Zompist Bboard Again

Knit Tie wrote: ↑Wed Nov 27, 2019 11:27 am Would people say that /u/ fortitioning to /kw/ word-finally and before voiceless consonants is a good idea?

How would this work? Surely /pu/ and /put/ wouldn’t turn into /pkw/ and /pkwt/?

bradrn wrote: ↑Wed Nov 27, 2019 5:04 pm

Knit Tie wrote: ↑Wed Nov 27, 2019 11:27 am Would people say that /u/ fortitioning to /kw/ word-finally and before voiceless consonants is a good idea?

How would this work? Surely /pu/ and /put/ wouldn’t turn into /pkw/ and /pkwt/?

Oh, right, forgot to specify that this applies only to the offglide in diphthomgs. A better way to put it would be to write

Vʊ̯ → Vkʷ / _[#,S]

Where S is a voiceless plosive or sibilant.

Knit Tie wrote: ↑Wed Nov 27, 2019 5:34 pm

bradrn wrote: ↑Wed Nov 27, 2019 5:04 pm

Knit Tie wrote: ↑Wed Nov 27, 2019 11:27 am Would people say that /u/ fortitioning to /kw/ word-finally and before voiceless consonants is a good idea?

How would this work? Surely /pu/ and /put/ wouldn’t turn into /pkw/ and /pkwt/?

Oh, right, forgot to specify that this applies only to the offglide in diphthomgs. A better way to put it would be to write

Vʊ̯ → Vkʷ / _[#,S]

Where S is a voiceless plosive or sibilant.

So you’re saying that /pau/ → /pakʷ/ and /paut/ → /pakʷt/? Sounds a bit odd, but not implausible.

Whether or not it's attested, you could explain it as /pau̯t → pawt → paɣʷt → pagʷt → pakʷt/, which looks realistic. And fortitioning offglide /u̯/ is certainly attested (e.g. Latin audiō → Aromanian avdu).

Darren wrote: ↑Wed Nov 27, 2019 11:59 pm Whether or not it's attested, you could explain it as /pau̯t → pawt → paɣʷt → pagʷt → pakʷt/, which looks realistic. And fortitioning offglide /u̯/ is certainly attested (e.g. Latin audiō → Aromanian avdu).

Thanks, that looks perfect!

What about going /pau̯t → pawt → paɣʷt → paxʷt → paxt/ instead?

Knit Tie wrote: ↑Thu Nov 28, 2019 12:28 am What about going /pau̯t → pawt → paɣʷt → paxʷt → paxt/ instead?

That would probably be more likely (having /gʷ/ before a voiceless stop isn't very common, unless it functions as an approximant). Albanian did the same thing (I think) except the final stage was xʷ → f instead of → x (e.g. Latin causa → kafshë).

Modern Greek also does this: αυτός [a'ftos]. I don't know of any examples for syllable-final yod, but there must be some.

I've definitely seen *Vw > *Vk(ʷ) before in Tibeto-Burman, some Eastern Malayo-Polynesian also have *Vy, *Vw > Vp.

dhok wrote: ↑Thu Nov 28, 2019 7:25 am Modern Greek also does this: αυτός [a'ftos]. I don't know of any examples for syllable-final yod, but there must be some.

Even if the Vj → Vʃ thing is unattested before, there certainly have been weirder examples of voicelessness spreading, like that one Native American language where /nt/ became /ht/ as the nasal became voiceless.

As an addendum to that, Icelandic also turned the old norse voiceless geminate consonants /pp tt kk/ into preaspirated clusters /hp ht hk/. This is how I derived my Vinlandic h+plosive clusters.

Yeah, voicelessness can apparently spread really well sometimes, hence all my conlang diachronic shenanigans with devoicing. Yod becoming a postalveolar fricative during devoicing isn't unrealistic, is it? I just can't find any examples of it at all.

I think it’s plausible (didn’t Vulgar Latin fortite *j to something similar?).

Knit Tie wrote: ↑Thu Nov 28, 2019 3:18 pm Yeah, voicelessness can apparently spread really well sometimes, hence all my conlang diachronic shenanigans with devoicing. Yod becoming a postalveolar fricative during devoicing isn't unrealistic, is it? I just can't find any examples of it at all.

Quite plausible. Rioplatense Spanish has basically done that, in several steps (j > ʝ > ʒ > ʃ), but you can cut some of the intermediate steps out without it being unrealistic. You can either do it through glide fortition (which is very very common) plus devoicing, or just a voiceless [j] becoming [ç] (which happens in most English dialects in words like "huge").

Këkkytir wrote: ↑Mon Nov 25, 2019 3:40 pm I have two questions:

1. Under what realistic conditions can fricatives become related stops (or nasals)? Examples: f → p, s → t, x → k. I want to get rid of most of the fricatives in my conlang somehow but I was told that an unconditional change into stops is very unlikely.

I think youve found your solution but i just wanted to share one more idea ... an asymmetrical change, where first /f s x/ all merge as /x/, and then you have one change like /x/ > /k/ (with or without intermediaries). This /k/ could vary according to surrounding vowels, so youd have extra /c/ near front vowels and extra /k/ elsewhere. Helps if you like unbalanced inventories since this will give you a lot more /k/.

Also, nobody mentioned nasals yet so ... I dont think getting nasals is very likely, ... if you dont make them into stops I would suggest approximants or simply deleting them.

Whimemsz wrote: ↑Thu Nov 28, 2019 8:23 pm

Knit Tie wrote: ↑Thu Nov 28, 2019 3:18 pm Yeah, voicelessness can apparently spread really well sometimes, hence all my conlang diachronic shenanigans with devoicing. Yod becoming a postalveolar fricative during devoicing isn't unrealistic, is it? I just can't find any examples of it at all.

Quite plausible. Rioplatense Spanish has basically done that, in several steps (j > ʝ > ʒ > ʃ), but you can cut some of the intermediate steps out without it being unrealistic. You can either do it through glide fortition (which is very very common) plus devoicing, or just a voiceless [j] becoming [ç] (which happens in most English dialects in words like "huge").

Unfortunately it doesn't apply to Vj diphthongs but onset jV, so more like maium [ˈmajjũː] > mayo [ˈmaʃo], iugum [ˈjʊgũː] > yugo [ˈʃuɣo]. I vaguely remember Nortaneous once mentioning an instance of Vj > Vʃ in one of his blogs though...

Pabappa wrote: ↑Fri Nov 29, 2019 1:26 am

Këkkytir wrote: ↑Mon Nov 25, 2019 3:40 pm I have two questions:

1. Under what realistic conditions can fricatives become related stops (or nasals)? Examples: f → p, s → t, x → k. I want to get rid of most of the fricatives in my conlang somehow but I was told that an unconditional change into stops is very unlikely.

I think youve found your solution but i just wanted to share one more idea ... an asymmetrical change, where first /f s x/ all merge as /x/, and then you have one change like /x/ > /k/ (with or without intermediaries). This /k/ could vary according to surrounding vowels, so youd have extra /c/ near front vowels and extra /k/ elsewhere. Helps if you like unbalanced inventories since this will give you a lot more /k/.

Also, nobody mentioned nasals yet so ... I dont think getting nasals is very likely, ... if you dont make them into stops I would suggest approximants or simply deleting them.

Thank you! Actually, I like this solution very much, because /k/ > /c/ before front vowels would fit well into the palatalization scheme of the language.

Ser wrote: ↑Fri Nov 29, 2019 12:23 pm

Whimemsz wrote: ↑Thu Nov 28, 2019 8:23 pm

Knit Tie wrote: ↑Thu Nov 28, 2019 3:18 pm Yeah, voicelessness can apparently spread really well sometimes, hence all my conlang diachronic shenanigans with devoicing. Yod becoming a postalveolar fricative during devoicing isn't unrealistic, is it? I just can't find any examples of it at all.

Quite plausible. Rioplatense Spanish has basically done that, in several steps (j > ʝ > ʒ > ʃ), but you can cut some of the intermediate steps out without it being unrealistic. You can either do it through glide fortition (which is very very common) plus devoicing, or just a voiceless [j] becoming [ç] (which happens in most English dialects in words like "huge").

Unfortunately it doesn't apply to Vj diphthongs but onset jV, so more like maium [ˈmajjũː] > mayo [ˈmaʃo], iugum [ˈjʊgũː] > yugo [ˈʃuɣo]. I vaguely remember Nortaneous once mentioning an instance of Vj > Vʃ in one of his blogs though...

True, but you've still got precedent for spontaneous devoicing of a palatal-region fricative (not even needing to be next to a voiceless consonant like Greek's [w] > [v] > [f] in falling diphthongs)* plus about a million precedents for fortition of glides. (And of course the alternate path offered by the English example which makes it pretty clear, although it should be intuitively obvious even if we didn't have real-world examples, that a devoiced [j] can very naturally becoming [ç] or a similar fricative.)

*Of course, now that I think of it I suppose Spanish already offered precedent for the exact same change a few hundred years earlier...

How common is /ᵐb ⁿd ᵑɡ/ without corresponding /ᵐp ⁿt ᵑk/? I have a proto-language with a fortis-lenis split in the occlusives, and in one branch the fortis nasals become prenasalized stops while the lenis nasals become regular nasals.

Further, suppose I wanted to back-derive exactly how the fortis-lenis distinction was pronounced in the occlusives given how they develop in the daughter branches. I'm having trouble thinking of how to explain this clearly and concisely (partly because I don't have the sound changes fully fleshed out), but here goes. Here, fortis consonants will be distinguished from lenis consonants with an overdot on the latter.

• What I describe above is how the nasals develop in the "northern" branch (in other words, *m *n *ŋ → /ᵐp ⁿt ᵑk/ and *ṁ *ṅ *ŋ̇ → /m n ŋ/); the stops *p *b *t *d *k *g become /p b t d k ɡ/, while *ṗ *ḃ *ṫ *ḋ *k̇ *ġ become fricative /f v θ ð x ɣ/ when lenis—pretty faithful except arguably for [ji ɨ̯ɨ wu] → /ʒi ɣɨ vu/.
• The "southern" branch preserves the fortis-lenis distinction only partially, though: the voiceless lenis stops become aspirate stops, *ḋ becomes /z/, *ṁ *ṅ → /w L/ conditionally (where L represent a liquid consonant—I have *ṅ → /r/ _r and *ṅ → /l/ where it elsewhere applies in mind), and the fortis-lenis distinction is otherwise lost.

Given the above, how might "fortis" and "lenis" be defined in the proto-language itself as opposed to in its daughters? Further, if I decide this is a branch of an even earlier proto-language, how might the distinction develop in the first place with the result that fortis consonants are favored word-initially and lenis consonants are favored word-finally (but the other way around not being phonotactically prohibited)? I should note that I have seven I miscounted—eight consonants in the proto-language that do not have a fortis-lenis distinction: *ʔ *h *s *r *l *w *ɉ *j. *ɉ is supposed to be reconstructed as /ɨ̯/, and the others I have reconstructed as in the IPA.

Actually, pretty good. Swahili lacks a prenasalized voiceless stop. In fact actually, the presence of prenasalized voiceless stop that is unusual.