

Some stuff put thru some sound changes on a whim, you can try hazarding a closer guess

akam chinjir wrote: ↑Fri Sep 06, 2019 4:04 am It appears that it's less common than I thought. Either that or I'm lousy at getting information out of PHOIBLE.

I've seen this get linked around. The search term you want is >1 +coronal;+delayed_release;-continuant;+anterior no +coronal;+delayed_release;-continuant;-anterior and no /tɕ/ and no /tsʲ/ and, which translates to "languages with more than one coronal anterior affricate and no coronal non-anterior affricates, and none of the specific consonants listed that I don't know how to make it exclude with the feature thing".

Nortaneous wrote: ↑Sat Sep 07, 2019 9:12 am I've seen this get linked around.

Thank you! That looks very useful.

Nortaneous wrote: ↑Sat Sep 07, 2019 9:12 am

akam chinjir wrote: ↑Fri Sep 06, 2019 4:04 am It appears that it's less common than I thought. Either that or I'm lousy at getting information out of PHOIBLE.

I've seen this get linked around. The search term you want is >1 +coronal;+delayed_release;-continuant;+anterior no +coronal;+delayed_release;-continuant;-anterior and no /tɕ/ and no /tsʲ/ and, which translates to "languages with more than one coronal anterior affricate and no coronal non-anterior affricates, and none of the specific consonants listed that I don't know how to make it exclude with the feature thing".

Wow, that looks so useful; I’ve been wanting something like this for ages! Pity it’s so hard to use though — for instance, I have no idea why and was made postfix.

bradrn wrote: ↑Sat Sep 07, 2019 7:28 pm for instance, I have no idea why and was made postfix.

Probably because postfix notation doesn't need parentheses, which makes it easier to parse.

Sharing is something, I guess?

And I miss looking at them. Great work as usual!

Huh!

I've been playing around a little with my old conworld, and I had the idea (which I've toyed with before) that a small, persecuted ethnic minority in one culture at the far west of the main continent (a social role sort of like the Jews with a hint of Roma, although the culture itself isn't specifically meant to evoke either) and a vast empire in the far east of the main continent should be (distantly!) related. [both the empire and the minority are descended from invading pastoralists, who I've decided should be related]

But the reason for my explanation is that I've just realised that their names, which I more or less randomly coined a good 15-20 years ago, can easily be interpreted as having the same origin in the proto-language!

The minority are the Hasuar. This is actually an exonym, but now I'll assume it's a fairly straight transcription of their endonym. The empire is Vajhoros. The -uar ending is also found in the exonyms for other tribes, and in Vajhoros several regions have names with -or, so I thought both might be related morphemes.

It turns out that, leaving the -os aside for a moment, the empire's name probably descends from something like wosi-wora. Now I'm thinking that this might actually represent an earlier kwosi-kwora. From this we eventually get wešwer, and due to the stress being on the first syllable this gives different outcomes for the two we- sequences: važor, which is what the spelling 'vajhor' indicates.

Meanwhile, however, although I haven't worked out anything about Hasuar yet, it's easy to imagine a sort of Germanic shift: kwosi-kwora > kʰosi-kʰora > xosi-xora. There's then a clockwise chain shift (which I know is less common than anticlockwise, but bear with me...): xosixora > xasuxar(e). Whether they actually call themselves the Xasuxar, and 'Hasuar' is a mishearing by their neighbours, or whether there's been a further lenition of x>h and then >0 intervocalically, I don't know, but either seems plausible.

Anyway, I think the original endonym for the parent tribe must have been kwosi. The protolanguage word kwora meant somethig like 'tribe', and the suffix retains this meaning for the Hasuar; in Vajhoros, it's instead gone 'tribe' > 'area controlled by a tribe' > 'area'. However, I suspect that before that semantic shift occured, an -os suffix was added specifically meaning 'home' or the like. [this could come from protolanguage kwosa, which obviously looks like it may somehow be related to the original tribal endonym in a plausible but unrecoverable way. Alternatively, something like CaCusai would be possible. ]


In any case, this admittedly isn't exciting to anyone but me, and mostly demonstrates that teenage me perhaps had poor imagination when it came to coining 'random' words. But it's a coincidence that's very pleasing to me!