The CBB

I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/

One idea would be to introduce a quality distinction which replaces the phonemicity of the glottalisation which is subsequently lost.

Assuming vowels were tense before glottal stops, for instance:
Long unglottalised vowels could diphthongise (sth like eɪ iːə oʊ uːə), and short unglottalised vowels could lower (perhaps a e i o u → a ɛ ɪ ɔ ʏ but glottalised a → ʌ).

Of course the vowel changes could be completely different.

DesEsseintes wrote: ↑01 Oct 2020 06:04

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/

One idea would be to introduce a quality distinction which replaces the phonemicity of the glottalisation which is subsequently lost.

Assuming vowels were tense before glottal stops, for instance:
Long unglottalised vowels could diphthongise (sth like eɪ iːə oʊ uːə), and short unglottalised vowels could lower (perhaps a e i o u → a ɛ ɪ ɔ ʏ but glottalised a → ʌ).

Of course the vowel changes could be completely different.

For the most part the vowels were tense in the language, (the short vowels were more lax than the long vowels though). I was sort of thinking the opposite, for the glottalized vowels to primarily diphthongize or lower. In English (looking æ-tensing) it looks like in glottal heavy environments like /bækʔ/ <back>, /fætʔ/ <fat> we get lax [bæk̚̚] and [fæʔ]. In the Index Diachronica, it looks like glottalized vowels between "Sin Sukchu" and Guānhuà generally experience more movement than the corresponding unglottalized vowels

Right now I'm leaning towards a shift like below (with more changes happening later)
eːˀ iːˀ oːˀ uːˀ → ɛɯ̯ ɜɪ̯ ɔə̯ oʊ̯
aˀ eˀ iˀ oˀ uˀ →æ ɛ ɪ ə ɔ

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

You could introduce tone. Glottalized vowels have the nice property that they can lead to high or low-toned vowels in daughter. You could have one daughter where glottalized vowels become high-toned and another daughter where glottalized vowels become low-toned. I think something similar happened in Athabaskan, IINM. If you don't like tone, you could get rid of it by having the leftmost/rightmost high toned vowel becoming stressed and deleting all other high tones. This would give you different stress systems in the daughter languages.

Daughter A
/déːgs/
/díg/
/dóːŋ/
/hét/

Daughter B
/dèːgs/
/dìg/
/dòːŋ/
/hèt/

Omzinesý wrote: ↑28 Sep 2020 14:10 What could be the origin of retroflex affricates?

In Slavic languages, they usually derive from palatals/postalveolars after new palatalization has pushed them to retroflexes.
In Swedish, retroflexes derive from /r/ + a velar, but they are not affricates.
I think English <tr> is pronounced as an affricate in some dialects.

What I'm thinking is having them without palatals contrasting with them.

Could some kind of RUKI law apply to dental stops as well?

I have a diachronic question concerning my SCA2 GMP. I rediscovered an altlang I once started where I took Proto-Germanic and made into a Bantu-like proto-language. Now, I noticed that the changes as they are don't give the desired result, and I was thinking about switching the order of large chunks of changes, but I am afraid this will cause changes in other parts of the language, that I did not intend. The basic idea is the following. All dipthongs, tripthongs, and sequences of vowels and glides become monophthongs (Simplification). In a different change, certain vowels are deleted to create new nasal+consonant sequences (creation of prenasalized stops). I only want vowels to be deleted that were monophthongs in Proto-Germanic, but right now the GMP includes monophthongization before vowel deletion. This does not give the desired results. In between are some changes that require references to vowels, such as palatalizations and a nasal shift (where some consonants before nasalized vowels become prenasalized). I still want these changes to work as before, but I intend to switch monophthongization and vowel deletion. Do you think this will interfere with the changes in between?

My GMP in SCA2 format (I use η and ȣ for the long lax mid vowels).

My advice is just to do it the hard way and get it over with. Better than having a nagging feeling in the back of your mind that your sound changes arent quite right. p.s. i think you broke a sentence in the middle towards teh end of your post but my advice is the same either way.

Well, you can run your changes and see, can't you?

But in any case, if you're just swapping two changes, that shouldn't change other changes that happen after both of them, should it?

Thank you both. I guess switching the changes and rechecking all the vocabulary is the most sensible option. It will take some time, but I will feel better after it

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

I gather from this and your other post that you'd rather have them affect the vowels somehow, but another possibility would be having them turn into consonants, e.g. by making them geminate the following consonant, or turning the glottal element into /k/ or something to that extent.

Creyeditor wrote: ↑02 Oct 2020 08:19 Thank you both. I guess switching the changes and rechecking all the vocabulary is the most sensible option. It will take some time, but I will feel better after it

As I say, I can't see how switching the changes could possibly have any effect on palatalisation, given that both changes happen before palatalisation. What is it you think might happen?

But in any case, the SCA has a 'highlight changes from last run' option. You run it one way, switch the changes, run it again with that option ticked, and anything that changes will pop up highlighted, so you don't have to compare by hand.

At least, that's how it should work. For me, that option no longer works, which is infuriating. [I have a bad habit of doing changes by half-arsing it, then (using a sufficiently large and varied wordlist) seeing what went wrong that I hadn't thought of beforehand, and fixing the problems that show up. This relies on the 'show changes' option, because it lets me see which changes happen that I didn't intend...]. However, it may just be a problem on my end, so hopefully it'll work for you.

Right, the option does not work anymore. It's the same for me. It used to not work only in cases of huge vocabulary lists (which I usually have). I think it might be part of the new update that introduced a lot of nice things otherwise, like explicit comments in the changes.

I actually already rechecked the vocabulary and it was much less of a hassle than I thought. There were only minor problems. One was that monophthongization before the switch did not block palatalization, so nucleus inital /i/ needed to stay after the change. In the revised version it does not need to keep the /i/, since palatalization has already applied. Relatedly, the palatalization block did include a rule that deleted prevocalic word intial /i/ since monophthongization did not get rid of it. This rule blocked some monophthongizations in the simple switched version, so I simply deleted it.

Creyeditor wrote: ↑02 Oct 2020 18:04 Right, the option does not work anymore.

How odd. It works for me. Words that are different from the last run are bolded.

DesEsseintes wrote: ↑02 Oct 2020 18:58

Creyeditor wrote: ↑02 Oct 2020 18:04 Right, the option does not work anymore.

How odd. It works for me. Words that are different from the last run are bolded.

For me, ALL words are bolded, except (for some reason) the first one. Previously that problem would happen sometimes when there was an error in the soundchanges - and also if there were lines left empty in the wordlist, or sometimes if I'd copied in the wordlist. I used to sometimes have to go through the rules and the worldlist making sure there were no spaces (or hidden characters?) at the end of lines. But now it always happens, no matter how simple the rules or how short the wordlist.


[I usedn't to have the problem, as Creyeditor did, with long wordlists. But to be fair, after a certain length, the page would just become really, really, really slow (not linearly, oddly, but it would quite rapidly switch to being slow after a certain threshold), so I'd open up another window and split the wordlist into two parts. So maybe I just never managed to work with a long enough wordlist to trigger the glitch?]

Or maybe my long word lists always included empty lines, incidentally. I might as well check for hidden characters now.

Aevas wrote: ↑02 Oct 2020 08:58

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

I gather from this and your other post that you'd rather have them affect the vowels somehow, but another possibility would be having them turn into consonants, e.g. by making them geminate the following consonant, or turning the glottal element into /k/ or something to that extent.

One of the issues with syllabifying the glottal element or making it into a consonant is that words are primarily trisyllabic, so a change like /eːˀ/ → /eʔe/ → /eke/ would introduce an additional syllable on an already large word, and assimilating them into the next consonant would create additional confusing clusters and weirdness. Also problematic is the fact that glottalization results from some of those consonants, and so reglottalizing them feels like a waste of time. I have begun to think thought that I'm trying to look for a elegant solution to a problem that has no simple answer. I wanted to do something like PIE coloring, but the language's shape (and the origin of the glottalization) doesn't make that too simple.

qwed117 wrote: ↑03 Oct 2020 01:39

Aevas wrote: ↑02 Oct 2020 08:58

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

I gather from this and your other post that you'd rather have them affect the vowels somehow, but another possibility would be having them turn into consonants, e.g. by making them geminate the following consonant, or turning the glottal element into /k/ or something to that extent.

One of the issues with syllabifying the glottal element or making it into a consonant is that words are primarily trisyllabic, so a change like /eːˀ/ → /eʔe/ → /eke/ would introduce an additional syllable on an already large word, and assimilating them into the next consonant would create additional confusing clusters and weirdness. Also problematic is the fact that glottalization results from some of those consonants, and so reglottalizing them feels like a waste of time. I have begun to think thought that I'm trying to look for a elegant solution to a problem that has no simple answer. I wanted to do something like PIE coloring, but the language's shape (and the origin of the glottalization) doesn't make that too simple.

You could do something like a "mixed change", e.g. glottalised syllables could be relatively shorter than their non-glottalised counterparts, and then they could also be coloured (most likely, this might include some sort of backing and/or lowering). So you might end up with:

/eː iː oː uː/ > /eː iː oː uː/ > /eː iː oː uː/ > /eː iː oː uː/
/a e i o u/ > /a e i o u/ > /a e i o u/ > /a e i o u/
/eːˀ iːˀ oːˀ uːˀ/ > /ɛːˀ ɪːˀ ɔːˀ ʊːˀ/ > /ɛˑ ɪˑ ɔˑ ʊˑ/ > /ɛ ɪ ɔ ʊ/
/aˀ eˀ iˀ oˀ uˀ/ > /ɑˀ ɛˀ ɪˀ ɔˀ ʊˀ/ > /ɑ˘ ɛ˘ ɪ˘ ɔ˘ ʊ˘/ > /ə ə ɨ ə ɨ/

So that's lowering followed by shortening, followed by a collapse of the vowel length system into one based predominantly on quality, with some long vowels.

Salmoneus wrote: ↑02 Oct 2020 20:02

DesEsseintes wrote: ↑02 Oct 2020 18:58

Creyeditor wrote: ↑02 Oct 2020 18:04 Right, the option does not work anymore.

How odd. It works for me. Words that are different from the last run are bolded.

For me, ALL words are bolded, except (for some reason) the first one. Previously that problem would happen sometimes when there was an error in the soundchanges - and also if there were lines left empty in the wordlist, or sometimes if I'd copied in the wordlist. I used to sometimes have to go through the rules and the worldlist making sure there were no spaces (or hidden characters?) at the end of lines. But now it always happens, no matter how simple the rules or how short the wordlist.


[I usedn't to have the problem, as Creyeditor did, with long wordlists. But to be fair, after a certain length, the page would just become really, really, really slow (not linearly, oddly, but it would quite rapidly switch to being slow after a certain threshold), so I'd open up another window and split the wordlist into two parts. So maybe I just never managed to work with a long enough wordlist to trigger the glitch?]

I'm having the same error with words being bolded except the first one.

sangi39 wrote: ↑03 Oct 2020 02:39

qwed117 wrote: ↑03 Oct 2020 01:39

Aevas wrote: ↑02 Oct 2020 08:58

qwed117 wrote: ↑01 Oct 2020 05:55 I’ve been working on a diachronic language that develops glottalized vowels from a variety of sources. I don’t particularly want glottalization to remain in the language, but I also don’t want to convert it to tone, at least not yet. What could I do with the glottalized vowels?

For context, the vowels are /a e i o u eː iː oː uː/, all of which can be glottalized to my knowledge. Some example syllables with the glottalized vowels:
/deːˀgs/
/diˀg/
/doːˀŋ/
/heˀt/
They are all part of larger words

I gather from this and your other post that you'd rather have them affect the vowels somehow, but another possibility would be having them turn into consonants, e.g. by making them geminate the following consonant, or turning the glottal element into /k/ or something to that extent.

One of the issues with syllabifying the glottal element or making it into a consonant is that words are primarily trisyllabic, so a change like /eːˀ/ → /eʔe/ → /eke/ would introduce an additional syllable on an already large word, and assimilating them into the next consonant would create additional confusing clusters and weirdness. Also problematic is the fact that glottalization results from some of those consonants, and so reglottalizing them feels like a waste of time. I have begun to think thought that I'm trying to look for a elegant solution to a problem that has no simple answer. I wanted to do something like PIE coloring, but the language's shape (and the origin of the glottalization) doesn't make that too simple.

You could do something like a "mixed change", e.g. glottalised syllables could be relatively shorter than their non-glottalised counterparts, and then they could also be coloured (most likely, this might include some sort of backing and/or lowering). So you might end up with:

/eː iː oː uː/ > /eː iː oː uː/ > /eː iː oː uː/ > /eː iː oː uː/
/a e i o u/ > /a e i o u/ > /a e i o u/ > /a e i o u/
/eːˀ iːˀ oːˀ uːˀ/ > /ɛːˀ ɪːˀ ɔːˀ ʊːˀ/ > /ɛˑ ɪˑ ɔˑ ʊˑ/ > /ɛ ɪ ɔ ʊ/
/aˀ eˀ iˀ oˀ uˀ/ > /ɑˀ ɛˀ ɪˀ ɔˀ ʊˀ/ > /ɑ˘ ɛ˘ ɪ˘ ɔ˘ ʊ˘/ > /ə ə ɨ ə ɨ/

So that's lowering followed by shortening, followed by a collapse of the vowel length system into one based predominantly on quality, with some long vowels.

That could work too. I'll start deciding soon

I don't mean to interrupt, but I just realized the definite affix in Yuraalian is a heavy syllable, and because it affixes onto nouns, it affects stress placement in certain words (and vowel quality in turn). For example:

jalun [ˈʑɒln̩] '(a) road'
ol-jalun [ˈolʑəˌlun] 'the road'

Is this naturalistic? Or should I fix it somehow?

aliensdrinktea wrote: ↑14 Oct 2020 03:14 I don't mean to interrupt, but I just realized the definite affix in Yuraalian is a heavy syllable, and because it affixes onto nouns, it affects stress placement in certain words (and vowel quality in turn). For example:

jalun [ˈʑɒln̩] '(a) road'
ol-jalun [ˈolʑəˌlun] 'the road'

Is this naturalistic? Or should I fix it somehow?

I can't think of any "problems with that", in terms of naturalism. Personally, I quite like the idea.