Sound Change Quickie Thread

[Nortaneous]

Yes. Some people write [ð̞] (cf Danish), but I think this is unnecessary because [ɹ] (or maybe [ɹ̪] if you want to be pedantic) does the job just fine.

except [ð̞] could be dental or interdental, whereas [ɹ̪] has to be dental and [ɹ] is almost always bunched r

[finlay]

[ɹ] is almost always bunched r

only in american english, not in the ipa

also the dental diacritic isn't specifically not interdental, there is nothing anywhere that says that. there is an interdental diacritic but only in extipa and only if you really need to be more specific.

[Risla]

Not quite a sound change, but I'm wondering if the vocalic epenthesis rule (especially for loanwords) I have in South Eresian is naturalistic, since it's somewhat harmonic and the language otherwise lacks vowel harmony. I think it's probably fine, just making sure. It's a height distinction, and the two potential vowels are /a/ and /e/, depending on if the preceding vowel is [-low] or [+low].

So basically, to randomly borrow a few words from Nalchast, a word like [ˈmɑrzdo] would be borrowed as [ˈmɑːɾasto], whereas [nexˈtræld] would be borrowed as [ˈneːketɬaltɑ].

[Jetboy]

Two very quick questions:

Firstly, is it realistic to have /f/ and /θ/ become /p/ and /t/, respectively, before /s/?
Secondly, is the change /z/ (> /ɾ/ or /ɹ/ )> /l/ plausible if the language already has /r/ and /l/?

[CaesarVincens]

Yes and yes.
/f/ and /θ/ to /p/ and /t/ before /s/ seems very plausible to me - a type of dissimilation perhaps.
/z/ (or even /s/) to /r/ with a merger of existing /r/ is attested, as are mergers of /r/ and /l/.

[Jetboy]

Yes and yes.
/z/ (or even /s/) to /r/ with a merger of existing /r/ is attested, as are mergers of /r/ and /l/.

Other way around; /z/ would become /l/, possibly by way of a dental/alveolar tap, flap, or approximant, in a language that already had /l/ and a dental/alveolar trill.

[8Deer]

Is metathesis a plausible way to get rid of ejective consonants? So that [t'a] becomes [taʔ]?

[Astraios]

Is metathesis a plausible way to get rid of ejective consonants? So that [t'a] becomes [taʔ]?

Doesn't sound likely to me, especially not in an open syllable. Ejectives are so very stuck-together that they aren't really clusters, they're coarticulated.

[finlay]

[tʼ] and [tʔ] aren't the same thing, not really. I'd say what you're more likely to have is the ejectives becoming another kind of plosive, although i've heard tell that it could be anything of plain, voiced or aspirated that it would end up as.

In my conlangs I got rid of them by merging them with the plain plosives.

[8Deer]

Doesn't sound likely to me, especially not in an open syllable. Ejectives are so very stuck-together that they aren't really clusters, they're coarticulated.

[tʼ] and [tʔ] aren't the same thing, not really. I'd say what you're more likely to have is the ejectives becoming another kind of plosive, although i've heard tell that it could be anything of plain, voiced or aspirated that it would end up as.

Ok, thanks. I was just trying to think of something interesting to do with them. I remember reading about Ket qiʔy and Athabaskan q'ey, both meaning birch bark, and thought I would go with that.

[Thomas Winwood]

If a language has ejectives, and somehow loses them, what would the ejectives usually turn into?

Thanks in advance.

Implosives or modal voiced consonants are both possible, as are just plain voiceless.

[Cedh]

Is metathesis a plausible way to get rid of ejective consonants? So that [t'a] becomes [taʔ]?

I could actually see this happening via creaky-voiced vowels at an intermediate stage, i.e. [t'a] > [ta̰] > [taʔ].

(My own conlang Tmaśareʔ does something similar for aspiration, with [tʰa] > [tḁ] > [tah] before voiceless consonants.)

[Risla]

I would think it would be possible for [tʼ] to become [tʔ], and then have an epenthetic vowel inserted.

[sirdanilot]

Yes. Some people write [ð̞] (cf Danish), but I think this is unnecessary because [ɹ] (or maybe [ɹ̪] if you want to be pedantic) does the job just fine.

except [ð̞] could be dental or interdental, whereas [ɹ̪] has to be dental and [ɹ] is almost always bunched r

How can you even pronounce that sound? And I am not too sure what exactly is meant by a 'bunched r': as far as I know, it's just an [ɹ] (though admittedly, English [ɹ] is different from the [ɹ] that is found in some dialects of Dutch, such as Leids, an urban Southern Hollandic dialect spoken in and around Leiden; these sound like something is happening with the palate as well)

As for the ejectives; a simple merger to produce plain voiceless stops seems the most plausible. It depends on what else you have. A system with voiced, voiceless and ejective stops is quite stable, I think; perhaps ejective -> aspirated is a way out. If you don't have voiced stops, you can make the voiceless ones voiced and the ejective ones voiceless.

Glottalization also seems to have to do something with stress; perhaps you can get rid of ejective stops when they become unstressed due to some morphological process. For example, if you have stress word-initially, but a prefix gets tacked on a word, the ejective stop becomes plain.

[finlay]

Bunched R is an R with secondary accompaniments, like velarisation or pharyngealisation. Many who have it don't have (strong) alveolar contact here, either. It's more common in America. In English [ɹ] is used as a shorthand because people are inconsistent in their use of one form or the other and the exact configuration of the tongue is often a bit nebulous and difficult to notate with a discrete system like the IPA. However, that absolutely does not preclude it from being used as an alveolar or dental approximant without secondary accompaniments, which is what Nortaneous is implying – although yes, it is infelicitous to L1-English learners who will pronounce it like their /r/ which is almost never a true [ɹ].

On that note, I've met about one person ever with a true [ɹ]; I have [ɻʷ] most often, most Brits have some form of [ɹʷ] or [ɻʷ], and Americans are more likely to have bunched R. Some Scots have [ɾ] and [r], and many English now have [ʋ].

There's also something to do with rhotics, which is the most nebulous phonetic category ever and difficult to define, but sometimes includes a lowered F3, if you know what that means – the secondary accompaniments often exaggerate this, which is why they come about. An [ɹ] without them might not have them. [ð̞] emphasises that it doesn't have these and therefore isn't an "R Sound". But as I say the category of "R Sounds" or rhotics is a bit nebulous and is often based on spelling and history instead of phonetics – like, American [ɾ] isn't a rhotic because it's actually /d/ or /t/.

The Holland [ɹ], IIRC, isn't quite "[ɹ]", strictly speaking either, because as you say it involves something with the palate. I think this counts as bunched R, but it's different from the American one. [ɹ] is certainly adequate for writing it phonetically, though.

[Mbwa]

I've got a vowel shift, but in order to preserve the inventory I want, the order needs to be kinda weird.

The starting inventory is /a e i o u a: e: i: o: u:/, with this vowel shift:

i: u: > ai au
e o > { a
e: o: > i u
i u > e o
a: a > a @

to end up with /a { e i o u @ ai au/. The order of changes seems to be a little weird to me. We have breaking of long high vowels, lowering of mid short vowels, then raising of mid long vowels before the short high vowels lower (what happens to /a: a/ doesn't really matter). Is that weird?

[Nortaneous]

In English [ɹ] is used as a shorthand because people are inconsistent in their use of one form or the other and the exact configuration of the tongue is often a bit nebulous and difficult to notate with a discrete system like the IPA. However, that absolutely does not preclude it from being used as an alveolar or dental approximant without secondary accompaniments, which is what Nortaneous is implying – although yes, it is infelicitous to L1-English learners who will pronounce it like their /r/ which is almost never a true [ɹ].

Except nobody uses IPA exactly as it's set out; how many people use <a> for a front vowel, as opposed to a central one?

Also, canIPA uses <ɹ> for bunched R, and some sort of odd 2-shaped thing for a true alveolar approximant.

The Holland [ɹ], IIRC, isn't quite "[ɹ]", strictly speaking either, because as you say it involves something with the palate. I think this counts as bunched R, but it's different from the American one.

Sound samples? That's probably postpalatal R (sounds like [ɨ̯]; listen to some Fionn Regan songs for examples), which is also pretty common in America. (I wonder if there's a pattern to the variation between pharyngeal (bunched) and postpalatal R in AE. Seems like bunched R is associated with the South and Midwest (I've seen it called "Texas R" once), whereas postpalatal R shows up more in the Northeast, especially among younger females.

[finlay]

lol, i wouldn't use canipa to support your argument, since i think it's bullshit....

[Cedh]

I've got a vowel shift, but in order to preserve the inventory I want, the order needs to be kinda weird.

The starting inventory is /a e i o u a: e: i: o: u:/, with this vowel shift:

i: u: > ai au
e o > { a
e: o: > i u
i u > e o
a: a > a @

to end up with /a { e i o u @ ai au/. The order of changes seems to be a little weird to me. We have breaking of long high vowels, lowering of mid short vowels, then raising of mid long vowels before the short high vowels lower (what happens to /a: a/ doesn't really matter). Is that weird?

It's not all that weird. Short vowels get lowered, long vowels get raised, basically. What you should keep in mind though is that with the order you've listed, [e: o:] actually become [e o], merging with original short , and that original short [o] ends up as [ə], merging with short [a].

If you want to keep these separate, I'd list the changes as follows:
[i: u:] > [ai au]
[e: o:] > [i: u:]
> [e o æ a ə]
[i: u: a:] >

[sirdanilot]

A sample of Leiden Dutch: http://www.youtube.com/watch?v=OQUH4di3OOU . Leids is the only Dutch dialect in which every /r/ is realized as [ɹ]; most other Holland dialects only have it in coda, or not at all.

I haven't heard this kind of [ɹ] in any dialect of English, I think.

[Tropylium⁺]

Doesn't sound likely to me, especially not in an open syllable. Ejectives are so very stuck-together that they aren't really clusters, they're coarticulated.

[tʼ] and [tʔ] aren't the same thing, not really. I'd say what you're more likely to have is the ejectives becoming another kind of plosive, although i've heard tell that it could be anything of plain, voiced or aspirated that it would end up as.

Ok, thanks. I was just trying to think of something interesting to do with them. I remember reading about Ket qiʔy and Athabaskan q'ey, both meaning birch bark, and thought I would go with that.

IIRC Vajda analyses Ket /ʔ/ as having a prosodic origin (low tone in original open syllables or something along those lines).

[Qwynegold]

A sample of Leiden Dutch: http://www.youtube.com/watch?v=OQUH4di3OOU .

It sounds like he's imitating American English.

I have question. I haven't been thinking about having this in any particular conlang, I'm just curious to know. Could voiceless consonants of those types that are rare in the world, like [r_0 n_0 j_0] and whatever, turn into breathy voiced consonants? I was trying to pronunce these things but it's so damn hard. Sometimes it seems to be become breathy or something for me.

[finlay]

A sample of Leiden Dutch: http://www.youtube.com/watch?v=OQUH4di3OOU . Leids is the only Dutch dialect in which every /r/ is realized as [ɹ]; most other Holland dialects only have it in coda, or not at all.

I haven't heard this kind of [ɹ] in any dialect of English, I think.

I wouldn't be particularly surprised to hear it in English, actually. I can't think if I specifically know of a place or accent which would actually include it, though, it just wouldn't necessarily sound out of place... It does kind of sound American to me too.

I'm starting to wonder if it's in my idiolect, but there's too much observers' bias at this point.... so, I doubt it.

Leiden is a nice town, incidentally.

[sirdanilot]

A sample of Leiden Dutch: http://www.youtube.com/watch?v=OQUH4di3OOU .

It sounds like he's imitating American English.

I have question. I haven't been thinking about having this in any particular conlang, I'm just curious to know. Could voiceless consonants of those types that are rare in the world, like [r_0 n_0 j_0] and whatever, turn into breathy voiced consonants? I was trying to pronunce these things but it's so damn hard. Sometimes it seems to be become breathy or something for me.

He isn't, this is the authentic Leids. Other than the weird /r/, many of its features are typical for Southern Hollandic city dialects. Nowadays, Leids is mostly spoken by people from the lower social classes and of course older people, a process happening to many dialects in the Netherlands.

I think voiceless nasals are reasonably stable, and don't really sound breathy. The only thing that's unstable about them is that it's hard to differentiate between various voiceless nasals. Welsh even has voiceless nasals as the product of assimilation in certain positions.

/j_0/, other than it being hard to explain how it would appear in any phonology in the first place, sounds like it could easily undergo fortition to /ç/. That leaves only [r_0] as likely to produce breathy voicing.

[Astraios]

I wouldn't be particularly surprised to hear it in English, actually. I can't think if I specifically know of a place or accent which would actually include it, though, it just wouldn't necessarily sound out of place... It does kind of sound American to me too.

My first thought was Cornish. West Country-ish.