Zompist Bboard Again

Noun classes!!!

So nice to have time to sit & think about conlangs a bit.

Ares Land wrote: ↑Sun Dec 27, 2020 4:13 pm For an extra consonant, how about /kw/? Or /c/, for symmetry?

Nah, not /c/ in such a small inventory. And I already have /tS/. But I have been thinking about /k_w/...

Unrelated to the previous thing, I've been working on a branch of a language family. I keep changing my mind about how to do some lenition rules. Right now I have decided about these rules in this order:

P > C[+affricate] / V(S)_V (S = semivowel)
C[+affricate] > F / V(S)_V
F[+voiced] > C[+approximant] / V(S)_V

It would lead to these changes in that environment:

p > f (> W > w)
b > w
t_d > s_d
d_d > l_d
k > x (> x\ > S)
g > w, j
ts_a > s_a (> s_d)
dz_d > l_d
tK > K
dK\ > l_a
tS > S
dZ > j

I had thought that the language would at one point distinguish between dental-laminal and alveolar-apical fricatives and affricates. The dental ones would be made with the tongue tip pointing towards the lower teeth, and the tongue bunched, so that a narrow channel is formed all the way between the tongue and the upper teeth, continuing a bit along the alveolar ridge. I was worried that this might be implausible, especially when contrasting with the alveolar-apicals. But then I read these things on WP:

The dental consonants [in Melpa] are "distributed", with closure along a considerable distance of the vocal tract. This presumably means that they are laminal, and that the alveolar consonants are apical. The dental consonants have palatalized allophones in free variation.

Most dialects in the Bengali–Assamese continuum distinguish between dental–laminal alveolar stops and apical alveolar stops.

I would then merge these two types of consonants as something that's produced like /s/ in Mandarin: with the tongue tip pointing towards the lower teeth, but without bunching the tongue.

I would also end up with the same distinction for /l/. I would merge those too, but now when I was trying to sound out the dental-laminal l, it sounded a little like [K]. So maybe that could happen in one daughterlang.

I revised Azgomennes' phonology! This is a big deal for me. I've revised everything about it - morphology, lexicon, syntax (although I normally don't get that far) but kept the same phonology all throughout every time I redo the grammar. This is the first major change to the phonology I've made in over 20 years.

One goal I have always had in mind with Azgomennes is a phonemic transcription that requires no special characters. I've had to put aside this idea for the vowels: I've gone from a regular stress rule and a five vowel /a e i o u/ system to /a e i ɪ o ɔ u ʊ/, which I'm writing as <a e i ì o ò u ù>. Stressed vowels heighten, so stressed /o/ is [ʊ] and stressed /ù/ is u. I have now created a dilemma for myself - not only do I have to mark vowel quality but ideally stress as well, in an orthography that is meant to look "friendly.' I haven't quite settled on the best way to do that yet.

Consonants have changed as well. They are now <b c d f g h j k l m n r s t v x z> where <c> is /t͡ʃ/, <j> is /ʝ/, <x> is /ʃ/, <h> is /x/ and worst of all, <z> is t͡θ̪. I know that I am making real problems for myself here, but (a) whatever, I can do what I want and (b) it's not as though natural language transcription systems don't have these kinds of quirks. At least now i have the consonant inventory I think I want (it makes historical sound changes easier) and I still can stay within the bounds of an ASCII character set.

The flow-on effects for this are going to be big: in particular, I've just outdated my entire lexicon and will have to redo it. But I think taking this bold step gets me closer to where I want to be. Once I've adopted the lexicon and morphology (there are a bunch of morphophonological changes that I'll need to redo as well) I think I will be much closer to something that I'm happy with.

Of course, the language itself now has to change names - to Asgomennes! That's the third change it's had in about 20 years.

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pm That's the third change it's had in about 20 years.

That's a lot!
changes major details about the same language several times a year every year, with up to four or five drafts of the orthography before my current version

mèþru wrote: ↑Wed Dec 30, 2020 9:06 pm

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pm That's the third change it's had in about 20 years.

That's a lot!
changes major details about the same language several times a year every year, with up to four or five drafts of the orthography before my current version

Haha, just the *name* which I always tried to keep consistent, but sometimes could not get away with not changing. It has an out of date grammatical ending -es when I was in a Euro phase that I now have had to explain away!

Oh I forgot to mention above, <t> is now laminal, while <d> is apical. I'm having to work hard to justify that voicing contrast on top of the place contrast, but I didn't want either to lose voicing or to have four stops in those positions because I don't think I have enough unaccented characters for that.

So Haleza Grise wrote: ↑Wed Dec 30, 2020 9:14 pm Oh I forgot to mention above, <t> is now laminal, while <d> is apical. I'm having to work hard to justify that voicing contrast on top of the place contrast, but I didn't want either to lose voicing or to have four stops in those positions because I don't think I have enough unaccented characters for that.

You could perhaps have fortition of [dʒ] > [d̺] in some earlier stage which, being perhaps very common, dragged the other [d] along with it?

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pm One goal I have always had in mind with Azgomennes is a phonemic transcription that requires no special characters. I've had to put aside this idea for the vowels: I've gone from a regular stress rule and a five vowel /a e i o u/ system to /a e i ɪ o ɔ u ʊ/, which I'm writing as <a e i ì o ò u ù>. Stressed vowels heighten, so stressed /o/ is [ʊ] and stressed /ù/ is u. I have now created a dilemma for myself - not only do I have to mark vowel quality but ideally stress as well, in an orthography that is meant to look "friendly.' I haven't quite settled on the best way to do that yet.

My suggestion: use ⟨a e i ì o ò u ù⟩ for the unstressed vowels, and ⟨á é í î ó ô ú û⟩ for the stressed ones. Still pretty easy to type (you can do it on a US-international keyboard), while retaining ‘friendliness’ and avoiding any particularly weird characters.

Consonants have changed as well. They are now <b c d f g h j k l m n r s t v x z> where <c> is /t͡ʃ/, <j> is /ʝ/, <x> is /ʃ/, <h> is /x/ and worst of all, <z> is t͡θ̪. I know that I am making real problems for myself here, but (a) whatever, I can do what I want and (b) it's not as though natural language transcription systems don't have these kinds of quirks. At least now i have the consonant inventory I think I want (it makes historical sound changes easier) and I still can stay within the bounds of an ASCII character set.

Personally, I think ⟨z⟩ for /t͡θ̪/ is perfectly fine, considering ⟨z⟩ has been used for both /ts/ and /θ/ before.

mèþru wrote: ↑Wed Dec 30, 2020 9:06 pm

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pm That's the third change it's had in about 20 years.

That's a lot!
changes major details about the same language several times a year every year, with up to four or five drafts of the orthography before my current version

Yep, that sounds about right for me as well. (If you look closely at my posts to the Conlang Fluency Thread, you’ll notice that none of them have quite the same grammar, due to all the changes I make.)

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pmI revised Azgomennes' phonology! This is a big deal for me. I've revised everything about it - morphology, lexicon, syntax (although I normally don't get that far) but kept the same phonology all throughout every time I redo the grammar. This is the first major change to the phonology I've made in over 20 years.

If it isn't broken, you don't need to fix it too much, ehehehehehehe.

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pmOne goal I have always had in mind with Azgomennes is a phonemic transcription that requires no special characters. I've had to put aside this idea for the vowels: I've gone from a regular stress rule and a five vowel /a e i o u/ system to /a e i ɪ o ɔ u ʊ/, which I'm writing as <a e i ì o ò u ù>. Stressed vowels heighten, so stressed /o/ is [ʊ] and stressed /ù/ is u. I have now created a dilemma for myself - not only do I have to mark vowel quality but ideally stress as well, in an orthography that is meant to look "friendly.' I haven't quite settled on the best way to do that yet.

Depending on the phonotactics, perhaps a digraphic orthography might help?

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pmConsonants have changed as well. They are now <b c d f g h j k l m n r s t v x z> where <c> is /t͡ʃ/, <j> is /ʝ/, <x> is /ʃ/, <h> is /x/ and worst of all, <z> is t͡θ̪. I know that I am making real problems for myself here, but (a) whatever, I can do what I want and (b) it's not as though natural language transcription systems don't have these kinds of quirks. At least now i have the consonant inventory I think I want (it makes historical sound changes easier) and I still can stay within the bounds of an ASCII character set.

It depends on how transparent you want things to be for your audience. For the stories I write, I usually spell out internal names and concepts in a way I think is transparent for English-speakers, even if it means not marking pitch accurately (cf. the name of the language on which I'm currently working, which is Ifsumé in English text — from a word in one of its daughter languages whose pitch pattern is better represented Ifsúme, because the thing couldn't be bothered having an internal name); sometimes, I find, you simply have to compromise...

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pmThe flow-on effects for this are going to be big: in particular, I've just outdated my entire lexicon and will have to redo it. But I think taking this bold step gets me closer to where I want to be. Once I've adopted the lexicon and morphology (there are a bunch of morphophonological changes that I'll need to redo as well) I think I will be much closer to something that I'm happy with.

Of course, the language itself now has to change names - to Asgomennes! That's the third change it's had in about 20 years.

Ah, the never-ending quest for satisfaction, or something approaching it.

So Haleza Grise wrote: ↑Wed Dec 30, 2020 9:14 pm Oh I forgot to mention above, <t> is now laminal, while <d> is apical. I'm having to work hard to justify that voicing contrast on top of the place contrast, but I didn't want either to lose voicing or to have four stops in those positions because I don't think I have enough unaccented characters for that.

/t d/ contrasting in passive articulator as well as voice is well attested from Austronesian and to some extent Papuan; there's also Pohnpeian where <d t> /t̪ t̠/.

Happy New Years, everyone!

(will also wish you the same on Lunar New Year)

So Haleza Grise wrote: ↑Wed Dec 30, 2020 8:48 pm I revised Azgomennes' phonology! This is a big deal for me. I've revised everything about it - morphology, lexicon, syntax (although I normally don't get that far) but kept the same phonology all throughout every time I redo the grammar. This is the first major change to the phonology I've made in over 20 years.

kudos on sticking with it for so long; I hope its more what you wanted now.

One goal I have always had in mind with Azgomennes is a phonemic transcription that requires no special characters. I've had to put aside this idea for the vowels: I've gone from a regular stress rule and a five vowel /a e i o u/ system to /a e i ɪ o ɔ u ʊ/, which I'm writing as <a e i ì o ò u ù>.

Maybe use underscores?

Nortaneous wrote: ↑Wed Dec 30, 2020 10:34 pm

So Haleza Grise wrote: ↑Wed Dec 30, 2020 9:14 pm Oh I forgot to mention above, <t> is now laminal, while <d> is apical. I'm having to work hard to justify that voicing contrast on top of the place contrast, but I didn't want either to lose voicing or to have four stops in those positions because I don't think I have enough unaccented characters for that.

/t d/ contrasting in passive articulator as well as voice is well attested from Austronesian and to some extent Papuan; there's also Pohnpeian where <d t> /t̪ t̠/.

I'm becoming more influenced by Papuan in a vague way at the moment, so this very useful, thanks.

Rounin Ryuuji wrote: ↑Wed Dec 30, 2020 9:43 pm You could perhaps have fortition of [dʒ] > [d̺] in some earlier stage which, being perhaps very common, dragged the other [d] along with it?

, so this is very useful to know.

Possibly, although the development of affricates/palatals at the moment is a little messy. I'll need to think about it some more.

bradrn wrote: ↑Wed Dec 30, 2020 10:07 pm My suggestion: use ⟨a e i ì o ò u ù⟩ for the unstressed vowels, and ⟨á é í î ó ô ú û⟩ for the stressed ones. Still pretty easy to type (you can do it on a US-international keyboard), while retaining ‘friendliness’ and avoiding any particularly weird characters.

Thanks, I think I will end up with something that is fairly close to that, although I'm still deciding..

Yep, that sounds about right for me as well. (If you look closely at my posts to the Conlang Fluency Thread, you’ll notice that none of them have quite the same grammar, due to all the changes I make.)

At least it's practice and vocabulary-building!

Happy new year / Creče lebe zon to all.

So Haleza Grise: Have you been working continuously on that conlang, little by little, or have you had long breaks from it?

Some vowel developments I've been thinking about (in convenient order, not chronological):

'i > 'i
'i: > 'i:
i > i
i: > i
'e > 'i
'e: > 'i:
e > e_↓ (I can't remember XSAMPA for lowered diacritic)
e: > i
'a > 'e_↓
'a: > 'ä: (central vowel)
a > ä
a: > ä
'u > 'u
'u: > 'u:
u > u
u: > o_↓
'o > 'u
'o: > 'u:
o > ä
o: > o_↓

After the above changes, we also have:
i > e / j_, _j
u > o / w_, _w
e > o / _w
(all of these when occurring in the same syllable)

I broke 534 words on the (third? Fourth? Fifth?) version of Classical Salvian, and started a whole new language based on an Anglo-Saxon/Frisian creole!

I am happy to report that I've met my goal of reaching 2500 words for Keševan by the end of the year. I've also been dabbling in Inkscape and finally familiarizing myself with how vector graphics work with the aim of creating a map of a conworld city, which in turn has helped fuel the creation of new urban words for the lexicon.

Classical Salvian: 549/2,500
Irthironian: 90/500

Yesterday I finished translating ‘The North Wind and the Sun’ into my current conlang! Admittedly not a particularly impressive feat, but that still makes it the longest text I’ve ever translated into any of my conlangs. (Mostly because my perfectionism means that I end up searching the literature for days on end before I add a new grammatical feature to my conlang, so it can take me quite a while to translate the more complicated sentences.) Now that’s done, I intend to clean it up a bit at some point and post it here (with commentary).

Also, I’ve been playing around with deriving a daughter language from this conlang to test a new SCA I’m making. I’m particularly pleased with some stem alternations this has given me:

əfɔ / ɛfyə
fɛw / fɛwše
nesi / nɛšə
ɔsæn / ɔsne
səw / sɛw
wel / weltse
yusɔ / yošə

(All these are verbal IMPF / PFV stems.)

In general, I quite like this aesthetic, which is quite different to any of the conlangs I’ve made before. To illustrate, here’s what happens to the first few sentences of the aforementioned translation when run through the sound changes:

*Sasat thaŋmu tlaquf Qalit thaŋ waqli gilut fawetlli, ‘naqeŋ khayiqenebey tshaal pawtiq?’, qaŋethinga yusaa tlaquf lamaq. Ŋiiyusaye wiiletsi fawetli tsah naaqa waq qaŋeth siifa lamaq, ŋay nii khayiq.

Sæs sɔmm šɔ Æl sæŋ wæl yel fɛwlle, ‘nɛəŋ kæyənəbəy sɔl pæss?’, ɛŋseŋæ yusɔ šɔ læm. Ŋiyošə weltse fɛwše tsæ nɔ wæ ɛŋ sef læm, ŋæy ni xæy.

(Admittedly this would most probably be thoroughly ungrammatical and anachronistic in the descendant, which I plan to be much more agglutinative than this, but it shows the aesthetic clearly enough.)

Qwynegold wrote: ↑Thu Dec 31, 2020 1:14 pm So Haleza Grise: Have you been working continuously on that conlang, little by little, or have you had long breaks from it?

I have had breaks in there definitely. I find it hard to find time to do some serious work, I can basically only find time outside of family and work so that's not very much. But mostly I get tempted to scrap everything and start over. So I don't have much to show for all the time!

Underscores are useful for showing underlying stress patterns in root words. More work to do there.

I have translated a decently complex phrase in a new conlang

Birds are a group of vertebrates which evolved from dinosaurs.
darha a kai uhanr so bak a so kamat yi na fet dinosaur so er i

darha

bird

a

PLUR

kai

have

uhanr

spine

so

of

bak

animal

a

PLUR

yi

to.be

so

of

kamat

group

na

which, such

fet

3P.SING.INF

dinosaur

dinosaur

so

of

er

come

i

past

birds a spine-having group of animals is, which it of dinosaur came.

and I got a new way to make relative clauses out of it

So Haleza Grise wrote: ↑Sun Jan 03, 2021 10:49 pmBut mostly I get tempted to scrap everything and start over. So I don't have much to show for all the time!

Ah, it seems like a lot of conlangers have that problem. You need to stop doing that!