Zompist Bboard Again

I... I'm not even going to dignify this question with a response.

KathTheDragon wrote: ↑Tue Dec 24, 2019 7:40 am I... I'm not even going to dignify this question with a response.

Why? It seems a perfectly valid question to me.

Grimm's Law is a very well-known sound change:

https://en.wikipedia.org/wiki/Grimm%27s_law

bradrn wrote: ↑Tue Dec 24, 2019 3:51 pm

KathTheDragon wrote: ↑Tue Dec 24, 2019 7:40 am I... I'm not even going to dignify this question with a response.

Why? It seems a perfectly valid question to me.

I didn't intend to aks a question, just account for my train of thought. Actually, I can think of several good reason for not packaging Grimm's law up as a single change. Firstly, the traditional formulation of the glottalic theory raises makes it a pull-push chain, and I'm not sure that they should be packaged up as single changes. Secondly, the glottalic theory raises significant questions as to its inputs. (It's even been claimed that conventional *d yields a preglottalised stop that survives into Estuarine English after short vowels.) Thirdly, if one takes voiced stops as an output, it interacts with Verner's law. What I was thinking of was the *t > [θ] part. (For the complications associated with it, it doesn't matter whether it was [t] > [θ] or [tʰ] > [θ].)

How might the numeral words influence each other phonetically/phonemically?:

https://www.frathwiki.com/Vrkhazhian#Numerals

Vijay wrote: ↑Tue Dec 24, 2019 5:06 pm Grimm's Law is a very well-known sound change:

https://en.wikipedia.org/wiki/Grimm%27s_law

You'd think so, wouldn't you? However, reading that article, I can see three or four fates for *tt - */θt/, /ss/ or /st/, and /tt/. It also states that the spirant law ‘continued to operate throughout the Proto-Germanic period’, which rather implies that it wasn’t part of Grimm's law! Th general outline is well known; it's the details that get fiddly. We've also got Kluge's law at work, which delivers voiceless geminate stops.

Morpheme-internal /tt/ is only known from the single word atta "father", which is almost certainly baby talk. I believe it should be excluded, as baby talk can defy sound changes, and because the word may not have existed at the time of PIE to begin with. So that eliminates that one. The other three outcomes, I suspect, can be explained mostly or entirely by different dates of construction ... the oldest reflex would be /ss/, but newer coinages that had not had adjacent /tt/ in PIE could have different outcomes. What looks messy can become neat when you put everything in its proper order.

Original *tt became *ss before Grimm's law took place. Secondary *tt from morpheme composition became *st as a partial regularisation of the *tt > *ss rule, which continued to be applied as a morphonological rule. *attô can have its *tt from Kluge's law, since the word is an n-stem.

bradrn wrote: ↑Tue Dec 24, 2019 3:51 pm

KathTheDragon wrote: ↑Tue Dec 24, 2019 7:40 am I... I'm not even going to dignify this question with a response.

Why? It seems a perfectly valid question to me.

The suggestion of using sound change appliers to measure anything is absurd.

Richard W wrote: ↑Tue Dec 24, 2019 5:14 pm Actually, I can think of several good reason for not packaging Grimm's law up as a single change. Firstly, the traditional formulation of the glottalic theory raises makes it a pull-push chain, and I'm not sure that they should be packaged up as single changes. Secondly, the glottalic theory raises significant questions as to its inputs. (It's even been claimed that conventional *d yields a preglottalised stop that survives into Estuarine English after short vowels.) Thirdly, if one takes voiced stops as an output, it interacts with Verner's law. What I was thinking of was the *t > [θ] part. (For the complications associated with it, it doesn't matter whether it was [t] > [θ] or [tʰ] > [θ].)

The evidence for the glottalic theory as it is usually presented is unconvincing to me, so I completely reject it. I don't really grok your objections anyway.

KathTheDragon wrote: ↑Tue Dec 24, 2019 6:29 pm The suggestion of using sound change appliers to measure anything is absurd.

It's less absurd than using just the Levenshtein edit distance to quantify the differences between languages, which does happen.

Richard W wrote: ↑Wed Dec 25, 2019 5:04 pm

KathTheDragon wrote: ↑Tue Dec 24, 2019 6:29 pmThe suggestion of using sound change appliers to measure anything is absurd.

It's less absurd than using just the Levenshtein edit distance to quantify the differences between languages, which does happen.

The fact computational linguists do things like that doesn't make them moral agents of the Good though.

Ser wrote: ↑Wed Dec 25, 2019 5:33 pm

Richard W wrote: ↑Wed Dec 25, 2019 5:04 pm

KathTheDragon wrote: ↑Tue Dec 24, 2019 6:29 pmThe suggestion of using sound change appliers to measure anything is absurd.

It's less absurd than using just the Levenshtein edit distance to quantify the differences between languages, which does happen.

The fact computational linguists do things like that doesn't make them moral agents of the Good though.

Yeah, this is not a good justification.

Is a global shift of fricative + fricative to fricative + stop reasonable? So something like ʃs to ʃt, sx to sk, xf to xp, and so on?

EastOfEden wrote: ↑Wed Dec 25, 2019 9:38 pm Is a global shift of fricative + fricative to fricative + stop reasonable? So something like ʃs to ʃt, sx to sk, xf to xp, and so on?

I already asked this:

bradrn wrote: ↑Wed Dec 18, 2019 2:06 pm I was asking whether this is plausible for all fricative sequences. In other words, [+fric] [+fric] → [+fric] [+stop] / _ (or → [+fric] [+affricate], as the case may be).

I don’t think I got an answer though.

Sound changes typically don't work in isolation. What is your phoneme inventory? How frequent are fricatives and fricative clusters? How frequent are fricative + stop clusters? If FS clusters are common, but FF clusters are rare, I can see the latter merging into the former.
But most of the time there are exceptions, or rather multiple developmnets. E.g. ʃs, sʃ could easily simplify to ʃ, s instead.

If, on the other hand, FF clusters are very frequent, maybe some of them will simplify to F, some of them will change to FS (spirant + non spirant → spirant + stop seems likely), and yet some others could remain unaffected or change in other ways.

Might it be naturalistic for a series of numerals, such as this...
tibsa, śiṭṭa, maǧla, ṣebāsa
(S-S S-S S-S S-L-S)
kurēya, tiǧta, menāpa, ḳerda
(S-L-S S-S S-L-S S-L-S)
yasīna, ʾañla, pisāwa, lamma
(S-L-S S-S S-L-S S-S)

...to influence each other to the point of becoming this?
tibsa, śiṭṭa, meǧāla, ṣebāsa
(S-S S-S S-L-S S-L-S)
kurya, tuǧta, menāpa, ḳerāda
(S-S S-S S-L-S S-L-S)
yasna, ʾañla, pesāwa, lamma
(S-S S-S S-L-S S-S)

It has a triconsonantal root system if it means anything.

bradrn wrote: ↑Wed Dec 25, 2019 11:40 pm

EastOfEden wrote: ↑Wed Dec 25, 2019 9:38 pm Is a global shift of fricative + fricative to fricative + stop reasonable? So something like ʃs to ʃt, sx to sk, xf to xp, and so on?

I already asked this:

bradrn wrote: ↑Wed Dec 18, 2019 2:06 pm I was asking whether this is plausible for all fricative sequences. In other words, [+fric] [+fric] → [+fric] [+stop] / _ (or → [+fric] [+affricate], as the case may be).

I don’t think I got an answer though.

There are three problems with [+fric][+fric] → [+fric][+stop]:

1) It seems unlikely for geminates.
2) It doesn't seem to work for the second fricative being /s/ - cited examples are sporadic and indirect.
3) I don't buy a phonetic change of (fric, fric) to (fric, plosive, fric) just on the basis of manner. Fricative plus affricate seems rather to tend to simplify to fricative plus fricative or to stop plus fricative.

I think all these answers have actually been offered.

bradrn wrote: ↑Wed Dec 25, 2019 11:40 pm

EastOfEden wrote: ↑Wed Dec 25, 2019 9:38 pm Is a global shift of fricative + fricative to fricative + stop reasonable? So something like ʃs to ʃt, sx to sk, xf to xp, and so on?

I already asked this:

bradrn wrote: ↑Wed Dec 18, 2019 2:06 pm I was asking whether this is plausible for all fricative sequences. In other words, [+fric] [+fric] → [+fric] [+stop] / _ (or → [+fric] [+affricate], as the case may be).

I don’t think I got an answer though.

It might be plausible for some fricatives, especially /θ/, but much less so for /ɸ f x h/, and unlikely for sibilants, unless there is very heavy influence from a neighbouring language where for some reason fricative + affricate sequences are common.

Or maybe something like this would be plausible:
-non-sibilant fricatives are formed phonetically from post-vocalic lenition (kápaki → káɸaxi, mátapa → máθaɸa)
-unstressed penult vowels are lost (káɸaxi → káɸxi, máθaɸa → máθɸa)
-the [x ɸ] are still allophones of [k p] which occur after vowels, so now that they are post-consonantal they revert to [kaɸ.ki] and [maθ.pa]
-[θ] becomes /s/ (máθpa → máspa)
-[ɸ x] are phonemicised somehow
In this scenario, the speakers would have to perceive [p k] as the same as [ɸ x]. It's a bit ugly, but it would appear that from the second stage to the third, /ɸx θɸ/ became /ɸk θp/. You could add affricates to the shift so you can get the /s ʃ/ in clusters.

Richard W wrote: ↑Thu Dec 26, 2019 2:37 pm 2) It doesn't seem to work for the second fricative being /s/ - cited examples are sporadic and indirect.

You might be able to make a case for a fricative dissimilation s > c from Section 3.4 of 'Oowekyala Segmental Phonology' by Darin Mathew Howe. These occur from the resolutions of /ss/ to /c/ and /ɬs/ to /ɬc/ at (someʔ) morpheme boundaries. This change doesn't occur after dorsal fricatives, and labial fricatives don't occur. As the latter change might be rather a neutralisation of the difference between stop and affricative, I am not sure it is a valid counter-argument to my disbelief.

How realistic is ʍ → hʷ → h? Also, can
ʃp ʃt ʃk → f ʃt x, i.e. can the middle cluster remain as the other two simplify?

Knit Tie wrote: ↑Sat Dec 28, 2019 11:31 am How realistic is ʍ → hʷ → h?

Very: I believe it happened in most English dialects.