Zompist Bboard Again

I had an idea that certain consonants could influence a vowel even if it's not directly adjacent to the vowel and this would result in the various ablaut grades in Vrkhazhian:

For example, you'd have paṭaṣ- (< puṭaṣ-) in one series of inflections but -pṭuṣ- in another series of inflections, with /a/ turning into /u/ under the influence of the bilabial. Though I don't know if velars would be i-mutating or neutral. Same with coronals. I feel like coronals (alveolar) should be neutral while back in the past there were postalveolars which were i-mutating.

However, these sorts of changes may be resisted to a degree if the vowel differences are grammatically important.

Ahzoh wrote: ↑Wed May 07, 2025 1:00 pm I had an idea that certain consonants could influence a vowel even if it's not directly adjacent to the vowel and this would result in the various ablaut grades in Vrkhazhian:

For example, you'd have paṭaṣ- (< puṭaṣ-) in one series of inflections but -pṭuṣ- in another series of inflections, with /a/ turning into /u/ under the influence of the bilabial. Though I don't know if velars would be i-mutating or neutral. Same with coronals. I feel like coronals (alveolar) should be neutral while back in the past there were postalveolars which were i-mutating.

However, these sorts of changes may be resisted to a degree if the vowel differences are grammatically important.

Just for the record, at times alveolar consonants are i-mutating in addition to postalveolar consonants, e.g. in the English right here, where e.g. /u/ > [y] after an alveolar, postalveolar, or palatal consonant (except before /l/ where no fronting occurs, or before a velar other than /l/, where then /u/ > [yu] instead).

As for velar consonants, I would make them neutral or backing myself, not i-mutating. E.g. /l/, which is velar in the dialect here, backs/prevents fronting of vowels that precede it.

Travis B. wrote: ↑Wed May 07, 2025 2:49 pm

Ahzoh wrote: ↑Wed May 07, 2025 1:00 pm I had an idea that certain consonants could influence a vowel even if it's not directly adjacent to the vowel and this would result in the various ablaut grades in Vrkhazhian:

For example, you'd have paṭaṣ- (< puṭaṣ-) in one series of inflections but -pṭuṣ- in another series of inflections, with /a/ turning into /u/ under the influence of the bilabial. Though I don't know if velars would be i-mutating or neutral. Same with coronals. I feel like coronals (alveolar) should be neutral while back in the past there were postalveolars which were i-mutating.

However, these sorts of changes may be resisted to a degree if the vowel differences are grammatically important.

Just for the record, at times alveolar consonants are i-mutating in addition to postalveolar consonants, e.g. in the English right here, where e.g. /u/ > [y] after an alveolar, postalveolar, or palatal consonant (except before /l/ where no fronting occurs, or before a velar other than /l/, where then /u/ > [yu] instead).

As for velar consonants, I would make them neutral or backing myself, not i-mutating. E.g. /l/, which is velar in the dialect here, backs/prevents fronting of vowels that precede it.

Well, alveolars are common so I don't want a majority of the vowels mutating. In fact, i sort of want to create a weak coronal-peripheral phonological behaviour contrast.

I don't know, I'm just trying to come up with semi-predictable reasons for the language's various ablaut alternations.

I have:

Transitive alternations (nonfuture > future):
a > a
a > u
a > i
Dynamic intransitive alternations (nonfuture > future):
u > a
u > u
Stative intransitive alternations (nonfuture > future):
i > a
i > i

Like I want them to be a little random, opaque, unpredictable but I don't want them to be too random, opaque, and unpredictable. At the same time I don't want them to be too predictable and transparent. It's really a fine line.

I do not want every instance of a root containing a bilabial to be of the a-u or u-u pattern nor do I want every instance of a root containing coronals to be of the a-i or i-i pattern. At the same time paṭaṣ- vs -pṭuṣ- feels right to me where paṭaṣ- vs -pṭiṣ- does not. Although, funnily enough, paraḫ- vs -priḫ- feels right to me but paraḫ- vs -pruḫ- does not, even though it has a /p/ like the other example. A lot of it is just vibes, but I do not like the uncertainty of vibes.

Ahzoh wrote: ↑Wed May 07, 2025 4:27 pm Well, alveolars are common so I don't want a majority of the vowels mutating. In fact, i sort of want to create a weak coronal-peripheral phonological behaviour contrast.

One possibility is the phenomenon called ‘prosodies’ in Chadic languages. Basically, these are word-level processes of palatalisation or labialisation, which have effects on both vowels and consonants within the word. The most extreme case is probably Moloko, which has one underlying vowel but ten surface vowels, which depend on whether the vowel is affected by labialisation, palatalisation, or is adjacent to /j/ or /w/. I can easily imagine a system like this giving rise to ablaut alternations which co-vary with consonants.

Another possibility is, quite simply, plain old vowel syncope, as happened in Germanic languages or northern Vanuatuan languages. Triconsonantal systems come from syncope anyway, so there should be some way to fit this in fairly easily.

bradrn wrote: ↑Wed May 07, 2025 10:46 pm

Ahzoh wrote: ↑Wed May 07, 2025 4:27 pm Well, alveolars are common so I don't want a majority of the vowels mutating. In fact, i sort of want to create a weak coronal-peripheral phonological behaviour contrast.

One possibility is the phenomenon called ‘prosodies’ in Chadic languages. Basically, these are word-level processes of palatalisation or labialisation, which have effects on both vowels and consonants within the word. The most extreme case is probably Moloko, which has one underlying vowel but ten surface vowels, which depend on whether the vowel is affected by labialisation, palatalisation, or is adjacent to /j/ or /w/. I can easily imagine a system like this giving rise to ablaut alternations which co-vary with consonants.

Another possibility is, quite simply, plain old vowel syncope, as happened in Germanic languages or northern Vanuatuan languages. Triconsonantal systems come from syncope anyway, so there should be some way to fit this in fairly easily.

I don't want Chaha. Nor do I want word-level secondary-articulation harmony.

Dissimilation is a possibility where, for example, bilabials disprefer /u/ and turn it into some other vowel, whereas adjacent consonants are subtly labialised with in turn labializes, and eventually backs, vowels.

Just to confirm, macrons signify long vowels, so "cāk" would be pronounce the same as cake, right? I'm a little hesitant to throw macrons around before checking.

AwfullyAmateur wrote: ↑Thu May 08, 2025 11:20 am Just to confirm, macrons signify long vowels, so "cāk" would be pronounce the same as cake, right? I'm a little hesitant to throw macrons around before checking.

Macrons can signify long vowels, yes. The <a> is cake is an /ei/, not /a:/.

Thanks.